W'Hi'lwvhpp

 

www.“

 

 

...~ ,
4:... r

.mmﬁ
.anuamm... :

crilivt timrn?

.3

a. ”5??qu Ct
.. 3.3.1:.

56.24.”...
t 1.. 3.}.

.
v:.1...! .

... .. 4..

r: |.

. .2 4

\ .

1e xllh
\t..l fl
4.13:2)... .1533...)

2.. ginngng

1 v13
.{.4“!.. vii ‘tlu ‘-
. enzfv a. 3.: I. 3 . . 2&0an
» . I .51.... 3|! ."ni..tv

» 1!:- arVi Iii—9K»

 

 

mm LIBRARY
3 Michigan State
100 2 University

 

 

 

This is to certify that the
dissertation entitled

AN INTERVENTION-BASED EXAMINATION OF
WORK-FAMILY DECISION-MAKING SELF-EFFICACY AND
ANTICIPATED WORK—FAMILY CONFLICT

presented by

ALYSSA JILL FRIEDE

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Psychology

 

 

///// /:y///

' ajor Professor” 3/ aure

445.0?

 

Date

MSU is an affirmative-action, equal-opportunity employer

--»-—-ur-‘—--—-»-.—.-.-—.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K:lPro;IAcc&Pres/ClRC/DateDueindd

AN INTERVENTION-BASED EXAMINATION OF
WORK-FAMILY DECISION-MAKING SELF-EFFICACY AND
ANTICIPATED WORK-FAMILY CONFLICT
By

Alyssa Jill Friede

A DISSERTATION

Submitted to
Michigan State University
in partial. fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

2008

ABSTRACT
AN INTERVENTION-BASED EXAMINATION OF
WORK-FAMILY DECISION-MAKING SELF-BF F ICACY AND
ANTICIPATED WORK-FAMILY CONFLICT
By '
Alyssa Jill Friede
These is considerable research about the conﬂict that individuals experience

managing the demands of work and family roles. However, this research fails to examine
the experiences of professional students before they transition into work and family roles.
Furthermore, there is little research on the effectiveness of attempts to help students with
this transition. In the current study, I investigate the anticipated work-family conﬂict of
medical students. I examine the effectiveness of an intervention designed to decrease
anticipated work-family conﬂict by increasing work-family decision-making self-
efﬁcacy. The ﬁndings from this study suggest that both the work-family intervention and
the control intervention are effective in enhancing work-family decision-making self-

efﬁcacy and decreasing anticipated time-based work-family conﬂict. Implications for

ﬁJture research and practice are discussed.

Copyright by
ALYSSA JILL FRIEDE
2008

TABLE OF CONTENTS

LIST OF TABLES ........................................................ . ..................................................... vii
LIST OF FIGURES ............................................................................................................. x
INTRODUCTION ............................................................................................................... 1
Does Anticipated Work- Family Conﬂict Matter? ....................................... 2

A Work- Family Perspective... .. 3

A Career Decision- Making Perspective .......................................... 4

What IS Anticipated Work- Family Conﬂict? ........................................................... 6

Relationship between AWFC and WFC12
A Conceptual Model ofKey Study Variables.............................................14

Demographics ..................................................................... I 6
Personality .......................................................................... 21
Work and Family Role Importance23
Role-Planning Knowledge ....................................................... 25
Work-Family Decision-Making Self-Efﬁcacy (WFDMSE) ................. 25
Intervention Inﬂuencing WFDMSE ..................................................... 27
Outcomes of AWF C and WFDMSE ..................................................... 33
Certainty Regarding Medical Specialty Plans and
Family Plans ....................................................................... 34
Information Gathering.................................. .......................... 39
Mediational Hypotheses ................................................................... 44
METHOD ............................................................................................... 50
Participants .................................................................................. 50
Design and Procedure ..................................................................... 52
Experimental Conditions .................................................................. 54
Data Cleaning ............................................................................... 56
Measures ..................................................................................... 60
Anticipated Work-Family Conﬂict ............................................. 60
Certainty Regarding Medical Specialty Plans ................................. 62
Certainty Regarding Family Plans ............................................... 64
Career Decision Self-Efﬁcacy Scale ............................................ 65
Work-Family Decision-Making Self-Efﬁcacy ................................. 67
Role—Planning Knowledge ....................................................... 68
Gender .............................................................................. 69
Age .................................................................................. 69
Marital Status ...................................................................... 70
Additional Personal Characteristics ............................................. 70
Core Self-Evaluations ............................................................ 70
Role Importance ................................................................... 71

iv

Work-Family Information Gathering ........................................... 72

Differentiation among Self-Efﬁcacy Scales ................................... 74
Manipulation Check ............................................................... 75
RESULTS ............................................................................................... 76
Predictors of Anticipated Work-Family Conﬂict ....................................... 81
Demographics ...................................................................... 8 1
Core Self-Evaluations ............................................................. 81
Role Importance .................................................................... 82
Role-Planning Knowledge ........................................................ 82
Work-Family Decision-Making Self-Efﬁcacy ................................. 86
Effects of the Intervention on Career and Work-Family
Decision-Making Self-Efﬁcacy ........................................................... 90
Effects of the Intervention on Anticipated Work-Family Conﬂict ................... 94
Factors Inﬂuencing Degree of Change in WFDMSE ................................ 101
Outcomes of AWFC and WFDMSE .................................................... 104
Certainty of Medical Specialty Plans .......................................... 104
Certainty Regarding Family Plans ............................................. 107
Information-Gathering during Medical School .............................. 110
Requests for Electronic Newsletter ............................................ 110
Viewing the Website ............................................................. 112
Information-Gathering during Two-Week Lag ............................... 113
Mediation of the Relationship between WFDMSE and Outcomes ................. 115
DISCUSSION ....................................................................................... 1 15
Predictors of Anticipated Work-Family Conﬂict ..................................... 115
Effects of Intervention .................................................................... 120
Factors Inﬂuencing the Degree of Change in WFDMSE ............................ 124
Outcomes of Anticipated Work-Family Conﬂict ..................................... 127
Potential Methodological Limitations .................................................. 129
CONCLUSIONS ................................................................................... 133
APPENDICES ...................................................................................... 134
Appendix A: Email Invitation from Dean to Participate in Study .................. 135
Appendix B: Consent Form Time 1 ..................................................... 136
Appendix C: Consent Form Time 3 ..................................................... 138
Appendix D: Email Invitation to Participate in Time 3 .............................. 139
Appendix E: Debrieﬁng Form ........................................................... 140
Appendix F: Electronic Newsletter Email ............................................. 141
Appendix G: PowerPoint Slides for Career Condition Including Tasks ............ 142
Appendix H: PowerPoint slides for work-family condition
Including Tasks ........................................................... 150
Appendix I: Anticipated Work-family Conﬂict Items ............................... 158

Appendix J: Anticipated Work-Family Conﬂict Item Intercorrelations. . . . . ......160
Appendix K: Anticipated Work-family Conﬂict CFAs at

Each Administration ....................................................

166

Appendix L: Certainty Regarding Medical Specialty Plans Items ................ 167
Appendix M: Certainty Regarding Medical Specialty Plans Item

Intercorrelations ......................................................... 1 69
Appendix N: Certainty Regarding Family Plans Items ............................. 170
Appendix 0: Certainty Regarding Family Plans Item Intercorrelations ......... 172
Appendix P: Career Decision Self-efficacy Items ................................... 173
Appendix Q: Career Decision Self-efﬁcacy Item Intercorrelations ............... 174
Appendix R: Results of CFA for Career Decision Self-efﬁcacy Items ........... 178
Appendix S: Work-family Decision-making Self-efﬁcacy Items. . . . . . . . . . ........179
Appendix T: Work-family Decision-making Self-efﬁcacy Item

Intercorrelations ......................................................... 1 80
Appendix U: Results of CFA for Work-family Decision-making

Self-efﬁcacy Items ...................................................... 185
Appendix V: Role-planning Knowledge Items ....................................... 186
Appendix W: Role-planning Knowledge Item Intercorrelations .................. 187
Appendix X: Factor Loadings for Role-Planning Knowledge Items... .... ........190
Appendix Y: Demographics items ...................................................... 191
Appendix Z: Core Self-evaluations Items ............................................. 193
Appendix AA: Core Self-evaluations Item Intercorrelations ....................... 194
Appendix BB: CF A for Core Self-evaluation Items ................................. 195
Appendix CC: Role-importance Items ................................................ 196
Appendix DD: Role-importance Item Intercorrelations ............................ 197
Appendix EE: Factor Loadings for Role-importance Items ........................ 198
Appendix F F: Work-family Information Gathering Time 2 Items ................ 199
Appendix GG: Work-family Information Gathering Time 2 Item

Intercorrelations ....................................................... 200

Appendix HH: Work-family Information Gathering Time 3 Items ................ 201
Appendix II: Work-family Information Gathering Time 3 Item

Intercorrelations ......................................................... 203
Appendix J]: Factor Loadings for Work-family Information Gathering

Time 3 ..................................................................... 204
Appendix KK: Discrimination Among Self-efﬁcacy-related Scales .............. 205
Appendix LL: Manipulation Check Items ............................................ 209

REFERENCES .................................................................................... 211

vi

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Table 10:

Table 1 I:

Table 12:

Table 13:

Table 14:

Table 15:

Table 16:

LIST OF TABLES

Expected Differences across Conditions in Key Study Constructs

at Three Assessment Points .............................................................. 30
Demographic Information by Medical School ....................................... 60
Reliabilities for Anticipated Work-family Conﬂict Scales ......................... 62
Descriptive Statistics for Medical Specialty Plans ................................... 63
Descriptive Statistics for Family Plans ............................................... 64
Results of Manipulation Check ........................................................... 76
Correlations among Study Variables ................................................... 77
Hierarchical Regressions for H14: Role-planning Knowledge

Predicting Anticipated Work-Family Conﬂict (N = 108) ........................... 83

Path Analysis for H14: Auto-correlated Cross-lagged Relationships
Between Anticipated Work-family Conﬂict and Work-family
Role-planning Knowledge over Time .................................................. 85

Hierarchical Regressions for H15: Work-family Decision-making
Self-efﬁcacy Predicting Anticipated Work-Family Conﬂict
(N = 108) ................................................................................. 87

Path Analysis for H15: Auto-correlated Cross-lagged Relationships

between Anticipated Work-family Conﬂict and Work-family

Decision—making Self-efﬁcacy over Time .......................................... 89
Descriptive Statistics for H15: WFDMSE by Condition ......................... 91
Descriptive Statistics for H17: CDSE by Condition ................................. 93
Descriptive Statistics and One-way ANOVAs for H18:

AWF C across Conditions .............................................................. 97
Paired-samples T-tests for Differences in Anticipated Work-family

Conﬂict over Time by Cond1t10n97
Moderated Regressions: H19 — H24 ................................................ 103

vii

Table 17:

Table 18:

Table 19:

Table 20:

Table 21:

Table 22:

Table 23:

Table 24:

Table 25:

Table 26:

Table .11:

Table J2:

Table J3:

Path Analyses for H25-26: Relationship between Anticipated
Work-family Conﬂict and Certainty Regarding Medical

Specialty over Time .............................................................

Path Analyses for H27-28: Relationship between Work-family
Decision-making Self-efﬁcacy and Certainty Regarding Medical

Specialty over Time ...............................................................

Path Analyses for H29- 30: Relationship between Anticipated
Work- family Conﬂict and Certainty Regarding Family Plans
over Time... ...

Path Analyses for H31-32: Relationship between Work-family
Decision-making Self-efﬁcacy and Certainty Regarding Family

Plans over Time ...................................................................

Path Analyses for H35-36: Relationship between Anticipated
Work-family Conﬂict and Requests for the Electronic Newsletter

over Time ..........................................................................

Path Analyses for H37—38: Relationship between WFDMSE and

Requests for the Electronic Newsletter over Time ...........................

Path Analyses for H39-40: Relationship between Anticipated

Work-family Conﬂict and Viewing the Website over Time .................

Path Analyses for H41-42: Relationship between Work-family

Decision-making Self-efﬁcacy and Viewing the Website over Time. .

Path Analyses for H43-44: Relationship between Anticipated
Work-family Conﬂict and Gathering Information

(Over Two Week Time Lag) over Time ........................................

Path Analyses for H41: Relationship between Work-family
Decision-making Self-efﬁcacy and Gathering Information

(Over the Two Week Time Lag) over Time ..................................
Anticipated Work-Family Conﬂict Item Intercorrelations for Time 1....
Anticipated Work-Family Conﬂict Item Intercorrelations for Time 2.

Anticipated Work-Family Conﬂict Item Intercorrelations for Time 3.........

Table K1: Estimates of Fit Indices for Anticipated Work-Family Conﬂict

at Time 1 ..........................................................................

viii

..... 106

....107

..109

110

111

112

112

113

114

114

...160

162

164

166

Table K2: Estimates of Fit Indices for Anticipated Work-Family Conﬂict

at Time 2 .............................................................................. 166
Table K3: Estimates of Fit Indices for Anticipated Work-Family Conﬂict

at Time 3 .............................................................................. 166
Table Q1: Career Decision Self-Efﬁcacy Item Intercorrelations for Time 1 ............. 174
Table Q2: Career Decision Self-Efficacy Item Intercorrelations for Time 2... . . . . .......175
Table Q3: Career Decision Self-Efﬁcacy Item Intercorrelations for Time 3 ............. 176
Table T1: Work-family Decision-making Self-efﬁcacy Item Intercorrelations

for Time 1 ............................................................................. 180
Table T2: Work-family Decision-making Self-efﬁcacy Item Intercorrelations

for Time 2 ............................................................................. 182
Table T3: Work-family Decision-making Self—efficacy Item Intercorrelations

for Time 3 ............................................................................. 183
Table W1: Role-planning Knowledge Item Intercorrelations for Time 1 ................. 187
Table W2: Role-planning Knowledge Item Intercorrelations for Time 2 ................. 188
Table W3: Role-planning Knowledge Item Intercorrelations for Time 3 ................. 189
Table X1: Factor Loadings for Time 1 Role-Planning Knowledge Items ................. 190
Table X2: Factor Loadings for Time 2 Role-Planning Knowledge Items ................. 190
Table EEl: Factor Loadings for Role-importance Items .................................... 198
Table EE2: Factor Loadings for Role-importance Items Forcing Two Factors .......... 198
Table KKl: Factor loadings of WFRPK, CDSE, and WFDMSE Items ................... 205
Table KK2: Factor loadings of CSE, CDSE, and WFDMSE Items ........................ 207

ix

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 6:

Figure 7:

Figure 8:

Figure 9:

LIST OF FIGURES

Conceptual Model of Key Constructs ................................................ 15

Relationship between Anticipated Work-family Conﬂict and Certainty
Regarding Medical Specialty Plans ................................................... 37

Relationship between Anticipated Work-family Conﬂict and Certainty
Regarding Family Plans ............................................................... 39

Relationship between Anticipated Work-family Conﬂict and Requests
for the Electronic Newsletter ........................................................... 41

Relationship between Anticipated Work-family Conﬂict and Viewing
the website .............................................................................. 43

Mediational Relationship between Work-family Decision-making
Self-efﬁcacy, Anticipated Work-family Conﬂict, and Certainty
Regarding Plans ........................................................................ 45

Mediational Relationship between Work-family Decision-making
Self-efficacy, Anticipated Work-family Conﬂict, and Gathering
Information during Medical School ................................................... 47

Mediational Relationship between Work-family Decision-making
Self-efﬁcacy, Anticipated Work-family Conﬂict, and Request for
Electronic Newsletter .................................................................. 48

Mediational Relationship between Work-family Decision-making
Self-efﬁcacy, Anticipated Work-family Conﬂict, and Viewing
the Website .............................................................................. 49

Figure 10: Mediational Relationship between Work-family Decision-making

Self-efﬁcacy, Anticipated Work-family Conﬂict, and Gathering
Information during Two Week Time Lag .......................................... 50

Figure 11: Path Analysis: Relationships between Anticipated Work-family

Conﬂict and Work-family Role-planning Knowledge Over Time .............. 84

Figure 12: Path Analysis: Relationships Between Anticipated Work—family

Conﬂict and Work-family Decision-making Self-efficacy
Over Time ............................................................................... 88

Figure 13: Work-family Decision-making Self-efﬁcacy Over Time ........................ 92

Figure 14:
Figure 15:
Figure 16:
Figure 17:
Figure 18:
Figure 19:
Figure 20:

Figure 21:

Figure 22:

Career Decision Self-efﬁcacy Over Time ......................................... 94

Time-based Work-interference-with-family Over Time ........................ 98

Time-based Family-interference-with-work over Time ......................... 98

Strain-based Work-interference-with-family Over Time ........................ 99
Strain-based Fami1y-interference-with-work Over Time ........................ 99
Behavior-based Work-interference-with-family Over Time ................... 100
Behavior-based Family-interference-with-work Over Time ................... 100
Interaction between Work-family Decision-making Self-efficacy

and Age ............................................................................... 103
Interaction between Work-family Decision-making Self-efﬁcacy

and Marital Status ................................................................... 104

xi

INTRODUCTION

In the past 25 years, the number of dual-earner couples has markedly increased as
has the total number of hours that couples are working (Bond, Thompson, Galinsky, &
Prottas, 2002). Women are working more than ever before, while men are taking on
increased responsibility for domestic responsibilities and childcare (Bond et al., 2002).
These facts highlight the importance of research on work-family conﬂict, deﬁned by
Greenhaus and Beutell (1985) as a type of inter-role conﬂict in which the competing
demands of work and family roles are incompatible and participation in one role makes
participation in the other role more difﬁcult. The effects of work-family conﬂict have
been shown to be severe, including higher depression, increased alcohol use, increased
psychological burnout, greater reporting of psychosomatic symptoms, decreased job
satisfaction, and increased intention to turnover at work. (cf., Burke, 1988; Hammer,
Brockwood, & Neal, 2001; Frone, Russell, & Cooper, 1993; see also Eby, Casper,
Lockwood, Bordeaux, & Brinley, 2005).

While considerable research attention has been paid to the construct of work-
family conﬂict, surprisingly little attention has been paid to the concept of anticipated
work-family conﬂict. Adopting the Greenhaus and Beutell (1985) deﬁnition leads to the
following deﬁnition of anticipated work-family conﬂict: the belief that participation in
one ’5 future work-role will interfere with participation in one '3 future family—role (and
vice versa). This anticipated work-family conﬂict could be experienced, for example, as
individuals prepare to enter the workforce after college, when a new child is expected,
when they plan career changes, when a spouse changes jobs, or when considering re-

entering the workforce. In any situation where individuals expect their work or family

roles to change, they can anticipate work-family conﬂict between these two roles in the
future. To be clear, anticipated work-family conﬂict can exist when only one role (work
or family) changes and the other stays the same or when both are changing
simultaneously. I

In the current study, I build upon research conducted by Friede (2005) which
examined the construct of anticipated work-family conﬂict (AWF C). Friede (2005)
examined the nature of the construct, its antecedents, and its consequences among a
sample of medical students. Building upon that research, in this study, I will conduct an
intervention aimed at reducing AWFC among medical students. To do this, I focus on a
the construct of work-family decision-making self-efﬁcacy (WFDMSE).

I begin by clarifying the concept of AWFC and discussing its relevance to the
work—family ﬁeld. I then highlight critical variables in the nomological net of this
construct. I then describe the experimental intervention that will be the focus of this
study and discuss the proposed effects of this intervention through the presentation of
study hypotheses.

Does Anticipated Work-Family Conﬂict Matter?

While the importance of work-family conﬂict as an individual and societal
problem has become clear, one must consider whether the concept of anticipated work-
family conﬂict is important in its own right. This concept stands at the conﬂuence of two
streams of research: work-family conﬂict and career decision-making. From each
perspective, it is clear that there are several reasons why this topic merits attention from

researchers and practitioners.

A work-family perspective. From the work-family perspective, there is
considerable evidence that work-family conﬂict has detrimental outcomes for the
individual and the organization (as mentioned above; Eby et al., 2005). Yet, most
attempts to alleviate employee work-family conﬂict have been generated at the level of
organizational policies and practices (e.g., on-site childcare, maternity leave, ﬂextime;
Kossek & Friede, 2006). However, recent arguments have been made that individuals
themselves may have the opportunity to exert control over their lives in a way that
enhances their ability to manage work and family demands. The creation of
developmental opportunities to help individuals learn strategies for managing work and
family demands may complement organizational-level interventions (Behson, 2002;
Friedman, Christensen, & Degroot, I998; Friedman & Friede, 2006).

However, once employees enter an organization and a particular position, they
may have somewhat limited latitude regarding the strategies that they can use to balance
work and family roles (e.g., one cannot use ﬂextime if the organization does not allow it
or the position is not amenable to it). On the other hand, as individuals enter new careers
(e.g., after graduating college, switching organizations), they may have considerably
more latitude in selecting careers, specialties, and jobs that will allow them to effectively
manage work and family demands. Similarly, once individuals enter a marriage or have
children, they may have limited latitude regarding the strategies that they can use to
balance work and family. However, prior to entering relationship and family roles,
individuals may be able to make relationship and family decisions (e. g., at what age to
marry, how many children to have, selecting a partner who is willing to participate in

childcare) that will allow them to better manage work and family demands. Thus,

understanding how individuals think about the conﬂict that they will experience between
multiple life roles before entering them may facilitate the prevention or avoidance of
actual work-family conﬂict (as opposed to waiting for it to occur and then seeking
palliative strategies).

Additionally, individuals who enter their careers and jobs with realistic
expectations about the work and family demands that they will face may be less affected
by work-family conﬂict than those who were unprepared for the challenges that they
face. Research in the work role entry literature has suggested that individuals who have
more realistic information about the jobs that they enter (i.e., some negative information
in addition to positive information) have better outcomes on-the-job, including lower
attrition from the recruitment process, higher performance, and decreased voluntary
turnover, (Phillips, 1998). However, this research has not focused on providing realistic
information about the speciﬁc challenges that individuals will face balancing work and
family in a new work role. To the extent that individuals are prepared with accurate
information about the work-family demands that they will face on-the-job (i.e., the
demands and challenges of balancing work and family for a particular job), these
individuals may experience less work—family conﬂict and other negative outcomes once
in their new roles. In other words, the anticipation of work-family conﬂict may be
related to on-the—job experiences because the accuracy of expectations of conﬂict may be
associated with the extent to which job expectations are met or unmet.

A career decision-making perspective. From a career decision-making
perspective, considerable attention has been paid to how individuals make choices about

what careers, specialties, and jobs to enter (Phillips & lmhoff, 1997). However, much of

this research has ignored the extent to which work-family concerns play a role in such
decision-making (Barnett, Gareis, James, & Steele, 2003). A process model of individual
career decision-making suggests that individuals who identify the need to ﬁnd a new
career engage in a number of decision-making steps. First, upon recognizing a problem
(in some cases dissatisfaction with the current position, in other cases, an external need to
ﬁnd a new job/career — e. g., graduating from college, layoffs), individuals formulate a
role-change strategy (e. g., identifying ways to change the situation, resources available),
engage in a search for alternate roles (e.g., using internal and external resources),
evaluate the possibilities and, in some cases, transition into a new position (Mihal, Sorce,
& Comte, 1984).

Expectations and beliefs about the intersection of work and family roles may
come into play at a number of points in this career decision-making process. For
example, to the extent that one’s current career interferes with effective functioning in the
family role (i.e., work-interference-with-family, Greenhaus & Beutell, 1985), a
discrepancy between one’s current position and one’s ideal position may exceed a level
that is considered tolerable. For individuals entering the workforce for the ﬁrst time,
expectations regarding the work-family demands of a future role may also result in an
expected disparity between ideal conditions for work-family balance and the
opportunities provided by the position. Additionally, when individuals engage in a
search for altemative roles, they may seek out information regarding the extent to which
potential roles will enhance or reduce conﬂict between work and family roles and they
may weigh that information in their decision—making. According to Gianakos (1999) “A

quality decisional outcome is more likely when individuals rationally and systematically

evaluate career-related information...” (p.245). Therefore, because work-family issues
are associated with important work-related outcomes (Eby et al.), a rational and
systematic evaluation of work-family career-related information prior to role-entry is
likely to be important.

However, little empirical research has examined the extent to which work-family
issues are considered in career decision-making. Phillips and Imhoff (1997) note that
women face increased challenges with career planning in recent decades due to the
growing importance that they place on work and the continued importance placed on
family. Russell and Rush (1987) found that women were aware of the need to address
family considerations in career decision-making. Additionally, research on future life
roles has tended to examine career and non-work choices separately. Barnett et al. (2003)
note that most of the literature on college students’ plans for the future focuses on choice
of career and or course of study, with little attention paid to the interaction between non-
work and work decision-making.

Overall, the issue of anticipated work-family conﬂict is also important from a
career decision-making perspective. Career counselors attempting to help individuals
identify and enter careers that will be satisfying may beneﬁt from a greater understanding
of how anticipated work-family conﬂict inﬂuences career decision-making.
Additionally, an understanding of the extent to which a consideration of multiple life
roles in career counseling beneﬁts individuals and their employing organizations may
also be helpful for career counselors.

What is A nticipated Work-F ami/ y Conflict?

Because research on anticipated work-family conﬂict is in its infancy, it is vital
that a fuller understanding of what this construct really means is reached. The few
published articles that have speciﬁcally discussed anticipated work-family conﬂict as a
construct have offered only a basic deﬁnition of anticipated inter-role conﬂict. For
example, in one anticipated work-family conﬂict article, the author states that “In contrast
to previous studies on current perceptions of work-family conﬂict, this study examined
future expectations for work—family conﬂict” (Burley, 1994, p.116). This ﬁrst study on
anticipated work-family conﬂict does not mention how the author adapted the work-
family conﬂict scale that was used to assess ﬁxture work-family conﬂict (Burley, 1994).
A 1996 follow-up article on the same topic also does not provide much detail regarding
the meaning of anticipated work-family conﬂict in a similar fashion to the 1994 paper
(Livingston, Burley, & Springer, 1996). In this article, however, the authors elaborate on
the measurement of anticipated work-family conﬂict in which they take a traditional
work-family inter—role conﬂict scale (Kopelman, Greenhaus, & Connolly, 1984) and
translate the sentences into a future tense. That is, a sample survey item reads “My future
work will take up time that I would like to spend with my family and friends” (Livingston
et al., 1996, p. 184, italics added).

Weer, Greenhaus, Colakoglu, and Foley (in press) examined the construct of
expected work-family conﬂict as a measure of “students’ perceptions of the potential for
conﬂict or interference between their work and family roles after they embark on their
careers” (p.14). Hallett and Gilbert (1997) also offer a future difficulties scale which was
developed by Gilbert, Dancer, Rossman and Thom (1991). It assesses the perceptions

that young adults hold regarding the challenges and realities associated with having a

dual-career maniage. This future difﬁculties scale is similar to the construct of
anticipated work-family conﬂict in that it assesses future challenges, yet distinct in that it
does not directly address inter-role conﬂict. That is, it focuses on challenges within a
domain (work or family) but does not reﬂect challenges that arise from conﬂict between
roles (i.e., the interference of family life with work or vice versa). Friede (2005)
examined the construct of “anticipated work-family conﬂict”, and the ﬁndings from this
research (which are the precursor to the current study) are discussed in greater detail
below.

Researchers have also addressed related, yet distinct constructs. For example,
career-marriage conﬂict focuses on the expected conﬂict between one’s career and the
career of one’s spouse (Barnett et al., 2003). Work-family balance self-efﬁcacy (Killian,
Casper, Sitzman, Ellsbury, Burtrum, & Stephens, 2003) focuses on the beliefs that
individuals hold about their ability to avoid work-family conﬂict and deal with work-
family conﬂict when it occurs.

A number of researchers have addressed the attitudes that young adults hold
towards their future roles. One such construct is called attitudes towards multiple role
planning (ATMRP; Weitzman, 1994). Multiple role planning reﬂects intentions to
balance work and family in the future. ATMRP is considered the general orientation that
an individual holds regarding planning for the combination of career and family roles,
such as conﬁdence in the ability to integrate work and family roles, commitment towards
having multiple roles, and feelings of autonomy regarding the ability to make choices

about family and work roles for oneself. Another related issue concerns the attitudes that

 

individuals hold towards future role-sharing (e.g., sharing responsibilities for childcare
equally with a spouse versus being the primary care-taker).

Beyond the limited research on anticipated work-family conﬂict, it is worthwhile
to examine research on actual work-family conﬂict to discover how it has been deﬁned
and measured and how this may inﬂuence our understanding of anticipated work-family
conﬂict. A 1985 article by Greenhaus and Beutell has largely been responsible for
deﬁning the work-family conﬂict construct. As mentioned earlier, they deﬁne work-
family conﬂict as a form of inter-role conﬂict in which there are competing demands
arising from an individual’s participation in different roles (in this case, the work and
family roles). Greenhaus and Beutell (1985) discuss three major forms of work-family
conﬂict. The ﬁrst is time-based conﬂict. This reﬂects the fact that individuals have a
ﬁnite amount of time and that time spent at work cannot be spent with the family and
vice versa. This form of inter-role conﬂict exists because time dedicated to one role
makes it challenging to ﬁll the requirements of a different role. Another type of work-
family conﬂict that Greenhaus and Beutell (1985) discuss is strain-based conﬂict. This
exists when strain in one role (e.g., tension, anxiety, fatigue, or irritability) makes
participating in the other role more difﬁcult. For example, stress from work may
spillover into the family role and affect one’s ability to have a healthy and satisfying
family life. Finally, Greenhaus and Beutell (1985) describe behavior-based conﬂict.
This is when certain patterns of behavior within one role are incompatible with behaviors
necessary for effective functioning in another role. For example, the authors discuss how
managers may be expected to display aggressiveness and objectivity in the workplace and

that they may be expected to be warm and nurturing in their home environments. The

incompatibility of these two types of behaviors may cause behavior-based work-family
conﬂict. Research has supported the existence of time- and strain-based conﬂict but
behavior-based conﬂict has been difﬁcult to operationalize and is less supported by
research (Kelloway, Gottlieb, & Barham, 1999).

Furthermore, Greenhaus and Beutell (1985) discuss another key issue in the
conceptualization of work-family conﬂict, the directionality of the conﬂict. The authors
distinguish between the perception that work is interfering with family and the perception
that family is interfering with work. They propose that an individual’s response to a
situation with competing role demands will determine whether it is perceived as family-
interference-with-work or work-inteifei'ence-with—family. For example, if a person
chooses to attend a meeting at work rather than his child’s school performance, this
situation will be perceived as work-interfering-with-family. However, if the opposite
choice was made, it would be perceived as family-interfering-with-work. To this end, the
authors argue that it is important to develop work-family conﬂict'scales that contain
items that reﬂect both directions of role interference.

Research in the work-family conﬂict area has attempted to predict the different
types of conﬂict and the directionality of the conﬂict from different antecedents and
examined the different outcomes expected from these types of conﬂict. For example, in a
longitudinal study of work-family conﬂict conducted by Kelloway et al. (1999), results
showed that work-interference-with-family could be distinguished from family-
interference-with-work and that strain-based conﬂict could be distinguished from time-
based conﬂict. The authors found that only strain-based family-interference-with-work at

Time I predicted stress and intent to tumover in a job at Time 2, while time-based work-

10

interference-with-family at Time 2 was predicted by stress at Time 1. Similarly, Hammer
et al. (2001) found that family-interference-with-work predicted depression while work-
interference-with-family did not. F rone et al. (1992) found that work-interference-with-
family was predicted by job stressors while family-interference-with-work was predicted
by family stressors and family involvement.

Thus, ﬁndings from the literature on work-family conﬂict suggest that these six
dimensions are relevant to the nature of that construct. For this reason, they may also be
relevant to the related construct of anticipated work-family conﬂict. Because anticipated
work-family conﬂict is being deﬁned as the belief that participation in one’s future work-
role will interfere with participation in one’s future family role (and vice versa), it is
important to recognize that individuals may anticipate time-, behavior-, and/or strain-
based conﬂict in either direction (work-interference-with-family and/or family-
interference-with-work) and to measure the construct accordingly. A measure of
anticipated work-family conﬂict was developed by F riede (2005).. An 18-item work-
family conﬂict scale created by Carlson, Kacmar, and Williams (2000) was altered to
reﬂect the measurement of “anticipated” conﬂict. Therefore, all items were rewritten in
the future tense. Six types of anticipated work-family conﬂict were measured including
anticipated behavior-based work-interference-with-family (behavior-based WIF'),
anticipated behavior-based family-interference-with-work (behavior-based F] W),
anticipated strain-based work-interference-with—family (strain-based WIF), anticipated
strain-based family-interference-with-work (strain-based F l W), anticipated time-based
work-interference-with-family (time-based WIF) and anticipated time-based family-

interference-with-work (tin-ze—based F I W).

Carlson et al. (2000) found that a structural equation model with six correlated
latent factors ﬁt their data best. Friede (2005) also conducted a CFA to examine whether
the same six factors emerged in the measure of AWF C. A six factor model was
compared to a three factor model (which reﬂected time-, strain-, and behavior-based
conﬂict — ignoring direction of conﬂict), a two factor model (work-interference with
family and family-interference-with-work — ignoring type of conﬂict), and a
unidimensional model (all anticipated work-family conﬂict items). Factors were allowed
to correlate. Results indicate that the six dimensional model of anticipated work-family
conﬂict ﬁt the data signiﬁcantly better than the alternative models (F riede, 2005). This
CFA uses the same methods as Carlson et al. (2000) and replicates their ﬁndings for the
anticipated work-family conﬂict scale. Based on the ﬁndings from the conﬁrmatory
factor analysis, scale scores for each of the six dimensions were created. The reliability
of the six scales ranged from a = .75 (time-based F 1W) to a = .88 (strain-based FIW).
Thus, there is sufﬁcient evidence for the existence of six forms of AWFC, similar to the
six forms of work-family conﬂict described by Greenhaus and Beutell (1985).

Relationship between A WF C and WF C. To this point, I have argued the
importance of the construct of AWFC to the work-family and career planning ﬁelds and
have deﬁned the construct in detail. However, as noted earlier, the purpose of this study
is to develop an intervention that addresses issues of AWFC among medical students.
One key question becomes whether greater levels of AWFC are associated with more or
less experienced work-family conﬂict (once in work and family roles). Operating under

the assumption that it is desirable to reduce actual work-family conflict, interventions

may be aimed at decreasing or increasing AWF C, depending upon its relationship to
actual work-family conﬂict.

On the one hand, AWF C may serve as a warning sign to individuals that directs
them to make choices that effectively allow them to avoid or cope with actual work-
family conﬂict. For example, individuals who anticipate a great deal of work-family
conﬂict may make decisions that help them to avoid experiencing the levels of conﬂict
that they anticipate (e.g., delaying childbirth, selecting a less demanding career).
Additionally, high levels of AWFC may act as a “realistic preview” of what is to come,
and help individuals to develop accurate expectations that are then met on-the-job
(Phillips, 1998). If this is the case, interventions might be aimed at raising levels of
AWFC to a realistic level.

On the other hand, AWF C may serve as a relatively accurate marker of the work-
family conﬂict that individuals will actually experience. In other words, individuals who
expect to experience more conﬂict actually do experience more conﬂict. In this case, the
AWFC construct provides insight into how much conﬂict individuals will experience.
Thus, interventions that genuinely providing individuals with the
skills/resources/information that are helpful in managing multiple life roles will reduce
levels of AWFC and actual work-family conﬂict. The reduction in AWFC, in this case,
serves as a genuine indicator that individuals have gained preparedness to handle multiple
roles in a way that allows them to avoid or alleviate actual work-family conﬂict.

There has been limited research connecting AWFC to actual work-family conﬂict.
Friede (2006) collected follow-up data from the medical students in her 2005 study

approximately one year after the initial data collection. At the time ofthe follow-up,

participants were asked to report their actual work-family conﬂict in their medical
residency. Friede (2006) found a positive correlation between anticipated behavior-based
conﬂict and actual behavior-based conﬂict in both directions (WIF and FIW).
Anticipated strain-based WIF family was signiﬁcantly. and positively correlated with
actual strain-based WIF, but the same was not true for strain-based F IW (i.e., the
correlation was not signiﬁcant). Anticipated time-based WIF and F IW were not
signiﬁcantly correlated with the actual experiences of these forms of work-family
conﬂict, respectively. Overall, Friede (in progress) found only positive or non—signiﬁcant
correlations between anticipated and actual work-family conﬂict (no negative
correlations), suggesting that individuals who expect more work-family conﬂict actually
do experience more work-family conﬂict, overall.

This supports the notion of AWF C as a “marker variable” that points to future
levels of actual work-family conﬂict. The current study therefore builds upon these
ﬁndings and attempts to develop an intervention that reduces levels of AWF C, with the
intention that this will also reduce levels of actual work-family conﬂict. The AWFC
variable will hopefully serve as an indicator that the intervention has taught students to
make decisions that will help them prepare for their multiple roles. Note that while it is
beyond the scope of the current study to examine the relationship between AWF C and
actual work-family conﬂict (because of the time-lag involved), longitudinal data
collection regarding the effectiveness of the intervention in alleviating actual work-family
conﬂict is planned.

A Conceptual Model ofKey Studv Variables

Figure 1 provides a basic conceptual model of the key constructs that will be
examined in this study (note that moderator variables will be addressed later in the
paper). Below, I describe the major components of this model and their relationship to
AWFC and highlight previous research supporting the relationships between variables

included in the model.

 

 

 

 

 

 

 

 

PREDICTORS Anticipated Work- OUTCOMES

. Demographics 7 Family Conﬂict o Certainty re.

0 Core self—evaluations (6 types) medical specialty

0 Role importance plans. .

o Role-planning 0 Certainty re. fam1ly
knowledge plans .

o WFDMSE 0 Inforrnatlon

gathering behaviors

 

 

 

 

 

 

Figure 1. Conceptual Model of Key Constructs

F riede (2005) examined some important predictors of AWF C. In the current
study, I build upon that research by replicating some of the analyses conducted by F riede
(2005; ie., demographics, core self-evaluations, role importance, and role-planning
knowledge as predictors of AWFC). Additionally, I introduce a new construct of interest,
work-family decision-making self-efﬁcacy (WFDMSE) and conduct an in-depth
examination of the role that WFDMSE plays in inﬂuencing AWF C. Speciﬁcally, I
conduct an experimental intervention to manipulate WFDMSE and examine the effects
on AWFC. This enhances our understanding of this construct as a predictor of AWF C,
because it allows for stronger inferences of causality.

In the following sections, I ﬁrst describe the relationships between predictors and

AWF C that are being replicated from Friede (2005). These relationships are being

15

analyzed only at the initial data collection (prior to the intervention). Next, I introduce
the construct of WFDMSE and present hypotheses regarding its relationship with AWFC.

Demographics. Of the limited research on AWFC and related constructs,
demographics variables are not frequently examined as predictors of AWFC because they
are controlled for (either statistically or by limiting the sample) and thus their direct
effects on AWFC are unknown. For example, McCracken and Weitzman (1997)
controlled for education level in their study of attitudes towards multiple role-planning
because they found it to be correlated with the construct. Also, the McCracken and
Weitzman (1997) study only included women while Hallett and Gilbert’s (1997) study
included only female, never-married, upper—classmen thus controlling for year in school,
gender, and marital status through the selection of the sample. While Kerpelman and
Schvaneveldt (1999) included both men and women in their study, they only included
participants who were never married, never a parent, and were within the age range of 18
to 25. Many studies assess the race of the participants, controlling for its effects on the
outcomes of interest. (cf., Livingston et al., 1996).

Research conducted on actual work-family conﬂict also tends to consider
demographics, yet limited conclusions have been reached regarding their inﬂuence on
work-family conﬂict. Demographics and background characteristics have been shown to
comprise almost 25% of the predictors included in research on the work-family interface
(e.g., age, race, gender, and marital status). Lockwood, Casper, Eby and Bordeaux
(2002) note that questions about children were “almost non-existent in this research” (p.
16). As one example of an investigation of demographics as a predictor of work-family

conﬂict, Burke (1988) found that being married was associated with increased work-

family conﬂict but that age, gender, education level, and years on the job were not.
However, parental status was not assessed in this study. Other studies have also
examined the inﬂuence of demographic variables with mixed results (cf., Frone et al.,
1992). For example, different researchers have found‘that women experience greater
conﬂict (e.g., Behson, 2002), less conﬂict (e.g., Parasurman & Simmers) and no
differences in conﬂict (e.g., Duxbury & Higgins, 1991) as compared to men.

In the current study, I assess the relationship between a number of demographic
variables and AWF C. Friede (2005) also examined these relationships. In cases where
F riede (2005) did not ﬁnd signiﬁcant results, I nonetheless provide directional hypotheses
about the expected relationships because the non-signiﬁcant ﬁndings of Friede (2005)
may not replicate in the current sample.

This study will examine the relationship between age and AWF C. Older
individuals are expected to be considering both work and family decisions more
seriously, because they are closer in proximity. Therefore, concerns about balancing
work and family may be particularly salient. For the current sample of medical students,
both the demands of managing a medical career and the demands of managing a family
may become more salient with age as greater exposure to those roles is gained. Friede
(2005) found that age was only signiﬁcantly positively correlated with one of the six
types of AWFC (strain-based F 1W). However, all correlations between age and AWFC
were in the positive direction. Therefore, I propose the following hypothesis:

Hypothesis 1: Older participants will anticipate greater (a) time-based WIF, (b) time-
based F] W, (c) strain-based WIF, (d) strain-based F] W, (e) behavior-based WIF, and (ﬂ

behavior-based F I W .

l7

In a similar vein, students who are farther along in school are closer to actually
having a job than those who are earlier in their education. The reality of balancing their
work and non-work lives as professionals is looming closer in the future. While age and
year in school are likely to be correlated, there may be particular insights that are gained
through progressing through medical school that are unrelated to age. For medical
students, those who are further along in school will have had greater exposure to the
different career options that lie ahead of them. They will have learned more about the
different specialties formally in the classroom as well as had more exposure to informal
education from other students about the demands of different medical careers (e.g.,
through knowing other students who have graduated already and begun their residency
programs). Therefore, we also expect students who are farther along in medical school to
anticipate greater work-family conﬂict. However, F riede (2005) did not ﬁnd any
signiﬁcant correlations between year in school and any of the six types of AWFC.
Hypothesis 2: Participants who are farther along in medical school will anticipate
greater (a) time-based WIF, (b) time-based F] W, (c) strain-based WIF, (d) strain-based
F] W, (e) behavior-based WIF, and (ﬂ behavior-based F] W.

In today’s society, women are still taking on the primary role of providing
childcare (Bond et al., 2002). And yet, although still working fewer hours outside of the
home than men, women in the workforce are more educated than men and are more likely
to be managers and professionals than men (men are more likely to hold blue-collar jobs
than women; Bond et al., 2002). Also, mothers have been shown to have less time to
themselves than men, according to the 2002 Highlights of the National Study for the

Changing Workforce (Bond et al., 2002). These societal-level differences between the

lives of men and women may mean that women can expect to be bearing the greater
burden for family-care. With regard to medical students, in the 2001-2002 academic
year, 47.6% of new entrants to medical school were women (Women in US. Academic
Medicine, 2004, para. 3). The women in this study are entering into a challenging
profession and yet will likely assume the primary responsibility for childcare in their
households (see above). Interestingly, Friede (2005) found that the only signiﬁcant
differences between men and women emerged in the relationship between time-based and
behavior-based work-interference-with-family. Women actually anticipated less work-
interference-with-family in those two cases. A reasonable explanation for this is that
women were planning to ensure that their medical career did not intrude on their family
life. Given this logic, women may be more concerned than men about their family lives
interfering with work (although these correlations were not signiﬁcant in the Friede
(2005) study). Given the ﬁndings from Friede (2005), I propose:

Hypothesis 3: Female participants will anticipate greater (a) time-based F I W, (b) strain-
based WIF, and (c) behavior-based WIF.

Hypothesis 4 .' Female participants will anticipate less (a) time-based WIF, (1)) strain-
based WIF, and (c) behavior-based WIF.

Marital status may also affect the anticipation of work-family conﬂict. Similar to
the predicted effects of age, those individuals who are married or have long-term partners
may be more likely to have children in the near future (as compared to those who have
not yet found a spouse). Further, those individuals who are married or have long-term
partners may have had important discussions with their partners about how the couple

will manage work and family responsibilities, whereas single individuals may have spent

19

considerably less time thinking and talking about this issue. Again, this attention placed
on future balance may lead to an awareness of the challenges of balancing work and
family, and therefore an increased reporting of anticipation of work-family conﬂict. With
regard to medical students, the reasoning above is expected to apply. Further, because of
the speciﬁc constraints associated with the completion of medical school (e.g., paying off
large debt from schooling, choosing a residency program, etc.), these married/partnered
students may be especially likely to have concerns about balancing work and family.
Interestingly, Friede (2005) found that married individuals anticipated signiﬁcantly less
behavior-based FIW but that none of the other correlations were signiﬁcant. However,
the relationships for anticipated strain-based conﬂict and behavior-based conﬂict were all
in the negative direction while the relationships for anticipated time-based conﬂict were
in the positive direction. Perhaps married individuals are more aware of the difﬁculties
of negotiating the time-demands of marriage and a family (because they are currently
negotiating the time-demands of marriage and medical school) but they have already
learned that the stress and behaviors from different roles do not inhibit performance in the
other roles. Based on the ﬁndings from Friede (2005), I predict:
Hypothesis 5: Married participants will anticipate greater (a) time-based WIF and (b)
time-based F I W.
Hypothesis 6: Married participants will anticipate less (a) strain-based WIF, (b) strain-
based F] W, (c) behavior—based WIF, and (d) behavior-based F I W.

Parents (as compared to non-parents) may anticipate different amounts of work-
family conﬂict because they may be more aware of the excessive demands that children

place on their time. Because prior research on future roles has tended to exclude parents

20

from their samples, this relationship will be particularly interesting and informative. In
the case of medical students, medical school is an extremely challenging environment
that may equal (if not exceed) the rigors of actually practicing medicine. Because
medical student—parents are already managing the demands of their school and family
lives, they may be more aware of the challenges that lie ahead and anticipate more
conﬂict. However, F riede (2005) did not ﬁnd any signiﬁcant relationships between
parental status and any of the six forms of AWF C. I maintain the original hypothesis that
parents will anticipate greater levels of conﬂict.

Hypothesis 7: Participants who are parents will anticipate greater (a) time-based WIF,
(b) time-based F] W, (e) strain-based WIF, (d) strain-based F l W, (e) behavior-based
WIF, and (f) behavior-based F] W.

One criticism of Friede (2005) is that the effects of being in a committed,

romantic relationship were unstudied. Indeed, individuals with signiﬁcant others may
have spent more time thinking about and discussing the challenges of balancing work and
family in the future. Additionally, individuals with signiﬁcant others may be more likely
to enter marital and parental roles in the nearer future. Thus, this may make concerns
about managing work and family demands more salient. Therefore, I also propose the
following hypothesis:
Hypothesis 8: Participants in romantic relationships will anticipate greater (a) time-
based WIF, (b) time-based Fl W, (e) strain—based WIF. (d) strain-based F] W, (e)
behavior-based WIF, and (f) behavior-based F I W.

Personality. There has been very little research examining the relationship

between personality and beliefs about future roles. Research from the literature on actual

21

work-family conﬂict can be useful in understanding which personality variables may
predict anticipated work-family conﬂict. Recently, researchers have begun to consider
the role of personality in the experience of work-family conﬂict. For example, research
has addressed the role of negative affectivity (cf., Carlson, 1999; Stoeva, Chiu &
Greenhaus, 2002) and the “Big Five” (cf., Bruck & Allen, 2003) in the experience of
work-family conﬂict.

A book chapter by Friede and Ryan (2004) argues that core self-evaluations will
predict work-family conﬂict and that people with more positive core self-evaluations will
experience less work-family conﬂict. Core self-evaluations are considered the
fundamental premises individuals hold about themselves, or the extent to which
individuals possesses a positive self-concept (Judge, Erez, & Bono, 1998). The four traits
that comprise this higher level construct are self-esteem, generalized self-efﬁcacy, locus
of control and emotional stability. Core self-evaluations are a latent, multivariate
construct or a compound personality variable in that it is comprised of four more speciﬁc
traits (Judge et al., 1998).

The work of Friede and Ryan (2004) argues that individualslwho have more
positive core self-evaluations will experience less work-family conﬂict. They suggest
that individuals with positive core self—evaluations may select environments for
themselves that are more supportive of demands in other domains, such as selecting a job
that is ﬂexible and allows the individual to meet family demands (Diener, Larson, &
Emmons, 1984; Judge et al., 1998). Thus, they experience less work-family conﬂict than
individuals with more negative core self-evaluations. It is also possible, according to

Friede and Ryan (2004), that individuals simply perceive their environments differently

based on their core self-evaluations. That is, individuals with negative core self-
evaluations perceive the environments from their multiple life domains more negatively
and as less supportive of obligations in other domains (Fogarty et al., 1999; Larsen, 1992;
Moyle, 1995). Therefore, they feel as though they have more conﬂict in their lives.
Further, individuals with positive core self-evaluations may use more effective coping
styles to handle the demands of their work and non-work lives, and in doing so, decrease
the amount of conﬂict that they experience (Aryee, Luk, Leung, & Lo, 1999). Friede
(2005) examined the relationship between core self-evaluations and anticipated work-
family conﬂict and found that core self-evaluations were signiﬁcantly negatively
correlated with all six forms of anticipated work-family conﬂict (with the strongest
relationships existing, not surprisingly, between core self-evaluations and strain-based
conﬂict). Based on these ﬁndings, I propose the following:
Hypothesis 9: Participants with higher core self-evaluations will anticipate less (a) time-
»based WIF, (b) time-based F] W, (c) strain-based WIF, (d) strain—based F] W, (e)
behavior-based WIF, and (f) behavior-based F] W.

Work and family role importance. An additional factor that may inﬂuence how
much conﬂict individuals anticipate is the importance that they place on work and family .
roles as critical components of their lives. In the small literature on future roles, role-
importance has emerged as a popular attitudinal construct. For example, Kerpelman and
Schvaneveldt (1999) compared the importance that males and females placed on their
career, marital, and parental roles. Rajadhyaksha and Bhatnagar (2000) examined gender
and age differences in role-importance in married dual-career couples in India. Bu and

McKeen (2000) investigated differences between Canadian and Chinese business

23

students in the importance that they placed on the three life roles. Friede (2005)
examined role importance and found that greater importance placed on the work role was
associated with greater anticipated behavior-based WIF, greater anticipated strain-based
WIF, and less anticipated time—based F IW (when controlling for demographics,
personality, and other attitudinal constructs). More broadly, this implies that individuals
who place a high importance on their work role expect that work will interfere with their
family lives but that family life will not interfere with work. Therefore, the following
hypotheses are proposed:
Hypothesis 10: Participants who place a greater importance on the work role will
anticipate greater (a) time—based WIF, (b) strain-based WIF, and (c) behavior-based
WIF.
Hypothesis [1: Participants who place a greater importance on the work role will
anticipate less (a) time-based FIW, (b) strain-based F I W, and (c) behavior-based F I W.
Friede (2005) also examined the relationship between the importance that
individuals place on marital and parental roles and their AWFC. F riede (2005) found that
individuals who place a greater importance on marital and parental roles anticipated
greater time-based FIW. Those who place a greater importance on the parental role also
anticipated less strain-based WIF. These ﬁndings parallel the ﬁndings from work role
importance in that the people who place a high importance on their family life expect
family to interfere with work but not the reverse. For the current study, 1 merge the
parental and marital roles into a combined “family” role importance. Therefore, I propose

the following hypotheses for the current study:

24

Hypothesis 12: Participants who place a greater importance on the family role will
anticipate greater (a) time-based F] W, (b) strain-based F] W, and (c) behavior-based
F I W.
Hypothesis 13: Participants who place a greater importance on the family role will
anticipate less (a) time-based WIF, (b) strain-based WIF, and (c) behavior-based WIF.

Role—planning knowledge. Another predictor of AWFC that F riede (2005)
investigated was role-planning knowledge. Role-planning knowledge is deﬁned as the
belief that one knows how to plan for holding multiple roles and it is measured as a self-
assessment of knowledge (Weitzman & Fitzgerald, 1996). Friede (2005) examined the
relationship between role-planning knowledge and the six forms of anticipated work-
family conﬂict. When controlling for demographics and personality, role-planning
knowledge explained incremental variance in all six forms of anticipated work-family
conﬂict. In all cases, greater role-planning knowledge was associated with decreased
anticipation of work-family conﬂict. Although Friede (2005) incorporated other
attitudinal constructs in her study, role-planning knowledge was the strongest and most
consistent predictor of the six forms of AWFC. Therefore, in the current study I predict:
Hypothesis 1 4: Participants with greater role-planning knowledge will anticipate less (a)
time-based WIF, (b) time-based F] W, (c) strain-based WIF, (d) strain—based F] W, (e)
behavior-based WIF, and (f) behavior-based F] W.

Work-family decision-making self—efﬁcacy (WFDMSE). An additional construct
that was not included in Friede (2005) is relevant to the prediction of AWFC. Although
this construct, career decision self-efﬁcacy (CDSE), is focused on career choice (without

a speciﬁc consideration of non-work roles), it serves as the foundation for thinking about

how self-efﬁcacy may inﬂuence AWF C. CDSE is deﬁned as one’s conﬁdence in his/her
ability to successfully investigate and decide upon a career (Taylor & Betz, 1983). The
Taylor and Betz (1983) measure of CDSE has ﬁve dimensions: self-appraisal, gathering
occupational information, planning, problem-solving, and goal selection (based on Crites’
1961 ﬁve Career Choice Competencies). Betz and Hackett (1981) found that as
individuals engaged in behaviors ’such as goal setting, gathering occupational
information, and problem-solving, their CDSE increased. Gianakos (1999) noted that
individuals with greater CDSE have been shown to have greater self-esteem overall, to
consider a wider range of occupations and are more committed to their career decisions
(Church, Teresa, Rosebrook, & Szendre, I992; Lent & Hackett, 1987; 1987; Robbins,
1985)

I argue that a new, parallel construct, work-family decision-making self-efﬁcacy
(WFDMSE), is likely to be related to individuals’ experiences of AWFC (Note that the
original authors of the CDSE scale used the term “decision-making” but changed it due to
copyright issues). I use the term “decision-making” for the work-family variable to
highlight the importance of the decision-making process. WFDMSE can be deﬁned as
the belief that one is capable of investigating and making decisions regarding the
management of work and family roles. Individuals who have greater conﬁdence in their
ability to make decisions that will allow them to balance work and family are expected to
anticipate less work-family conﬂict. Similar to CDSE, WFDMSE can be thought of as
having ﬁve components: self-appraisal regarding goals for balancing work and family,
gathering information about balancing work and family in different careers and family

situations, selecting goals for how to balance work and family, planning for entering

26

work and family roles that meet one’s goals, and problem-solving when challenges arise
to pursuing one’s work and family goals. In the current study, I choose to focus on self-
appraisal, gathering information, and goal selection due to the time constraints associated
with measuring and manipulating self-efﬁcacy along all ﬁve dimensions.

Note that this construct is similar to role-planning knowledge (which emerged as
an important construct in earlier research). The key difference between role-planning
knowledge and WFDMSE is that WFDMSE directly assesses conﬁdence whereas role-
planning knowledge is an evaluation of perceived knowledge level. These constructs are
expected to be positively correlated. Both of these constructs are also distinct from
AWF C. While the AWFC construct is focused on how much conﬂict one expects to
experience, WFDMSE is focused on conﬁdence in one’s own ability to seek out
information and plan for making decisions regarding work and family roles, and role-
planning knowledge focuses on self-assessed knowledge. The reason that I choose to
focus on WFDMSE in the current study (in addition to role-planning knowledge) is the
emphasis that research on careers has placed on investigating CDSE and the existing
evidence that CDSE can be increased through interventions. Therefore, I propose the
following hypotheses:
Hypothesis 15: Participants with greater WFDMSE will anticipate less (a) time-based
WIF, (b) time-based Fl W, (e) strain-based WIF, (d) strain-based F1 W, (e) behavior-
based WIF, and (f) behavior—based F] W.
Intervention Inﬂuencing WFDMSE

A critical component of the current study is the fact that I conduct an intervention

intended to increase participants” WFDMSE to examine the causal effects on AWFC.

27

This is valuable for a number of reasons. One major criticism of the work-family ﬁeld is
the reliance on self-report cross-sectional data. Therefore, one important area for future
research is to manipulate key predictor variables to examine their relationship with
AWFC and other important outcome variables (e. g., certainty regarding relationship,
family, and career plans). Experiments have long been considered an important research
method in psychological research because they facilitate the inference of causality which
is critical in determining the nature of the relationship between constructs (Shadish,
Cook, & Campbell, 2001). Additionally, the use of experimental manipulations
overcomes many of the limitations of cross-sectional self-report studies (e. g., common
method bias; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).

Additionally, an important component of psychological research is the evaluation
of the effectiveness of experimental and real-world interventions. Within the American
Psychological Association, there is a division called the “Evaluation, Measurement, and
Statistics Division” (American Psychological Association, n.d.).j The American
Evaluation Association (n.d.) also emphasizes the evaluation of training programs. Such
evaluations are valuable for the improvement of the program or intervention and for
making decisions regarding future programs or interventions (Donaldson & Christie,
2006). There is a growing recognition of the importance of evaluation for the progress
and effectiveness of psychological research. For example, research focused on employee
training and development often considers the effectiveness of training programs.
Kirkpatrick (I994) highlighted the importance of evaluating a number of training
outcomes, including reactions, learning, transfer of training, and bottom-line results.

Therefore, as a supplement to correlational research on work—family topics, it is an

28

important step to evaluate the effectiveness of work—family interventions, whether they
are at the individual, group, or organizational level.

More speciﬁcally, in the current study, I randomly assigned medical students to
one of two intervention conditions: a CDSE intervention condition (CDSE; Career
condition), and a WFDMSE intervention condition (Work-family condition). The goal of
the Work-family condition is to raise participant levels of WFDMSE relative to the
Career condition.

I have chosen to focus on career decision self-efﬁcacy as a control condition
because it parallels the Work-family condition in important ways that control for
alternative explanations of the impact of the intervention. For example, the Work-family
condition might boost mood or inﬂuence participants overall levels of self-efﬁcacy (as
opposed to self-efficacy speciﬁcally for work-family decision-making) and this might be
related to AWFC. By focusing on self-efﬁcacy in the Career condition as well, this
problem is mitigated. In other words, it is critical to understand whether inﬂuencing the
more speciﬁc construct of self-efﬁcacy related to decision making about work andfamily
is important. Additionally, many professional students may already be getting support
related to decision-making about careers (that does not address work and family issues)
and a practical question is whether the additional focus on work and family roles is
important. It is important to note that I will assess key study variables prior to the
intervention (Time 1), immediately following the intervention (Time 2), and after a two
week time-lag (Time3). Prior research focused on self-efﬁcacy related to career
decision-making has utilized this same time-lag (Luzzo, Funk, & Strang, 1996).

Therefore, for the purposes of comparing this study to previous research and addressing

29

the potential criticism that the effects of the intervention are limited to the immediate
time and setting, I ﬁnd it valuable to also include a follow-up after two weeks.

The intervention conditions were designed utilizing three of Bandura and Adams’
(1977) mechanisms for enhancing self-efﬁcacy: performance accomplishments,
observational learning, and verbal persuasion (described in greater detail in the Method
section). The content of the interventions will focus on enhancing self-efﬁcacy for work-
family decision-making or career decision-making in the following areas: self-appraisal,
gathering information, and goal selection (based on Crites’ 1961 career competencies). I
will examine the effects of the intervention conditions on key study variables. Table 1
depicts the expected results of the intervention.

Table 1

 

Expected Differences Across Conditions in Key Study Constructs at Three Assessment
Points.

 

 

Variable Time 1 Time 2 Time 3

WFDMSE WF = Career WP > Career WF > Career
CDSE WF = Career WF < Career WF < Career
AWFC (6 types) WP = Career WF < Career WF < Career

 

Note. WFDMSE = Work-family Decision-making Self-efﬁcacy. CDSE = Career
Decision Self-efﬁcacy. AWFC = Anticipated Work-family Conﬂict. WF = Work-family
condition. Career = Career condition.

When examining the effects of the intervention, it is important to consider
whether the Work-family intervention does, in fact, increase WFDMSE more so than the
Career intervention. Conversely, the Career intervention should increase CDSE more so
than the Work—family intervention. Thus, the following hypotheses are proposed:
Hypothesis [6: Participants in the Work-family condition will show greater WFDMSE

than those in the Career condition after the intervention.

30

Hypothesis 1 7: Participants in the Career condition will show greater CDSE than those
in the Work-family condition after the intervention.

As noted earlier, I expect higher WFDMSE to be associated with lower AWF C.
Therefore, I anticipate that one effect of the Work-family condition will be to lower
AWFC.

Hypothesis 18: Participants in the W ark-family condition will anticipate less (a) time-
based WIF, (b) time-based F! W, (c) strain-based WIF, (d) strain—based F] W, (e)
behavior-based WIF, and (I) behavior-based F] W after the intervention than those in the
Career condition.

With regard to the impact of the intervention, it is important to consider the
possibility that not all individuals will react in the same way to the intervention. In fact,
there may be systematic differences between individuals that inﬂuence their reaction to
the intervention. A number of variables that were discussed above as predictors of
AWFC may also be potentially critical moderators of the impaCt of the intervention in
increasing WFDMSE. Broadly, I hypothesize that those individuals who are the most
motivated to learn how to make decisions regarding work and family (i.e., those most
concerned about how they will balance work and family) will show the greatest increases
in WFDMSE from the work-family intervention (based on Noe, 1986). In other words,
the intervention will be most effective in increasing WFDMSE for those who are
motivated to learn about work-family decision-making.

Four demographic characteristics may inﬂuence the extent to which individuals
are motivated to learn about work-family decision-making: gender, age, marital status,

and relationship status. With regard to gender, there is evidence that women play a

31

greater role in caring for children and taking care of household chores than do men (Bond
et al., 2002). Since all participants in this study intend to become doctors, one can argue
that work-family concerns are more salient to women and that they are therefore more
likely to be motivated to learn about work-family deCision-making. Thus, the
intervention is therefore potentially more meaningful and important for them. The
following hypothesis is proposed:

Hypothesis 19: In the Work—family condition, female participants will show a greater
increase in WFDMSE than male participants.

Additionally, older students, those who are married, and those with signiﬁcant
others are more likely to be close to starting a family than those who are younger or
single. Therefore, these students may be more motivated to learn about work-family
decision-making. Thus, the following hypotheses are presented.

Hypothesis 20: In the Work-family condition, older participants will show a greater
increase in WFDMSE than younger participants.

Hypothesis 21 : In the Work-family condition, married participants will show a greater
increase in WFDMSE than unmarried participants.

Hypothesis 22: In the Work—family condition, participants with signiﬁcant others will
show a greater increase in WFDMSE than participants without significant others.

Additionally, individuals differ in the extent to which they place importance on
the family role. I hypothesize that individuals who place a higher importance on the
family role will be more motivated to learn from the work-family intervention and that it
will therefore have a stronger effect on those individuals. Therefore, I propose the

following hypothesis:

Hypothesis 23:1n the Work-family condition, participants higher in family role
importance will show a greater increase in WFDMSE than participants lower in family
role importance.

Finally, personality may inﬂuence the effectiveness of the work-family
intervention in increasing levels of WFDMSE. Individuals with more positive core self-
evaluations feel like they are in control over their lives and that they have the capacity to
inﬂuence their lives in positive ways. Therefore, individuals with positive core self-
evaluations who are in the Work-family condition may be more likely to believe that they
have the capacity to make decisions about balancing work and family. Therefore, they
are more likely to pay attention to the skills and guidance provided in the intervention
than those lower in core self-evaluations (who are more likely to ignore the intervention,
feeling that it cannot help them). In other words, individuals with more positive core
self-evaluations are expected to have a greater increase in self-efﬁcacy as a result of the
intervention than those lower in core self-evaluations.

Hypothesis 24: In the Work-family condition, participants higher in core self-evaluations
will show a greater increase in WFDMSE than participants lower core self—evaluations.
Outcomes of A WF C and WFDMSE

F riede (2005) also examined potential outcomes of AWF C. In the current study, I
replicate some of the more prominent ﬁndings regarding the outcomes of AWF C in that
study and also examine the impact of the WFDMSE on those outcomes of interest. It is
important to consider the potential outcomes of anticipated work-family conﬂict. While
the anticipation of conﬂict may be distressing in its own right, if this construct turned out

to be unrelated to the career and family decision-making of individuals and their eventual

33

work-family conﬂict (once in work and family roles), its merit as a research focus and
opportunities for intervention would be limited. On the other hand, if anticipated work-
family conﬂict is shown to be a driver of important outcomes, research aimed at
addressing work-family issues prior to role entry is extremely valuable.

I have argued that higher WFDMSE is expected to result in decreased AWF C.
Therefore, all hypotheses regarding the relationship between AWF C and outcomes
should also replicate for the relationship between WFDMSE and these same outcomes in
the opposite direction. Resulting hypotheses are therefore presented below following the
hypotheses regarding the relationship between AWF C and outcome variables.

Certainty regarding medical specialty plans and family plans. Both theory and
research suggest that AWFC may be related to work and family decisions. The social
cognitive theory of career development sheds light onto this relationship (Lent, Brown, &
Hackett, 1994). This theory argues that individuals will pursue careers that they think
they will be able to be successful in (self-efﬁcacy) and that they believe will lead to the
outcomes that they desire (outcome expectations; Lent et al., 1994). To the extent that
individuals expect that certain personal and professional choices will lead to an
undesirable outcome (work-family conﬂict) they will be motivated to seek out and pursue
alternative careers.

An example of this process might be a medical student who plans to become an
Obstetrician and decides to talk to an Obstetrician about the pros and cons of this
specialization. Upon talking with the Obstetrician, the medical student learns about the
demanding and unpredictable hours of this specialization, and therefore anticipates work-

family conﬂict. Because the medical student does not want to experience considerable

34

work-family conﬂict, she might become less certain about her decision to be an
Obstetrician and seek out alternative specialization options. Additionally or alternatively,
she might become less certain about her relationship and family plans and seek out
alternatives (e.g., delaying marriage or childbirth, reducing the number of desired
children).

Thus, theory suggests that individuals who anticipate greater levels of work-
family conﬂict may be less certain about their work and family plans. Friede (2005)
examined the relationship between AWFC and certainty regarding medical specialty and
family (marital and parental) plans. In the current study, I also examine these types of
certainty. With regard to certainty regarding medical specialty plans, F riede (2005)
found that greater certainty about one’s choice of medical specialty was associated with
lower levels of four types of AWF C: strain—based conﬂict (both directions) and time-
based conﬂict (both—directions). In other words, those individuals who expected the
greatest amount of time-based and strain—based conﬂict were the least certain about their
future career plans. This could imply that individuals who are anticipating a great deal of
work-family conﬂict are continuing to search for potential alternatives that might result in
less conﬂict. Note that the reverse direction of causality is also possible (i.e., that
individuals who are unsure about which career path they plan to follow expect the most
work-family conﬂict, perhaps because they are unable to prepare for and develop
strategies for balancing work and family since they are uncertain of their work plans).

In the F riede (2005) study it was not possible to test the direction of causality
because it was a correlational study. Because I gather data at three time points in the

current study, I can examine the direction of causality between these two constructs.

35

Based on the social cognitive theory of career development, I expect the data to support
the notion that AWF C causes a lack of certainty regarding medical specialty plans (rather
than the reverse direction of causality; see Figure 2). Therefore, the following
hypotheses are proposed regarding the relationship between AWFC and certainty
regarding medical specialty plans. As noted above, hypotheses regarding the relationship
between WFDMSE and this outcome are also included (see Figure 2).

Hypothesis 25 : Individuals who anticipate greater (a) time-based WIF, (b) time-based
F] W, (c) strain-based WIF, (d) strain-based FI W, (e) behavior—based WIF, and (f)
behavior-based F] W will report less certainty in their choice of medical specialty
(regardless of condition).

Hypothesis 26: The paths M anticipated (a) time-based WIF, (b) time-based F] W, (c)
strain-based WIF, (d) strain—based F] W, (e) behavior-based WIF, and (f) behavior-based
F I W t_o certainty regarding medical specialty plans (i. e., the solid black paths in Figure
2) will be stronger than the paths fr_om certainty regarding medical specialty plans t_o
anticipated (a) time-based WIF, (b) time-based F1 W, (c) strain-based WIF, (d) strain-
based F] W, (e) behavior-based WIF, and (f) behavior-based F] W (i. e., the dashed black
paths in Figure 2).

Hypothesis 27: Over time, individuals with greater WFDMSE will report more certainty
in their choice of medical specialty (regardless of condition).

Hypothesis 28: The pathsm WFDMSE t_o certainty regarding medical specialty plans
will be stronger than the paths M certainty regarding medical specialty plans IL)

WFDMSE.

36

 

 

 

Time 1 Time 2 Time 3
AWFC

F
>
S
'1']
0
V
>
2
“n.
o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time 1 Time 1 Time I
Certainty re. Certainty re. Certainty re.
medical ~ medical - medical
specialty specialty specialty
plans plans plans

 

 

 

 

 

 

 

 

 

Figure 2. Relationship Between Anticipated Work-family Conﬂict and Certainty
Regarding Medical Specialty Plans.
Note. Separate analyses will be conducted for each of the six types of AWF C.

Friede (2005) also investigated the relationship between'certainty regarding
family plans and AWF C. Certainty regarding family plans was associated with four of
the six types of AWF C (F riede, 2005). Individuals who anticipated less behavior-based
FIW, less strain—based conﬂict (both directions), and less time-based WIF were more
certain about their family plans. Again, the causal direction of this relationship could not
be determined by the Friede (2005) study, although certainty was conceptualized as an
outcome of AWF C. Again, the social cognitive theory of career development supports
the notion of certainty as an outcome of AWFC. Therefore, a causal hypothesis is
proposed below (see Figure 3). Additionally, hypotheses regarding the relationship

between WFDMSE and certainty regarding family plans are presented (see Figure 3).

37

Hypothesis 29: Over time, individuals who anticipate greater ((1) time-based WIF, (b)
time-based F] W, (c) strain-based WIF, (d) strain-based F] W, (e) behavior-based WIF,
and (f) behavior-based F] W will report less certainty regarding family plans (regardless
of condition).

Hypothesis 30: T he paths fr_om anticipated (a) time-based WIF, (b) time-based FIW, (c)
strain-based WIF, (d) strain-based F] W, (e) behavior-based WIF, and (f) behavior-based
F] W t_Q certainty regarding family plans (i. e., the solid black paths in Figure 3) will be
stronger than the paths M certainty family plans {9 anticipated (a) time-based WIF", (b)
time-based F] W, (c) strain-based WIF, (d) strain-based FI W, (e) behavior—based WIF,
and (f) behavior-based F] W (i. e., the dashed black paths in Figure 3).

Hypothesis 3] : Over time, individuals with greater WFDMSE will report greater
certainty regardingfamily plans (regardless of condition).

Hypothesis 32: The pathsfr;o_m WFDMSE t_o certainty regarding family plans will be

stronger than the pathsjrom certainty regarding family plans t_o WFDMSE .

38

 

 

 

Time 1 Time 2 Time 3
AWFC

v

>
€
'11
O
V

>
é
"1'1
0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time 1 Time 1 Time 1
Certainty re. Certainty re. Certainty re.
family plans 7 family plans L family plans

 

 

 

 

 

 

 

 

 

Figure 3. Relationship Between Anticipated Work-family Conﬂict and Certainty
Regarding Family Plans.
Note. Separate analyses will be conducted for each of the six types of AWF C.

Information gathering. In the current study, I examine additional potential
outcomes of AWFC. I argue that individuals who anticipate greater levels of work-
family conﬂict will be more motivated to gather information that will help them balance
work and family in the future than those who anticipate less work—family conﬂict,
because they are likely to be more concerned about their balance between work and
family and less certain about their work and family choices. In the current study, there
are several measures of the extent to which individuals seek out information regarding
work-family balance for doctors. First, they have the opportunity to self-report how
much time they have spent during medical school gathering information about balancing
work and family from different sources prior to the intervention. I expect that individuals
with greater incoming levels of anticipated work-family conﬂict also report gathering

more information in the past. I have argued that it is the anticipation of work-family

39

conﬂict that causes individuals to be motivated to seek out information. However,
because this construct is a self-report of past behaviors it does not make sense to examine
its relationship to assessments of AWFC at different points in time. Therefore, this
hypothesis is only assessed prior to the intervention. This hypothesis is also assessed for
WFDMSE.

Hypothesis 33: Individuals with greater anticipated (a) time-based WIF, (b) time-based
F] W, (c) strain-based WIF, (d) strain-based F1 W, (e) behavior-based WIF, and (f)
behavior-based F] W will report gathering more information about balancing work and
family during medical school.

Hypothesis 34: Individuals with greater WFDMSE will report gathering less information
about balancing work and family during medical school.

Information was also gathered after the intervention that represents a desire to
gain additional information about work-family balance. Following the intervention,
participants had the opportunity to sign up to receive an electronic newsletter that
provides information about how doctors can effectively balance work and family roles. I
propose that individuals who have the highest levels of AWFC will be the most likely to
desire this additional information. I have argued that the direction of causality in this
relationship is that AWF C motivates individuals to seek out information. This causal
relationship can be examined in comparison to its converse. Therefore, I propose the
following hypotheses (see Figure 4). Furthermore, hypotheses are replicated for
WFDMSE in the opposite direction.

Hypothesis 35: Individuals with greater anticipated (a) time-based WIF, (b) time-based

F] W, (e) strain—based WIF, (d) strain-based F] W, (e) behavior-based WIF, and (D

40

behavior-based FI W will be more likely to request the electronic newsletter (regardless
of condition).

Hypothesis 36: T he path ﬁ'o_m_ anticipated (a) time-based WIF, (b) time-based FI W, (c)
strain-based WIF, (d) strain-based F1 W, (e) behavior-based WIF, and (f) behavior-based
F1 W t_o requesting the electronic newsletter (i. e., the solid black path in Figure 4) will be
stronger than the path fﬂﬂ requesting the electronic newsletter t_o anticipated (a) time-
based WIF, (b) time-based F] W, (c) strain-based WIF, (d) strain-based FI W, (e)
behavior-based WIF, and (f) behavior-based FI W (i. e., the dashed black path in Figure
4 ).

Hypothesis 3 7: Individuals with greater WFDMSE will be more likely to request the
electronic newsletter (regardless of condition).

Hypothesis 38: The pathm WFDMSE {Q requesting the electronic newsletter will be

stronger than the path ﬁom requesting the electronic newsletter t_o WFDMSE.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time 1 Time 2 Time 3
AWF C ~ AWF C AWFC
I ’v
Time 2
Requests for
electronic
newsletter

 

 

 

Figure 4. Relationship Between Anticipated Work-family Conﬂict and Requests
for the Electronic Newsletter.
Note. Separate analyses will be conducted for each of the six types of AWFC.

41

Furthermore, for those participants who receive the electronic newsletter, the
newsletter included a link to a website with additional information about work-family
balance issues. This represents another form of information-gathering. Given the logic
presented previously, the following hypotheses are therefore presented (see Figure 5).
The hypotheses are replicated in the opposite direction for WFDMSE.

Hypothesis 39: Individuals with greater anticipated (a) time-based WIF, (b) time-based
F] W, (c) strain-based WIF, (d) strain-based F] W, (e) behavior-based WIF, and ([9
behavior-based F] W will be more likely to view the website (regardless of condition).
Hypothesis 40: The pathﬁgoﬂ anticipated (a) time-based WIF, (b) time-based F] W, (c)
strain—based WIF, (d) strain-based F] W, (e) behavior-based WIF, and (f) behavior-based
F] W t_o viewing the website (i. e., the solid black path in Figure 5) will be stronger than
the pathﬂo_m viewing the website _l_Q anticipated (a) time—based WIF, (b) time-based F] W,
(c) strain-based WIF, (d) strain-based F] W, (e) behavior-based WIF, and (f) behavior-
based F] W (i. e.. the dashed black path in Figure 5).

Hypothesis 4]: Individuals with greater WFDMSE will be less likely to view the website
(regardless of condition).

Hypothesis 42: The pathﬁm WFDMSE _l_0_ viewing the website will be stronger than the

path [ram viewing the website t_o WFDMSE.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time 1 Time 2 Time 3
AWFC AWFC - AWF C
I If
Time 2
Views the
website

 

 

 

Figure 5. Relationship Between Anticipated Work-family Conﬂict and Viewing
the website.

Note. Separate analyses will be conducted for each of the six types of AWF C.

At Time 3, participants were asked to report how much time they spent gathering
information about balancing work and family in the two-week time lag between the two
parts of the experiment. Again, individuals who anticipate greater conﬂict are expected
to have spent more time gathering information. Given the fact that this information was
collected at Time 3, it is not possible to examine the direction of causality between
AWFC and gathering information. One would require a time-lag after the assessment at
Time 3 to look at the causal relationship. Therefore, the relationship is examined at a
single time point (at Time 3). Again, this hypothesis is replicated for WFDMSE.
Hypothesis 43: Individuals with greater anticipated (a) time-based WIF, (b) time-based
F] W, (c) strain-based WIF, (d) strain-based F] W, (e) behavior-based WIF, and (f)
behavior-based F l W will report spending more time gathering information during the

two-week time lag.

43

Hypothesis 44: Individuals with greater WFDMSE will report spending less time
gathering information during the two-week time lag.
Mediational Hypotheses

Finally, the hypotheses presented above all suggest that WFDMSE inﬂuences
level of AWFC and that level of AWF C inﬂuences the outcomes described above. In
other words, I argue that AWFC mediates the relationships between WFDMSE and the
outcomes of interest. Partial mediation is hypothesized because WFDMSE could also
have a direct effect on the outcomes of interest. Therefore, the following hypotheses are
proposed. See Figure 6 to see a graphic depiction of the hypotheses for certainty
regarding medical specialty plans and certainty regarding family plans (which are all
assessed at three points in time). Hypotheses 45 through 50 are depicted in Figure 6
through 10, respectively.
Hypothesis 45: Anticipated (a) time-based WIF, (b) time-based F] W, (c) strain-based
WIF, (d) strain—based Fl W, (e) behavior-based WIF, and (f) behavior-based F I W
partially mediate the relationship between WFDMSE and certainty regarding medical
specialty plans (regardless of condition).
Hypothesis 46: Anticipated (a) time-based WIF, (b) time-based F] W, (e) strain-based
WIF, (d) strain-based F] W, (e) behavior-based WIF, and (f) behavior-based Fl W
partially mediate the relationship between WFDMSE and certainty regardingfamily

plans (regardless of condition).

44

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time 1 Time 2 Time 3
WFDMSE - WFDMSE WFDMSE
Time 1 'me 2 Time 3
AWF C A AWF C
Time 1 Time 2 Time 3
Certainty re. Certainty re. Certainty re.
plans ’ plans ’ plans

 

 

 

 

 

 

 

 

 

Figure 6. Mediational Relationship Between Work-family Decision-making Self-
efﬁcacy, Anticipated Work-family Conﬂict, and Certainty Regarding Plans.
Note. Separate analyses will be conducted for each of the six types of AWF C and
both types of certainty regarding plans (medical specialty and family).

Hypothesis 47: Anticipated (a) time—based WIF, (b) time—based F] W, (c) strain-based
WIF, (d) strain-based F I W, (e) behavior-based WIF, and (f) behavior-based F] W
partially mediate the relationship between WFDMSE and information—gathering during
medical school (regardless of condition).

Hypothesis 48: Anticipated (a) time-based WIF, (b) time-based F] W, (e) strain-based
WIF, (d) strain-based F l W, (e) behavior-based W IF, and (/) behavior-based F I W
partially mediate the relationship between WFDMSE and requests to receive the

electronic newsletter (regardless of condition).

45

Hypothesis 49: Anticipated (a) time-based WIF, (b) time-based F1 W, (c) strain-based
WIF, (d) strain-based F] W, (e) behavior-based WIF, and (f) behavior-based F] W
partially mediate the relationship between WFDMSE and the likelihood of viewing the
website (regardless of condition). '

Hypothesis 50: Anticipated (a) time-based WIF, (b) time-based F] W, (c) strain—based
WIF, (d) strain~based FI W, (e) behavior-based W IF, and (f) behavior-based F] W
partially mediate the relationship between WFDMSE and self-reports of gathering

information during the two week time lag (regardless of condition).

46

 

Time 1
WFDMSE

 

 

 

 

 

Time 1
AWFC

 

 

 

 

Ir

Time 1
Gathering
information
during
Medical
School

 

 

 

 

 

Figure 7. Mediational Relationship Between Work-family Decision-making Self-
efﬁcacy, Anticipated Work-family Conﬂict, and Gathering Information During
Medical School. -

Note. Separate analyses will be conducted for each of the six types of AWFC.

47

 

Time 1
WFDMSE

 

 

 

 

Time 2
AWFC

Time 2
Request for

Electronic
Newsletter

 
  

 

 

 

 

 

 

 

Figure 8. Mediational Relationship Between Work-family Decision-making Self-
efﬁcacy, Anticipated Work-family Conﬂict, and Request for Electronic
Newsletter. .

Note. Separate analyses will be conducted for each of the six types of AWFC.

48

 

Time 1
WFDMSE

 

 

 

 

   

 

 

 

Time 2

Viewing
website

 

   

 

 

 

Figure 9. Mediational Relationship Between Work-family Decision—making Self-
efﬁcacy, Anticipated Work-family Conﬂict, and Viewing the Website.
Note. Separate analyses will be conducted for each of the six types of AWFC.

49

 

Time 1
WFDMSE

 

 

 

 

 
  

 

 

 

\

Time 3
Gathering
Information
During 2
Week Lag

 

 

 

 

Figure 10. Mediational Relationship Between Work-family Decision-making Self-
efﬁcacy, Anticipated Work-family Conﬂict, and Gathering Information During
Two Week Time Lag.

Note. Separate analyses will be conducted for each of the six types of AWFC.

Method
Participants
The sample for this study was comprised of medical students of all years at
Michigan State University in the College of Human Medicine (CHM; approximately 480
medical students total) and the College of Osteopathic Medicine (COM; approximately
713 medical students total). Medical students were selected as the target sample for this
study because they are required to make concrete career decisions prior to leaving

medical school (i.e., which medical specialization to enter). Further, there are a limited

50

number of discrete specialties available for selection. For these reasons, medical students
are a practical sample for examining career decision-making, as opposed to
undergraduates or working employees who select jobs and careers from a broader range
of options, thus limiting the ability to make comparisons across participant career
choices.

The necessary sample size that is needed to reach an acceptable level of power for
the study hypotheses was examined. There are three major types of analyses being
performed in this study: correlations (including regressions and mediation), ANOVAs,
and general linear modeling (i.e., a continuous variable interacting with a categorical
variable). Power analyses were conducted to examine the necessary sample size for each
of these analyses. In each case, I examined the sample size necessary to detect small and
medium effect sizes, as described by Cohen (1969).

To determine the necessary sample size for correlational analyses, I used the
following information: alpha for a one-tailed test = .05 (a one-tailed test is used because
the correlational hypotheses are directional), power = .80, effect size = .1 (small), .3
(medium; Cohen, 1969). The necessary n to detect a small effect size is 618, while the
necessary n to detect a medium effect size is 68 (Cohen, 1969).

To determine the necessary sample size for the analyses involving ANOVAs, I
used the following information: alpha = .05, power = .80, u (k —1; where k = the number
of groups) = 2, effect size = .1 (small), .25 (medium; Cohen, 1969). The necessary n to
detect a small effect size is 322, while the necessary n to detect a medium effect size for

these analyses is 52 (Cohen, 1969).

51

Finally, to determine the necessary sample size for the GLM analysis (in which a
continuous variable interacts with a categorical variable), the power analysis for
differences among correlations can be used as a proxy (N. Schmitt, personal
communication, November 2, 2006). In this case, Iiused the following information: alpha
for a one-tailed test = .05 (a one-tailed test is used because the hypotheses are
directional), power = .80, effect size = .1 (small), .3 (medium; Cohen, 1969). The
necessary n to detect a small effect size is 1,240, while the necessary n to detect a
medium sample size is 140 (Cohen, 1969).

The ﬁnal sample for the current study is 108 participants. Therefore, medium
effect sizes should be detectable in the analyses using ANOVAs and correlations. A
medium-effect size using GLM will not be detectable given the current sample.

Design and Procedure

A link to a web-based questionnaire was sent via email to all medical students
from Deans of their respective medical schools. The email deScribed the nature of the
study and explained how students could participate (see Appendix A). Participation in
the survey was voluntary and conﬁdentiality of responses was assured to the participants.
Individuals who participated in Time 1 (pre-test immediately preceding the intervention)
and Time 2 (post-test immediately following the intervention) data collections received a
$5 check in the mail. Individuals who participated in the 2 week follow-up (Time 3)
received an additional $5 (i.e., they received a $10 check in the mail).

The web-survey automatically randomly assigned participants to one of two
conditions. At Time 1, participants completed pre-intervention survey items, participated

in an intervention (Work-family or Career; the interventions are described later), and then

completed the post-intervention survey items (Time 2). Note that there ideally would
have been a greater time lag between the administration of the pre-intervention survey
and the administration of the post-intervention survey (e. g., to reduce memory effects).
However, due to concerns regarding participant attrition, the Time 1 items were
completed immediately prior to the intervention, while the Time 2 items were completed
immediately after the intervention. See Appendices B and C for the consent forms that
were completed Time 1 and Time 3.

For those individuals who chose to participate in the study and provided consent
to be re-contacted, a follow-up email was sent to them two weeks after their initial
participation asking them to visit the website again to participate in Time 3 of the study
(see Appendix D). 91 out of the 108 original participants completed Time 3 of the study
(84.3%). Each individual had a two week window in which to participate in Time 3 of
the study. Participants who completed Time 3 read a debrieﬁng letter at the end of the
survey (see Appendix E). Those who did not participate in the Time 3 received the
debrieﬁng letter via email after the window had closed on participation for all
participants. It was not possible to provide a ﬁll] debrieﬁng at the end of the initial
survey completion so that study hypotheses were not revealed prior to the end of the
experiment.

After completing the Time 2 survey, participants were asked if they would like to
receive an electronic newsletter about the work-family challenges facing medical
students. 55.6% of participants agreed to receive the newsletter. There was no difference
across conditions in the likelihood that students consented to receive the newsletter (F =

.59, p = .44).

53

Participants who consented to receive the electronic newsletter, received an email
the day that they completed the survey (see Appendix F). The email provided a link to a
website about the work-family challenges facing medical students. The website tracked
which students visited the site based on their ID number. 204% of all participants visited
the website. Of those who consented to receive the email, 36.7% visited the website.
There was no difference across conditions in the likelihood that students visited the
website (among those who received the newsletter (F = .15, p = .70)).

Experimental Conditions

Only the interventions differed across experimental conditions. Time 3 of the
study was the same regardless of experimental condition.

In the Career condition, participants were asked to complete the pre-intervention
survey items (Time I), engage in the Career experimental intervention, and then
immediately complete the post-intervention survey items (Time 2). The design for this
condition was based on two critical features. First, Crites (1961) identiﬁed ﬁve
components of career decision self-efﬁcacy: conﬁdence in self-appraisal, gathering
information, goal selection, planning, and problem-solving. The current study focuses on
the ﬁrst three of those components: self-appraisal, gathering information, and goal
selection. Additionally, Bandura and Adams (1977) suggested that self-efﬁcacy can be
enhanced in four distinct ways: performance accomplishments, observational learning,
emotional arousal, and verbal persuasion. Betz (1992) discussed the relevance of
applying these four methods to the enhancement of CDSE. Performance
accomplishments refer to providing opportunities for individuals to undertake the task

that they lack self-efﬁcacy for (Bandura & Adams, 1977; Betz, 1992). According to

54

Bandura and Adams (1977), it is important that these performance accomplishments
result in success for the participant, thus increasing their conﬁdence (see also Betz,
1992). Observational learning refers to Ieaming from role models, who demonstrate
success in the area in which the individual lacks conﬁdence (Bandura and Adams, 1977,
Betz, 1992). Emotional arousal refers to the importance of calming the anxiety of
individuals with regard to the area in which he/she lacks self-efﬁcacy. Relaxation
techniques, adaptive self-talk, and thought stopping are examples of mechanisms that are
expected to increase self-efﬁcacy through emotional arousal (Bandura & Adams, 1977,
Betz, 1992). Finally, verbal persuasion refers to the encouragement, support, and
positive reinforcement provided to improve self—efﬁcacy (Bandura & Adams, 1977, Betz,
1992). Note that emotional arousal was not included in the current intervention because
of the time necessary to effectively teach and implement relaxation techniques.
Appendix G provides the powerpoint slides utilized in the Career condition along
with a transcript of the automated recorded audio lecture that accompanied each slide.
There were three sets of powerpoint slides included (called tutorials; i.e., self-appraisal,
information-gathering, and goal selection). After viewing each set of powerpoint slides,
participants completed a brief task before moving onto the next set of slides (Appendix G
also includes the tasks that followed each Tutorial). In the Career condition, the average
time to complete Time 1 and Time 2 surveys (including participating in the intervention)
of the study for the ﬁnal sample was 53.13 minutes (sd = 24.99 minutes).
In the Work-family condition (see Appendix H for intervention materials),
participants were asked to complete the pre-intervention survey items (Time I), engage in

the Work-Family experimental intervention, and then immediately complete the post—

55

intervention survey items (Time 2). Note that the work-family intervention was designed
to be parallel to the Career condition in all ways except the content was altered from a
focus on career choices broadly to a focus on work-family issues in career choices. In the
Work-family condition, the average time to complete Time 1 and Time 2 surveys
(including participating in the intervention) for the ﬁnal sample was 50.54 minutes (sd =
30.0 minutes).

Although participants were randomly assigned to condition, I nonetheless
examined whether the participants differed by condition across any of the key
demographic variables. There were no differences between condition in gender (F = 1.54,
p = .22), school (F = 0.04, p = .85), age (F = 0.01, p = .93), marital status (currently
married or not; F = .08, p = .78), relationship status (i.e., whether in a committed
relationship or not; F = 0.62, p = .44), or parental status (parent or not; F = .97, p = .33).
Percentage of White participants varied across conditions. The Career condition was
88.9% White, whereas the Work-Family condition was 72.2% White (F = 4.92, p = .03).
This difference is likely due to chance and it is therefore not considered further unless
speciﬁed.

Data Cleaning

A number of steps were taken to clean the data prior to analyses. First, I
examined whether any participants names appeared in both conditions. Because the
survey was web-based, students sometimes accessed the survey, looked it over, and
exited without completing the survey. Then, they revisited the website later to actually
complete the survey. In these cases. their names appeared in both conditions. By

reviewing whether they ﬁlled out the survey and how long they took to do so, it is

56

apparent which survey they intended to complete fully. Seven participants were deleted
from one condition for this reason and their data was kept in the other condition. One
participant was deleted from both conditions in which her name appeared because
surveys were not completed in either condition.

Next, I examined the data for participants whose times to complete Time 1 and
Time 2 surveys (including participating in the intervention) of the survey were more than
one standard deviation below the mean completion time in their respective condition
(Work-family or Career). The mean completion time for the Career condition for all
participants was 45.33 minutes (sd = 29.36). Ten participants had completion times more
than one standard deviation below the mean in the Career condition. Upon further
examination of the responses of these participants, all participants showed large
quantities of missing data and they were therefore not included in the ﬁnal data set. The
same procedure was followed for the Work-family condition. The mean time to
complete Time I and Time 2 surveys (including participating in the intervention) of the
Work-Family condition for all participants was 42.69 minutes (sd = 10.79). Nine
participants in this condition completed these surveys more than one standard deviation
below the mean. Again, an examination of their data showed large quantities of missing
data and they were not included in the ﬁnal data set.

At this point, the data set consisted of 1 14 participants. The next step was to
examine the responses of remaining participants with large quantities of missing data.
Seven participants were identiﬁed as missing considerable amounts of data. Of those
seven, four terminated participation prior to the experimental intervention. Therefore,

their data was deleted. The remaining three participants terminated participation after the

57

experimental intervention. Therefore, I was able to examine their responses to the tasks
associated with the experimental interventions to determine whether they appeared to be
taking the study seriously. All three participants completed those tasks seriously.
Therefore their data was retained (two participants from the Career condition and one
from the Work-family condition).

Next, I examined the data from Time 3 of the study. I examined the data of
participants who completed Time 3 in an amount of time more than one standard
deviation below the mean for each condition (Note that all participants completed the
same survey in Time 3 regardless of condition). The mean completion time for
participants in the Career condition was 10.23 minutes (sd = 5.77). Four participants
completed the survey more than one standard deviation below the mean. An examination
of their data showed that they appeared to have taken the survey seriously and thus their
data was retained. In the Work-family condition, the mean time for completion was
calculated by excluding the data of two participants whose time to complete the survey
was exceptionally long (most likely they left the survey open on their computers and
returned to complete it at a later time). Excluding these two values, the mean time to
complete Time 3 in the Work-family condition was 8.00 minutes (sd = 3.12). Five
participants had completion times more than one standard deviation below the mean.
Again, an examination of their data showed that they appeared to take the survey
seriously and therefore their data was retained.

The next step was to see if there were any other participants with large quantities

of missing data in Time 3 of the study. One participant was identiﬁed who did not

58

complete the vast majority of the Time 3 survey and this participant’s Time 3 data was
deleted.

Finally, I examined whether there were any individuals whose data was kept for
Time 3 but deleted from the data collected for Time 1 and Time 2. This was the case for
one participant and that participant was therefore excluded from the study. The ﬁnal data
set consisted of 108 participants in Time 1 and Time 2 and 91 participants in Time 3.

Table 2 shows demographic information for the ﬁnal sample.

59

Table 2
Demographic Information by Medical School

 

 

COM CHM Total
Final number of participants in Time 1 and 2 55 53 108
Final number of participants in Time 3 47 44 91
Work-family condition 49.1 50.9 50.0
lSt year participants 50.9 30.2 40.7
2nd year participants 32.7 32.1 32.4
3rd year participants 16.4 22.6 19.4
4‘h year participants 0 15.1 0
Age: 22 or younger 0 0 0
Age: 23 — 24 56.4 39.6 48.1
Age: 25 — 26 27.3 30.2 28.7
Age: 27 — 28 9.1 15.1 12.0
Age: 29 or older 7.3 15.1 11.1
Female 70.9 66.0 68.5
Currently married 18.2 32.1 25.0
In a relationship 65.5 56.6 61.1
Having one or more children that live with 7.3 13.2 10.3
them
African-American, Black 3.6 3.8 3.7
Asian 7.3 3.8 5.6
Hispanic 0 i 1.9 .9
Caucasian, White 93.6 77.9 80.6

 

Notes. Numbers in table are percentages except for Final number of participants in

Time 1 and 2 and Time 3. COM = College of Osteopathic Medicine. CHM =

College of Human Medicine.
Measures

Anticipated Work-Family Conﬂict. F riede (2005) adapted the 18-item work-

family conﬂict scale created by Carlson et al. (2000) to reﬂect the measurement of
“anticipated” conﬂict. All items were rewritten in a future tense. Six types of anticipated
work-family conﬂict were measured including anticipated behavior-based work-

interference-with-family (behavior—based WIF), anticipated behavior-based family—

interference—with-work (liehc'tvior-based F I W). anticipated strain-based work-

60

interference-with-family (strain-based WIF), anticipated strain-based family-
interference—with-work (strain-based F] W), anticipated time-based work-interference-
with-family (time—based WIF) and anticipated time-based family-interference-with-work
(time—based F! W). The measure consists of 18 items ranging from 1 (strongly disagree) to
5 (strongly agree). The measure was administered at three points during the study: Time
1, Time 2, and Time 3. See Appendix I for the anticipated work-family conﬂict items.

Carlson et al. (2000) found that a structural equation model with six correlated
latent factors ﬁt their data measuring actual work-family conﬂict best. F riede (2005)
conducted a conﬁrmatory factor analysis to examine whether a six factor structure ﬁt the
measure of anticipated work—family conﬂict, as well A six factor model was compared to
a three factor model (which reﬂected anticipated time-, strain-, and behavior-based
conﬂict — ignoring direction of conﬂict), a two factor model (anticipated work-
interference with family and family-interference-with-work — ignoring type of conﬂict),
and a unidimensional model (all anticipated work—family conﬂict items). Factors were
allowed to correlate. Friede (2005) found the six factor model to have the best ﬁt.

In the current study, the factor structure at each time point was examined (see
Appendix] for item intercorrelations). Conﬁrmatory factor analyses were conducted to
examine whether the data collected from this study also fall into this six-dimensional
factor structure in each administration. A six factor model was compared to a three factor
model, a two factor model, and a one factor model at each time point. Factors were
allowed to correlate (in the six factor model, factors were only allowed to correlate with
factors that reﬂected the same type and direction of anticipated conﬂict — the sample size

was not large enough to support correlations among all latent variables). The models

6|

were not nested; therefore chi—square difference tests are not appropriate. Nonetheless, a
comparison of the ﬁt indices indicate that at all three time points, the six dimensional
model had the best ﬁt (see Appendix K). This replicates the ﬁndings of Friede (1995) for
the “anticipated” version of the measure. Six dimensional scales are therefore used at

each time point. Scale reliabilities are presented in Table 3.

 

 

Table 3
Reliabilitiegfor Anticipated Work-family Corylict Scales
Time I Time 2 Time 3

Time—based WIF .77 .86 .86
Time-based F IW .73 .80 .87
Strain-based WIF .82 .89 .89
Strain-based F IW .93 .92 .94
Behavior—based WIF .79 .85 .80
Behavior-based FIW .93 92 88

 

Notes. WIF = Work-interference-with-family. F IW = F amily-interference-with-work.

Certainty regarding medical specialty plans. Each participant was asked to
report what medical specialty they felt that they were most likely to enter (given 25
options; see Appendix L) in Time 1. Frequencies are provided for descriptive purposes

(see Table 4).

62

Table 4
Descriptive Statistics for Medical Specialty Plans

 

 

Specialty COM CHM Career Work- Total
‘ Family
Number of participants 46 46 48 44 92
Anesthesiology 0 2.2 0 2.3 1.1
Dermatology 0 0 0 0 0
Emergency medicine 10.9 21.7 14.6 18.2 16.3
Family practice 13.0 10.9 10.4 13.6 12.0
Internal medicine 10.9 0 4.2 9.1 6.5
Medical genetics 0 0 0 0 0
Neurological surgery 0 2.2 2.1 0 1.1
Neurology 6.5 0 2.1 4.5 3.3
Nuclear medicine 0 0 0 0 0
Obstetrics and gynecology 13.0 10.9 16.7 6.8 12.0
Ophthalmology 0 0 0 0 0
Orthopedic surgery 0 2.2 2.1 0 1.1
Otolaryngology 2.2 0 2.1 0 1.1
Pathology 0 0 0 0 0
Pediatrics 8.7 8.7 8.3 9.1 8.7
Physical medicine and rehabilitation 6.5 0 0 6.8 3.3
Plastic surgery 2.2 0 2.1 0 1.1
Psychiatry 2.2 4.3 4.2 2.3 3.3
Radiation oncology 0 2.2 . 2.1 0 1.1
Radiology — diagnostic 6.5 8.7 10.4 4.5 7.6
Surgery — general 4.3 4.3 6.3 2.3 4.3
Urology 0 2.2 O 2.3 1.1
Other 10.9 10.9 12.5 9.1 10.9
Unsure 2.2 6.5 0 9.1 4.3

 

Notes. All values are presented as percentages except the number of participants. COM =
College of Osteopathic Medicine. CHM = College of Human Medicine. Career = C areer.
condition. Work-family = Work-family condition.

Three items assessed certainty regarding medical specialty (see Appendix L).
These items were assessed in Time 1, Time 2, and in Time 3. The second and third items
were recoded so that higher scores represent greater certainty.

A principal axis factor analysis using a varimax rotation was conducted for each

administration of the measure to examine whether the three items can be combined into a

63

single scale (see Appendix M for item intercorrelations). For all three administrations of
the measure, a single factor was extracted. Therefore, an overall scale score representing
certainty about medical specialty was created for each administration. The reliability of
the measure in Time 1 was a = .91. The reliabilityiof the measure in Time 2 was at = .88.
The reliability of the measure in Time 3 was a = .88.

Certainty regarding family plans. Participants answered ﬁve questions about their
plans for their future family lives (see Appendix N) in Time 1. Frequencies are provided
for descriptive purposes (see Table 5). Participants also answered three questions
regarding their certainty about their family plans at Time 1, Time 2, and Time 3 (see
Appendix N). The second and third items were recoded so that higher scores indicate
greater certainty.

Table 5
Descriptive Statisticsfor Family Plans

 

COM CHM Career Work- Total

 

. Family
% that plan to marry 98.2 86.8 96.3 88.9 92.6
% that plan to marry before age 30 52.7 30.2 42.6 40.8 41.7
% that plan to have children 87.3 81.1 90.7 77.8 84.3
% that plan to have 2 or 3 children 63.6 62.3 72.2 53.7 63.0

% that plan to have children before age 30 34.5 9.4 24.1 20.4 22.2

 

Notes. COM = College of Osteopathic Medicine. CHM = College of Human Medicine.
Career = Career condition. Work-family = Work-family condition.

A principal axis factor analysis using varimax rotation was conducted for each
administration of the measure to examine whether the three items can be combined into a
single scale (see Appendix 0 for item intercorrelations). For all three administrations of
the measure, a single factor was extracted. Therefore, an overall scale score representing

certainty about family plans was created for each administration. The reliability of the

64

measure in Time 1 was a = .85. The reliability of the measure at Time 2 was a = .86. The
reliability of the measure at Time 3 was a = .89.

Career decision self-efficacy scale. The Betz et al. (1996) measure of CDSE was
adapted to reﬂect the choice of medical specialization (as opposed to choice of major and
career in the original measure). The measure has ﬁve facets based upon Crites’ (1961)
ﬁve Career Choice Competencies (Betz et al., 1996). They are: l) accurate self-
appraisal; 2) gathering occupational information; 3) goal selection; 4) making plans for
the future; and 5) problem solving. Betz and colleagues offer two forms of the scale: the
long form (50 items, 10-point response scale) and the short form (25 items, S-point
response scale). The short form of the measure has been shown to have equivalent
reliability and validity to the original measure (Betz et al., 1996). Items are rated on a 5-
point scale ranging from 1 (no conﬁdence at all) to 5 (complete conﬁdence). According
to Betz et al. (1996), subscale scores (the mean of the ﬁve items that reﬂect each facet)
can be used and an overall mean (across all 25 items) can also be used. Two unpublished
doctoral dissertations (Paulsen, 2001 and Smith, 2001; from Taylor & Betz, 2006) found
each of the overall measures and each ﬁve subscales to have reliabilities over a = .80 in
samples of 603 and 423 students, respectively. Prior research has shown the
effectiveness of this measure as a tool for evaluating the effectiveness of interventions
(pre— and post-intervention measures; Bergeron & Romano, 1994; Betz, 1992; Betz &
Luzzo, 1996; McAuliffe, 1991; Peterson, 1993).

In the current study, 1 adapted items from the shorter form of this measure. Due
to the fact that only three of the ﬁve facets were included in the intervention (self-

appraisal, gathering infomiation, and goal selection), only the ﬁve items representing

each of those facets were included (15 items total). For the current study, I adapted the
measure to reﬂect the career decision-making of medical students focusing on their
choice of medical specialization (see Appendix P). Items were rated on a 5-point scale
ranging from 1 (no conﬁdence at all) to 5 (complete conﬁdence). The measure was
administered at Time 1, Time 2, and Time 3.

As mentioned above, Betz et al. (1996) reported that the overall scale measure
could be used as a general measure of career decision-making self-efﬁcacy. See
Appendix Q for item intercorrelations and Appendix R for the results of the CFA. CFAs
were conducted for the measure at each time point. At each time point, CDSE items
from the three assessed facets (self-appraisal, gathering information, and goal selection)
each were loaded on to their respective latent factors. A higher order latent factor
representing overall CDSE was also included in the model. Error terms within a facet
were correlated based on modiﬁcation indices (this accounts for the differing degrees of
freedom across time periods). At Time 1 and Time 3, the error variance of the goal
selection latent variable was ﬁxed to .05 because the solution was not admissible without
this adjustment (i.e., it was a Heywood case). At all three time points, each of the three
latent factors signiﬁcantly loaded onto the second-order latent factor representing career
decision-making self-efﬁcacy. Also, the ﬁt of the model is good at all time points. These
ﬁndings support the notion that there is some higher-order career decision self—efﬁcacy
construct that is measured across all three of the facets. Therefore, I will use a measure
of overall CDSE (mean of all items) as opposed to the subscales. The reliability for the

Time 1 overall scale was a = .91. The reliability for the Time 2 overall scale was a = .93.

66

The reliability for the Time 3 overall scale was a = .91. Note that in all cases, these
reliabilities are equal to or higher than the reliabilities of the 5-item facet scales.

Work-family decision—making self-efficacy. This measure was adapted from the
ﬁnal version of the CDSE scale used in the current study (described above). Items were
re-written to focus on career decision-making with speciﬁc regard to work-family issues.
Similar to the career decisions self-efﬁcacy scale described above, only the competencies
of self-appraisal, gathering information, and goal selection were measured for work-
family decision-making self-efﬁcacy. This measure consists of 15 items rated on a 5-
point scale from I (no conﬁdence at all) to 5 (complete conﬁdence). Items are presented
in Appendix S.

Given that this measure is an adaptation of the CDSE scale described above, the
same analyses were conducted to examine the factor structure of this measure at each of
the three time points. At each time point, a CFA was conducted in which the work-
family decision-making self—efﬁcacy items from the three asSessed facets (self-appraisal,
gathering information, and goal selection) each were loaded on to their respective latent
factors. A higher order latent factor representing overall work-family decision-making
self-efﬁcacy was also included in the model. Error terms within a facet were correlated
based on modiﬁcation indices (this accounts for the differing degrees of freedom across
time periods). Each of the three latent facets signiﬁcantly loaded onto the second-order
latent factor. See Appendix T for item intercorrelations and Appendix U for the results of
the C F A. The ﬁt of the model is also good, thus supporting the idea that there is an
overarching WFDMSE construct that is indicated by the three facets. The reliability for

the Time 1 overall scale was a = .95. The reliability for the Time 2 overall scale was (1 =

67

.96. The reliability for the Time 3 overall scale is a = .94. Again, these reliabilities are
equal to or greater than the reliabilities of any of the S-item facet scales. I will use a
measure of overall WFDMSE (mean of all items) as opposed to the subscales for the
current study.

Role-planning knowledge. F riede (2005) measured role-planning knowledge.
This was taken from the “knowledge/certainty” subscale of the Attitudes Towards
Multiple Role Planning scale created by Weitzman and Fitzgerald (1996). This scale
measures beliefs about the extent to which one is knowledgeable about how to plan for
multiple roles and conﬁdence in the ability to do so (Weitzman & Fitzgerald, 1996).
Friede (2005) found. role-planning knowledge to be a signiﬁcant predictor of all forms of
anticipated work-family conﬂict. Friede (2005) found this measure to be unidimensional
with a reliability of a = .90. The same measure was used in the current study. The
measure is 10 items and responses are indicated on a 5-point scale from 1 (strongly
disagree) to 5 (strongly agree). This measure is administered in Time 1, Time 2, and
Time 3 (see Appendix V for items)

A principal axis factor analysis with a varimax rotation was conducted to assess
the factor structure of the Time 1 items (see Appendix W for item intercorrelations and
Appendix X for results of the factor analysis). Two factors were extracted. All items had
their highest loading on the ﬁrst factor except for Items 7 and 9. The ﬁrst factor explains
more than ﬁve times the variance of the second factor. The reliability for the overall
scale was a = .90. The reliability of the scale with items 7 and 9 removed was examined.
The deletion of these items did not improve the overall scale reliability. All items were

retained for the creation of a single measure of Time 1 role-planning knowledge.

68

The same steps were followed to examine the item characteristics in the Time 2
administration of the measure (see Appendix W for item intercorrelations and Appendix
X for results of the factor analysis). Again, the results of a principal axis factor analysis
with a varimax rotation showed that two factors were extracted. This time, items 5, 8,
and 10 had their highest loadings on the second factor. The reliability of the overall scale
with all ten items included is a = .89. An examination of the three items that loaded on
the second factor showed that the removal of these items did not improve the reliability
of the overall scale. Therefore, all items were retained in the overall scale.

The same steps were followed to examine the item characteristics in the Time 3
administration of the measure (see Appendix W for item intercorrelations and Appendix
X for results of the factor analysis). A principal axis factor analysis with a varimax
rotation extracted one factor. The reliability of the overall scale with all ten items
included is a = .91.

Gender. Participants were asked to report their gender (male, female) in Time 1
and it was coded male = 0, female = 1 (see Appendix Y). This was assessed prior to the
intervention. The overall sample was 68.5% female at the start of the assessment. In
Time 3 of the study, the sample was 68.1% female (see also Table 1). There was no
differential attn'tion from the study based on gender (F = .04, p = .84).

Age. Participants were asked to select their age on a 5-option scale ranging from
1 (22 or younger) to 5 (29 or older; in Time I; see Appendix Y). This was treated as a
continuous variable. See Table l for the percentage of participants falling into each age

range. There was no differential attrition from the study based on age in a comparison of

69

those under 25 with those over 25 (approximately half the sample was under 25 and the
other half was 25 or older; F = .00, p = .97).

Marital status. Participants were asked to report their current marital status given
the following ﬁve options (never married, currently married, divorced or separated,
widowed, long-term live-in relationships in Time 1; see Appendix Y). 74.1% of the
participants were never married, 25.0% were currently married, and .9% were divorced or
separated (see also Table 1). No participants reported being widowed or in a long-term
live-in relationship. This was assessed in Time 1. For analytical purposes, I coded this
variable 1 = currently married, 0 = all other options). There was no differential attrition
from the study based on marital status (F = 1.88, p = .17).

Additional personal characteristics. Additional personal characteristics were
assessed in Time 1 for the purposes of describing the sample. These are: year in medical
school (treated as a continuous variable), parental status (coded 0 = do not have a child
living with me, 1 = have a child living with me) ethnicity (O = all other ethnicities, l =
White), and age of youngest child (this variable was not utilized because it correlated r =
.89 with parental status). See Appendix Y for items and Table 1 for a description of the
sample along these characteristics.

Core selflevaluations. Core self—evaluations was measured in Time 1 using the
Core Self-Evaluations Scale (CSES) developed by Judge, Erez, Bono, and Thoresen
(2003). The CSES is a unidimensional l2-item measure representing four more speciﬁc
core traits (generalized self-esteem, self-efﬁcacy, neuroticism, and locus of control).

Responses are rated on a scale of 1 (strongly disagree) to 5 (strongly agree; see Appendix

Z).

70

I conducted a conﬁrmatory factor analysis to examine whether the current
measure of core self-evaluations measure can be treated as unidimensional,
similar to existing measures (Judge et al., 2003). Item intercorrelations are
presented in Appendix AA. I conducted a CFA with all twelve core self-
evaluations items loading onto a single latent factor (to represent the common
construct of core self-evaluations which are hypothesized to underlie all items). I
correlated item uniquenesses based upon modiﬁcation indices. The results of the

CFA indicate that model ﬁt is good and that all items load signiﬁcantly onto the

latent factor, thus supporting the use of a single measure of CSE (see Appendix

BB). The reliability for the overall measure is a = .86.

Role-importance. For the current study, I adapted a measure of role-importance
from Amatea, Cross, Clark, and Bobby (1986). Their original Life Role Salience Scale
(LRSS) was used by Kerpelman and Schvaneveldt (1999) and is one of the more
prominent measures of role-importance. Their original scale addresses four major life
roles — the occupation, the marital, the parental, and the homecare roles. However, in the
current study, we are focusing on two major life roles (work and family). Family can be
considered a combination of the marital and parental roles. Therefore, the items were
adapted to reﬂect those two life roles. Furthermore, item content was reﬁned so that
items were parallel across the two life roles.

Response-options were on a 5-point scale ranging from I (strongly disagree) to 5
(strongly agree). The measure was assessed in Time 1 intervention (see Appendix CC for
items). Appendix DD presents the intercorrelations among the items and Appendix EE

presents the results of the principal axis factor analysis with a varimax rotation. The

71

results of the factor analysis showed that all of the family items sorted onto one factor
while the work items sorted onto three different factors. I examined the items in the work
importance scale and it appears that the three factors reﬂect differences in item wording
rather than distinct constructs. Therefore, I conducted a second principal axis factor
analysis with a varimax rotation and forced two factors. The results of this show that the
work and family items do sort onto two separate factors (see Appendix EE). The scale
for the work items does have a slightly lower reliability than desired (a = .68). There was
no item whose removal from the scale would improve the reliability. Therefore, the
original items are all retained in the ﬁnal work role-importance scale. The reliability of
the family role-importance scale was a = .79. However, the family role importance scale
is negatively skewed, indicating that the majority of participants rated family role
importance extremely highly (mean = 4.37, 8d. = .68, skewness statistic = -l .34).
Work-family information gathering. At the end of Time 1, I assessed the
frequency with which participants gathered information on balancing their medical
careers with their family life since entering medical school (called work-family
information gathering Time 2; see items in Appendix FF). A principal axis factor
analysis with a varimax rotation was conducted with the ﬁve items and one factor was
extracted (see Appendix CC for item intercorrelations). The reliability of the ﬁve items
is a = .83. Therefore, a scale was created to represent overall work-family information
gathering behaviors at Time 2 (with higher scores reﬂecting greater infonnation-
gathering). There were no signiﬁcant differences across condition in responses to these

items.

In Time 3 (two weeks following the intervention), I asked a second set of work-
family information gathering questions (called work-family information gathering Time
3; see Appendix HH for items). These questions assessed whether the participants
received and read the email newsletter and visited the website since Time 2.
Additionally, they were asked the same ﬁve questions as asked at Time 2 about
information-gathering, except that these focused on the behaviors that were exhibited in
the past two weeks (i.e., between Time 2 and Time 3). 60 participants agreed to receive
an email newsletter about work-family balance issues facing medical students. However,
only 47 participants reported receiving the email. Of those 47, 42 participants reported
reading some or all of the email newsletter. 22 participants visited the website during the
two-weeks in between Time 2 and Time 3 (these 22 participants also self-reported
visiting the website). Participants self-reported spending on average between 5 — 10
minutes on the website.

A principal axis factor analysis with a varimax rotation was conducted on the ﬁve
information-gathering items that match the ﬁve items assessed in Time 2. The factor
analysis extracted two factors (see Appendix II for item intercorrelations and Appendix J]
for the results of the factor analysis). After reviewing the results of the factor analysis
and examining the inter-item correlations, it appears that Item 9 (time spent using other
resources) is problematic. Additionally, the reliability of the measure is a = .73 with Item
9 and a = .74 without Item 9. A second factor analysis was conducted omitting Item 9
and a one factor solution was extracted. Therefore, the decision was made to create an

overall scale only using Items 5, 6, 7, and 8.

73

Differentiation among self-efficacy scales. There are several scales included in
this study that tap (either directly or indirectly) perceptions of self-efﬁcacy. First, the
core self-evaluations scale includes a generalized self-efﬁcacy subscale (Judge etal.,
2003). Both the career decision self-efﬁcacy scale and the work-family decision-making
self-efﬁcacy scale are designed to tap self-efﬁcacy towards speciﬁc types of decision-
making. Note that generalized self-efﬁcacy (a general feeling that one is capable of
effectively accomplishing goals) is likely to inﬂuence more speciﬁc types of self-
efﬁcacy. Further, work-family role—planning knowledge may also have a self-efﬁcacy
component because a self-efﬁcacy towards making decisions about work and family may
inﬂuence perceptions of knowledge. Therefore, it is important to examine whether these
four constructs should be treated as distinct in the analyses that follow.

All items from the core self-evaluations, CDSE, WFDMSE, and role-planning
knowledge scales (Time 1) were entered into a principal axis factor analysis with a
varimax rotation. This factor analysis failed to converge due to the high number of items
relative to the sample size. Therefore, I conducted a factor analysis with only the CDSE,
WFDMSE, and role-planning items. Eight factors were extracted (see Appendix K for
factor loadings). The results of this factor analysis show that the items from separate
scales do not cross-load onto the same factors. CDSE items loaded onto factors 3, 5, and
7. WFDMSE items loaded onto factors 2, 4, and 6. Role-planning knowledge items
loaded onto factor 1. No items had high loadings on the 8th factor. This suggests that the
WFDMSE, CDSE, and role-planning knowledge scales are distinct and can be

differentiated from one another despite high correlations amongst the scales. In other

74

words, even though the items do share a common self-efﬁcacy factor, there is still a
meaningful distinction between the scales.

Additionally, I conducted a principal axis factor analysis with a varimax rotation
with the WFDMSE, CDSE, and core self-evaluations items (see Appendix K). In this
case, nine factors with eigen-values greater than 1.0 were extracted. No items had their
highest factor loading on the 9th factor. The CDSE items loaded onto factors 1, 7, and 8.
The WFDMSE items loaded into factors 2, 3, and 5. The CSE items loaded onto factors
4 and 6. Again, this suggests that the three scales can be differentiated from one another,
despite high correlations amongst the scales. Based on these ﬁndings, it is appropriate to
treat these four scales as distinct, while recognizing that a common self-efﬁcacy factor
may play a role in inﬂuencing responses to all items.

Manipulation check. Participants were given a manipulation check at the
beginning of Time 2 (immediately after the intervention) to see if they were able to
discern what content was included in the intervention that they watched (Career or Work-
family). Manipulation check items can be found in Appendix KK. Each of the six items
was rated on a scale of I (deﬁnitely not included) to 5 (deﬁnitely included). The items
represented content covered in the two conditions on each of the three major topics (self-
appraisal, gathering information, goal selection). Note that the Career items could also be
seen as having been covered in the Work-family condition, whereas the reverse is not
true. On items 1, 3, and 5, it was expected that there would either be no difference
between conditions or the participants in the Career condition would endorse this item
more strongly (i.e., be more certain that the content was included). On items 2, 4, and 6,

it was expected that the participants in the Work-family condition would endorse this

75

item more strongly (i.e., be more certain that the content was included). Table 6 shows

the mean ratings provided by each group. Results indicate that the differences across the

two conditions were, in fact, detected by the participants. As expected, the Career group

was signiﬁcantly higher than (or equal to) the Work-family group on items 1, 3, and 5.

The Work-family group was signiﬁcantly higher than the Career group on items 2, 4, and

 

 

6.
Table 6
Results of manipulation check
Career Work-family ANOVA
Mean (sd) Mean (sd) (F, p)
Item 1, C SA 4.56 (0.79) 4.13 (0.89) 6.88, .01
Item 2, WF SA 3.00 (1.30) 4.33 (0.80) 41.03, .00
Item 3, C G1 4.80 (0.53) 4.33 (0.80) 12.58, .00
Item 4, WF G1 3.51 (1.25) 4.56 (0.69) 28.84, .00
Item 5, C GS 4.46 (1.86) 4.36 (0.86) 0.40, .53
Item 6, WF GS 3.59 (1.27) 4.24 (0.97) 8.91, .00

 

Notes. Career, C = Career condition. Work-family, WF = Work-family condition. SA =

Self-appraisal. G1 = Gathering information. GS = Goal selection.

RESULTS

Table 7 presents the correlations among the key study variables.

76

Table 7
Correlations among Study Variables

 

 

Variable Mean SD 1 2 3 4 5 6 7
1. Condition — Time 1 0.50 0.50

2. Sex - Time 1 1.69 0.47 .12 -

3. Age — Time 1 2.86 1.02 -.02 -.10

4. Marital - Time 1 0.25 0.44 .02 -.07 .44

5. Relationship — Time 1 1.39 0.49 .08 .01 -.26 -.46

6. School — Time 1 0.49 0.50 .02 -.05 .16 .16 .09

7. Year — Time 1 1.94 0.95 .19 -.03 .29 .47 -. I7 .30

8. Parent — Time 1 0.10 .031 .10 -.17 .25 .51 -.27 .10 .35
9. Ethnic — Time I 0.81 0.40 -.21 .07 -.12 .07 -. I4 -.08 -.ll
10. CSE — Time 1 3.59 0.55 -. l 6 -.09 .07 .07 -.15 .07 .13
ll. Worklmp — Time 1 3.30 0.69 -.10 -.10 -.21 -.23 .16 .07 -.19
12. Famlmp — Time 1 4.37 0.68 -.02 -.01 -. 14 .09 .00 .00 .23
13. CDSE — Time 1 3.72 0.57 -.21 .05 .14 -.01 -.14 .01 .03
I4. CDSE — Time 2 3.82 0.60 —.30 .06 ..17 .06 -.16 -.12 .03
15. CDSE — Time 3 3.83 0.50 —.24 -.08 .16 .10 -.07 .00 .08
16. WFDMSE — Time ‘1 3.50 0.63 -.20 -.01 .14 .00 -.06 -.07 .04
17. WFDMSE — Time 2 3.72 0.66 -.24 .10 .12 .10 -.22 -.15 .05
18. WFDMSE — Time 3 3.68 0.57 -.22 -.06 .14 .09 -.09 -.05 .01
19. WFRPK — Time 1 2.87 0.76 -.11 -.11 .23 .09 -.15 -.03 .12
20. WFRPK — Time 2 3.05 0.72 -.10 -.04 .28 .12 -.20 -.03 .12
21. WFRPK — Time 3 3.16 0.74 -.14 -.02 .24 .09 -.06 -.10 .08
22. AWFC TWIF — Time I 3.40 0.75 .l 1 -.01 .05 -.05 .06 .16 .03
23. AWFC TFIW — Time 1 3.06 0.73 .07 .12 .04 .16 -.10 .05 .01
24. AWFC SWIF — Time 1 2.57 0.88 .17 .13 -.02 .10 -.02 .06 .09
25. AWFC SFIW — Time 1 2.24 .083 .20 .09 -.03 .04 -.01 -.01 .08
26. AWFC BWIF — Time 1 2.43 0.84 .l 1 -.25 -.05 -.05 .07 .09 .04
27. AWFC BFIW — Time 1 2.30 0.74 .03 -.21 -.04 -.04 .09 .06 -.02
28. AWFC TWIF — Time 2 3.12 0.86 .06 -.02 .02 .03 -.04 .09 .07
29. AWFC TFIW ~ Time 2 2.85 0.79 .06 -.01 .12 .09 -.01 -.02 .08
30. AWFC SWIF — Time 2 2.47 0.88 .24 .11 .01 .05 .05 .01 .14
31. AWFC SFIW — Time 2 2.21 0.72 .27 ~01 -.01 -.06 .12 -.01 .04
32. AWFC BWIF — Time 2 2.38 0.78 .18 -.23 .01 -.07 .21 .07 .04
33. AWFC BFIW — Time 2 2.29 0.80 .16 -.22 -.02 -.07 .11 .08 .05
34. AWFC TWIF - Time 3 3.19 0.87 .04 -.08 .07 .14 -.21 .14 .06
35. AWFC TFIW — Time 3 2.96 0.88 -.03 -.06 .03 .29 -. 16 .12 .12
36. AWFC SWIF — Time 3 2.49 0.91 .28 .05 .00 .07 -.08 -.06 .07
37. AWFC SFIW — Time 3 2.26 0.84 .30 -.04 -.06 .08 -. 1.5 .03 .12
38. AWFC BWIF — Time 3 2.40 0.74 .20 -.22 .01 .01 .03 .08 .00
39. AWFC BFIW — Time 3 2.33 0.73 .03 —.33 -.03 -.01 -.02 .12 -.01
40. MedCert — Time 1 3.18 1.10 -.08 -.01 .14 .10 -.20 .06 .26
41. MedCert — Time 2 3.19 1.02 -.07 -.03 .15 .10 -.14 .12 .27
42. MedCert — Time 3 3.19 1.01 -.05 -. 14 .17 .1 l -.19 .02 .21
43. FamCert — Time 1 3.47 1.06 -. 10 .01 .23 .33 -.39 .09 .25
44. FamCert — Time 2 3.46 1.07 -. 15 .04 .20 .34 -.38 .07 .31
45. FamCert — Time 3 3.53 1.11 -.05 .04 .22 .30 -.36 .17 .35
46. Info — Time 1 2.84 0.84 .13 .22 .30 .27 -.24 -.05 .34
47. Info — Time 3 2.13 0.79 .15 .01 -.01 .25 -.15 -.04 .15
48. Receive newsletter 0.56 0.50 .07 .20 .00 .04 -.20 -.05 -.02
49. Visit website 0.20 0.40 .00 .05 .05 .13 -.17 .06 .08

 

Notes. Notes presented on following pages.

77

Table 7, continued

Correlations among Study Variables

 

 

Variable 8 9 10 11 12 13 14 15 16
9. Ethnic — Time 1 -.14

10. CSE — Time 1 .04 .20

ll. Worklmp — Time 1 -.27 —.02 -.05 ,

12. Famlmp — Time I .23 -.05 .05 -.09

13. CDSE — Time 1 .01 .1 l .46 -.13 -.01

14. CDSE — Time 2 .12 .06 .48 -.09 .01 .81.

15. CDSE — Time 3 .08 .09 .38 -.07 -.08 .76 .79

16. WFDMSE — Time 1 .02 .07 .48 -.06 .02 .75 .73 .67

17. WFDMSE — Time 2 .14 .11 .50 -. 14 .07 .70 .83 .74 .78
18. WFDMSE — Time 3 .02 .06 .41 ~. 12 -.07 .69 .72 .86 .73
19. WFRPK — Time 1 .11 -.07 .51 —.04 -. 14 .64 .55 .46 .71
20. WFRPK — Time 2 .19 -. 12 .42 -.04 -. 14 .60 .56 .47 .65
21. WFRPK — Time 3 .10 -.18 .35 -.15 -.09 .53 .48 .48 .55
22. AWFC TWIF - Time 1 .12 -.I4 -.14 .07 .03 -.22 -.21 -.22 ~31
23. AWFC TFIW — Time 1 .19 .19 -.16 -.11 .08 -.39 -.32 -.37 —.42
24. AWFC SWIF — Time 1 -.01 .02 -.50 .05 .11 -.46 —.47 -.44 -.51
25. AWFC SFIW — Time 1 -.02 -.14 -.50 .08 .12 -.44 -.42 -.42 -.47
26. AWFC BWIF — Time 1 .11 -.32 —.41 .01 .18 -.35 -.33 -.26 -.34
27. AWFC BFIW — Time 1 .09 -.30 -.44 —.01 .06 -.35 -.32 -.25 -.35
28. AWFC TWIF -— Time 2 .11 .03 —. 18 .04 .05 —.22 —.28 -.19 -.30
29. AWFC TFIW — Time 2 .28 -.10 -.20 -.11 .19 —.33 —.33 —.31 -.34
30. AWFC SWIF - Time 2 -.02 -.02 -.44 .01 .08 -.40 -.46 -.37 -.47
31. AWFC SFIW —— Time 2 -. 14 -.25 -.44 .15 .00 -.40 -.48 -.37 -.41
32. AWFC BWIF — Time 2 .07 -.34 —.42 .00 .05 -.32 -.37 —.21 -.33
33. AWFC BFIW —— Time 2 .1 1 —.37 -.31 .04 -.02 -.23 —.25 -.16 -.20
34. AWFC TWIF — Time 3 .20 .05 -.02 .08 -.01 -.15 -.20 -.16 -.20
35. AWFC TFIW — Time 3 .28 .18 -.08 -.l l .14 -.32 -.27 -.28 -.34
36. AWFC SWIF — Time 3 .01 .07 -.35 .13 .08) -.35 -.33 -.37 -.42
37. AWFC SFIW — Time 3 .05 —.20 -.40 .21 .08 -.35 -.33 -.39 —.45
38. AWFC BWIF — Time 3 .13 —.17 -.20 .03 .06 -.1 1 -.16 -.11 -.21
39. AWFC BFIW — Time 3 .15 -.29 —.22 .13 .02 —.10 —. 10 -.10 -.'I9
40. MedCert — Time 1 .28 ~05 .25 -.03 .14 .53 .42 .49 .28
41. MedCert - Time 2 .24 -.08 .24 .01 .16 .55 .44 .54 .28
42. MedCert — Time 3 .18 -.17 .27 .01 .09 .54 .45 .51 .38
43. FamCert — Time 1 .37 .03 .33 -.20 .40 .30 .31 .18 .31
44. FamCert 4 Time 2 .39 .00 .29 -.26 .36 .39 .37 .28 .34
45. FamCert — Time 3 .36 -.09 .30 -.30 .34 .31 .31 .29 .24
46. Info — Time 1 .13 .00 .20 -.30 .25 .15 .10 .06 .16
47. Info — Time 3 .21 .01 .17 -.37 .24 -.02 .03 .05 .07
48. Receive newsletter .06 .22 .03 -.1 I -.04 -.03 .08 .02 -.01
49. Visit website .13 .13 .01 -.15 -.04 -.01 .04 .06 -.01

 

Notes. For correlations involving a Time 3 variable, N = 91. For correlations involving between Time land
Time 2 variables (pre and/or post), N = 105 to 108. When N = 91, correlations greater than 1.21! are
signiﬁcant at p = .05 and correlations greater than l.27l are signiﬁcant at p = .01. Marital = Marital status (0
= non—married, l = married). Relationship — Relationship status (0 = not in a relationship. I = in a
relationship). School = Medical school (0 = College of Osteopathic Medicine. 1 = College of Human
Medicine). Year = Year in school. Parent = Parental status (0 = no children in home, I = children in home).
Ethic = Ethnicity (0 = non-white, l = white). CSE = Core self-evaluations. Worklmp = Work importance.
Famlmp = Family importance. CDSE = Career decision self-efﬁcacy.

78

Table 7, continued

Correlations among Study Variables

 

 

Variable 17 18 19 20 21 22 23 24 25
17. WFDMSE — Time 2

18. WFDMSE — Time 3 .77

19. WFRPK - Time 1 .56 .52

20. WFRPK — Time 2 .60 .54 -.85

21. WFRPK — Time 3 .53 .58 -.74 -.79

22. AWFC TWIF — Time 1 -.26 —.28 -.30 -.30 -.19

23. AWFC TFIW — Time 1 -.27 -.28 -.42 -.37 -.26 .33

24. AWFC SWIF — Time 1 -.52 -.52 -.64 -.54 -.47 .34 .36

25. AWFC SFIW — Time 1 -.41 -.39 -.54 -.41 -.33 .09 .36 .72

26. AWFC BWIF — Time 1 -.34 -.34 -.44 -.41 -.40 .13 .09 .38 .45
27. AWFC BFIW — Time I -.33 -.28 -.41 -.34 -.27 .17 .18 .43 .48
28. AWFC TWIF — Time 2 -.35 -.25 —.30 -.37 -.22 .76 .44 .35 .14
29. AWFC TFIW — Time 2 —.32 -.30 -.35 -.38 —.21 .39 .72 .32 .27
30. AWFC SWIF — Time 2 -.51 -.44 -.58 -.55 -.41 .37 .33 .85 .65
31. AWFC SFIW — Time 2 -.51 -.37 -.47 -.42 -.27 .17 .28 .63 .85
32. AWFC BWIF - Time 2 -.40 -.31 -.39 -.41 —.42 .10 .06 .41 .39
33. AWFC BFIW — Time 2 -.31 -.20 -.27 -.26 -.31 .14 .10 .37 .35
34. AWFC TWIF — Time 3 -. 19 -.22 -.21 -.24 -.32 .67 .30 .20 .04
35. AWFC TFIW — Time 3 -.24 -.29 -.35 -.34 -.38 .31 .66 .34 .23
36. AWFC SWIF — Time 3 -.41 -.46 -.47 -.44 -.48 .35 .24 .78 .57
37. AWFC SFIW — Time 3 -.40 -.47 -.40 -.33 -.43 .19 .21 .61 .72
38. AWFC BWIF — Time 3 —. 1 7 -.27 -.24 -.27 -.44 .14 .08 .30 .20
39. AWFC BFIW — Time 3 -.16 -.21 —.21 -.22 —.37 .15 .05 .29 .21
40. MedCert— Time 1 .37 .38 .31 .29 .27 -.15 -.15 -.26 -.16
41. MedCert — Time 2 .37 .41 .32 .31 .32 -.10 -.16 -.29 -.21
42. MedCert — Time 3 .45 .46 .39 .39 .41 —.24 -.24 -.34 -.27
43. FamCert — Time 1 .38 .23 .32 .37 .32 -.05 .00 -.18 -.18
44. FamCert — Time 2 .39 .29 .36 .36 .35 —.09 -.09 -.20 -.24
45. FamCert — Time 3 .35 .33 .27 .30 .35 -.()7 -.04 -.21 -. l 7
46. Info — Time 1 .16 .09 .14 .11 .14 .12 .07 -.03 -.10
47. Info — Time 3 .08 .16 .06 -.01 .07 -.08 .00 -.02 -.10
48. Receive newsletter .16 .l l -.03 -.07 -.03 .10 .08 -.08 -.02
49. Visit website .08 .04 -.08 -.09 -.05 .05 .10 -.06 -.09

Notes, continued. WFDMSE = Work—family decision-making self-efﬁcacy. WFRPK = Work-family role-

planning knowledge. AWFC = Anticipated work-family conﬂict (T = time-based, S = strain-based, B =

behavior-based, WIF = work-interference-with-family, FIW = family—interference-with-work). Medcert =
Certainty regarding medical specialty plans. Famcert = Certainty regarding family plans. Info =

Information gathering behaviors.

 

79

Table 7, continued

Correlations among Sturb» Variables

 

 

 

 

 

Variable 26 27 28 29 30 31 32 33 34
27. AWFC BFIW — Time 1 .80

28. AWFC TWIF — Time 2 .13 .18

29. AWFC TFIW - Time 2 .22 .28 .56 -

30. AWFC SWIF — Time 2 .33 .37 .47 .37

31. AWFC SFIW — Time 2 .44 .43 .26 .32 .70

32. AWFC BWIF — Time 2 .75 .75 .22 .28 .44 .47

33. AWFC BFIW — Time 2 .70 .79 .27 .28 .42 .47 .88

34. AWFC TWIF — Time 3 .1 l .08 .69 .33 .20 .11. .05 .13

35. AWFC TFIW — Time 3 .17 .20 .41 .63 .27 .20 .14 .19 .52
36. AWFC SWIF — Time 3 .30 .28 .33 .18 .77 .58 .34 .31 .30
37. AWFC SFIW — Time 3 .45 .38 .16 .18 .53 .69 .37 .36 .24
38. AWFC BWIF — Time 3 .54 .41 .16 .15 .26 .23 .68 .59 .23
39. AWFC BFIW — Time 3 .55 .48 .16 .13 .22 .27 .65 .68 .21
40. MedCert — Time 1 -.06 -.08 -.09 -.05 -.22 -.20 -. l6 -. 10 -.04
41. MedCert — Time 2 -.08 -. 12 -.10 -.06 -.25 -.22 -. 18 -. l 3 -.03
42. MedCert — Time 3 -.l l -.06 -.18 -.09 -.28 -.22 -.10 -.05 -.22
43. FamCert — Time 1 -.21 -.22 -.05 .05 -.22 —.27 -.28 -.19 -.04
44. FamCert — Time 2 -.20 -.20 -. 12 .01 -.25 -.35 -.24 -. l 7 -.09
45. FamCert — Time 3 -. l 2 -. 12 -.08 .01 -. l 8 -.26 —.12 -.05 -.11
46. Info — Time 1 -.11 -. l 7 .14 .10 .04 -.07 -.14 -.14 .08
47. Info - Time 3 .08 .03 .02 .01 .00 -.16 -.01 .05 .05
48. Receive newsletter -.09 -.05 .05 -.07 -.09 -.18 -.18 -.18 .10
49. Visit website .03 -.01 .03 -.04 -.12 -.14 -.05 -.07 .16
Table 7, continued

Correlations among Study Variables

Variable 35 36 37 38 39 40 41 42 43
36. AWFC SWIF — Time 3 .34

37. AWFC SFIW — Time 3 .34 .72

38. AWFC BWIF — Time 3 .29 .41 .44

39. AWFC BFIW — Time 3 .23 .35 .44 .80

40. MedCert — Time I -.18 -. l 8 -.06 -.10 -.03

41. MedCert — Time 2 -.18 -.21 -.12 -.05 -.03 .93

42. MedCert — Time 3 -.26 -.31 -.20 -.15 -.12 .84 .85

43. FamCert — Time 1 -.05 -. l 6 -.23 -. 15 -.12 .25 .32 .23

44. FamCert -— Time 2 -.()8 -.24 -.28 -.11 -.07 .31 .39 .30 .88
45. FamCert — Time 3 -.07 -.25 -.28 -.10 -.ll .28 .36 .32 .84
46. Info — Time 1 .00 -.02 -. 15 -.21 -.26 .12 .09 .13 .34
47. Info — Time 3 .03 .03 -.07 -.04 -.10 .05 .02 .02 .10
48. Receive newsletter -.02 —.01 -.07 —.12 -.28 -.15 -. 15 -.17 .06
49. Visit website .06 -.07 -.06 .1 1 -.07 -.04 -.03 -.()5 .04

 

80

Table 7, continued
Correlations among Study Variables

 

 

Variable 44 45 46 47 48
45. FamCert - Time 3 .86

46. Info - Time 1 .30 .29

47. Info - Time 3 .18 .1. 3 .37 ~

48. Receive newsletter .06 .07 .24 .23

49. Visit website .04 .05 .12 .18 .45

 

Predictors of A nticipated Work-family Conflict

In the following section, I examine a set of potential predictors of anticipated
work-family conﬂict. These relationships are all examined using data collected at Time
1.

Demographics. Hypotheses 1 through 8 address the relationship between
demographics (i.e., age, year in school, gender, marital status, parental status, relationship
status) and anticipated work-family conﬂict. Out of 36 correlations, only two were
signiﬁcant. Women reported anticipating signiﬁcantly less behavior-based F IW (r = -.21,
p = .03). This relationship is in the expected direction. Thus, Hypothesis 3c is supported,
but not Hypothesis 3a or 3b. However, women also reported anticipating signiﬁcantly
less behavior-based WIF (r = -.25, p = .01). However, this relationship is in the opposite
direction than what was expected. Therefore, Hypothesis 4 is not supported. Overall,
Hypotheses 1 through 8 (except for 3c) were rejected showing that demographics were
largely uncorrelated with any of the six types of AWFC. However, women did anticipate
less behavior-based conﬂict in both directions.

Core self-evaluations. Hypothesis 9 predicted that individuals with higher core
self-evaluations would anticipate less work-family conﬂict. At Time 1, core self-
evaluations were signiﬁcantly correlated with anticipated strain-based conﬂict in both

directions and anticipated behavior-based conﬂict in both directions, but not with

81

anticipated time-based conﬂict in either direction. Participants with higher core self-
evaluations anticipated less strain-based WIF (r = -.50, p = .00), less strain-based FIW (r
= -.50, p = .00), less behavior—based WIF (r = -.41, p = .00) and less behavior-based F IW
(r = -.44, p = .00). Therefore, Hypothesis 9a and 9b (time-based conﬂict) are not
supported whereas Hypotheses 9c, 9d, 9e, and 9f (strain- and behavior-based conﬂict) are
supported.

Role importance. Hypotheses 10 through 13 examine the relationship between
anticipated work-family conﬂict and role importance (in both the family and work roles).
I expected that individuals with greater work role importance would anticipate greater
work-interference-with-fami1y and less fami1y-interference-with-work. Relationships in
the opposite direction were predicted for greater family role importance. Of the twelve
relationships examined, none were signiﬁcant. Work and family role importance were
not associated with any of the six types of AWFC. In other words, Hypotheses 10
through 13 were not supported.

Role-planning knowledge. Hypothesis 14 predicted that greater role—planning
knowledge would be associated with less AWFC. At Time 1, individuals with greater
role-planning knowledge did anticipate less time-based WIF (r = -.30, p = .00), less time-
based F IW (r = -.42, p = .00), less strain-based WIF (r = -.64, p = .00), less strain-based
F 1W (r = -.54, p = .00), less behavior-based WIF (r = —.44, p = .00), and less behavior-
based F IW (r = -.41, p = .00). Table 8 shows the results of a hierarchical regression
examining the role-planning knowledge as a predictor of each of the six types of AWFC
at Time 1, when controlling for core self-evaluations. Controlling for core self-

evaluations allows us to examine the incremental predictive validity of role-planning

82

knowledge while acknowledging (and controlling for) the effects of more generalized
self-efﬁcacy (as captured by the core self-evaluations scale). Note that demographics are
not controlled for because they are largely uncorrelated with AWFC. The results of these
hierarchical regressions indicate that role-planning knowledge explains a signiﬁcant
amount of incremental variance in each of the six types of AWFC, when controlling for
core self-evaluations. Furthermore, note in Table 2 that role-planning knowledge at each
administration (Time 1, Time 2, Time 3) is negatively correlated with almost all forms of

AWFC at each administration. Given this information, Hypothesis 14 is supported.

Table 8

Hierarchical Regressions for H I 4 .' Role-planning Knowledge Predicting
Anticipated Work-Family Conflict (N = 108)

 

R2 A R2 p

 

DV: Time 1 time-based WIF

Step 1 — Core Self-evaluations .02 .02 .14

Step 2 — Time 1 Role-planning Knowledge .09 .07 .01
DV: Time 1 time-based FIW

Step 1 — Core Self-evaluations .02 .02 .11

Step 2 — Time 1 Role-planning Knowledge .18 .16 .00
DV: Time 1 strain-based WIF

Step 1 — Core Self-evaluations .25 .25 .00

Step 2 - Time 1 Role—planning Knowledge .45 .20 .00
DV: Time 1 strain-based FIW

Step 1 — Core Self—evaluations .25 .25 .00

Step 2 — Time 1 Role-planning Knowledge .36 .11 .00
DV: Time 1 behavior-based WIF

Step 1 — Core Self-evaluations .17 .17 .00

Step 2 — Time 1 Role-planning Knowledge .24 .07 .00
DV: Time 1 behavior-based WIF

Step 1 — Core Self-evaluations .20 .20 .00

Step 2— Time 1 Role-planning Knowledge .24 0.4 .02

 

Notes. DV= Dependent variable WIF: Work- interference- with- family. FIW=
Family-interference- with- work.

Supplemental analyses were conducted to examine the relationship over time

between role-planning knowledge and the six types of AWFC. The ﬁgure representing

83

lllllllli ll i

[.1 lllllf:|l l... ill.

this analysis is presented in Figure 11. Note that this is a path analysis and that no latent
variables are modeled due to the limited sample size. Table 9 shows that while role-
planning knowledge and AWF C are signiﬁcantly negatively correlated at Time 1 (for all
six types of AWFC), the paths from role-planning. knowledge to AWF C are not
signiﬁcant over time. In other words, there is no evidence that role-planning knowledge

is causing AWF C over time (given the time frame examined here).

 

 

 

T1 RPK _ T2 RPK T3 RPK

 

 

V

 

 

 

 

 

 

 

 

 

 

 

 

Tl AWFC T2 AWFC T3 AWFC

 

 

 

 

 

 

 

 

 

 

 

Figure 1 1. Path Analysis: Relationships Between Anticipated Work-family
Conﬂict and Work-family Role-planning Knowledge Over Time

Notes. T1= Time 1. T2 = Time 2. T3 = Time 3. RPK = Role-planning knowledge.
AWF C = Anticipated work-family conﬂict. Each of the six types of AWF C was

84

Table 9

Path Analysis for H14: Auto-correlated Cross—lagged Relationships Between Anticipated

Work-family Conﬂict and Work-family Role-planninLKnowledge Over Time

 

 

 

 

 

 

 

Type of AWFC N 36 Df P CFI . RMSEA COV.atT1 SE (p)
Time-based 91 27.26 6 .00 .94 .20 -.17 (.01)
WIF
TlRPK 9 T2TWIF -.03 (.64)
TITWIF 9 T2RPK -.02 (.71)
T2RPK 9 T3TWIF -.04 (.65)
T2TWIF9 T3RPK .02 (.82)
Time-based 91 29.39 6 .00 .94 .21 —.24 (.00)
FIW
TlRPK 9 T2TFIW .00 (.98)
TlTFIW 9 T2RPK -.03 (.65)
T2RPK 9 T3TFIW -.15 (.07)
T2TFIW9 T3RPK .06 (.39)
Strain-based 91 36.56 6 .00 .94 .24 -.45 (.00)
WIF
TlRPK 9 TZSWIF -.02 (.63)
TlSWIF 9 T2RPK .03 (.63)
T2RPK 9 T3SWIF —.05 (.50)
T2SWIF 9 T3RPK .00 L96L
Strain-based 91 33.07 6 .00 .94 .22 -.35 (.00)
FIW ,
TlRPK 9 TZSFIW .06 (.35)
TISFIW 9 T2RPK .08 (.20)
T2RPK 9 T3SFIW -.10 (.22)
TZSFIW 9 T3RPK .01 (.94)
Behavior-based 91 14.20 6 .03 .98 .12 -.28 (.00)
WIF
TlRPK 9. T2BWIF -.03 (.66)
TIBWIF 9 T2RPK -.03 (.59)
T2RPK 9 T3BWIF -.04 (.66)
TZBWIF 9 T3RPK -.17 (.01)
Behavior-based 91 16.25 6 .01 .97 .14 -.21 (.00)
FIW
Tl RPK 9 T2BFIW .09 (.21)
TIBFIW 9 T2RPK .02 (.77)
T2RPK 9 T3BFIW -.10 (.18)
TZBFIW 9 T3RPK -.17 (.01)

 

Notes. COV. = Covariance at T1. SE = Standardized estimate (direction of arrows
reﬂects direction ofthe path). T1 = Time 1. T2 = Time 2. T3 = Time 3. RPK = Role-
planning knowledge. T = time-based, S = strain-based, B = behavior-based, WIF =
work-interference—with-family, F [W = family-interference-with-work.

85

Work-family decision-making self-efﬁcacy. Hypothesis 15 addresses the
relationship of anticipated work-family conﬂict to WFDMSE. At Time 1, individuals
with greater WFDMSE did anticipate less time-based WIF (r = -.3 1, p = .00), less time-
based F IW (r = -.42, p = .00), less strain-based WIF (r = -.51, p = .00), less strain-based
F [W (r = -.47, p = .00), less behavior-based WIF (r = -.34, p = .00), and less behavior-
based F 1W (r = -.35, p = .00). Therefore, Hypothesis 15 is supported. Table 10 shows
the results of a hierarchical regression examining WFDMSE as a predictor of each of the
six types of AWFC at Time 1 when controlling for core self-evaluations (to control for
the effects of more generalized self-efﬁcacy). Note that demographics are not controlled
for because they are largely uncorrelated with AWFC. Results indicate that WFDMSE
explains a signiﬁcant amount of incremental variance in four of the six types of AWF C

(time-based WIF, time-based FIW, strain-based WIF, and strain-based FIW).

86

Table 10

Hierarchical Regressions for H15: Work-family Decision-making Self-efficacy
Predicting Anticipated Work-Family Conﬂict (N = 108)

 

 

R2 A R2 p

DV: Time 1 time-based WIF

Step 1 - Core Self-evaluations .02 .02 .14

Step 2 — Time 1 Work-family Decision-making Self-efﬁcacy .10 .08 .00
DV: Time 1 time-based FIW

Step 1 — Core Self-evaluations .02 .02 .11

Step 2 — Time 1 Work-family Decision-making Self-efﬁcacy .18 .15 .00
DV: Time 1 strain-based WIF

Step 1 — Core Self-evaluations .25 .25 .00

Step 2 - Time 1 Work-family Decision-making Self-efﬁcacy .34 .09 .00
DV: Time 1 strain-based F IW

Step 1 — Core Self-evaluations .25 .25 .00

Step 2 — Time 1 Work-family Decision-making Self-efﬁcacy .32 .07 .00
DV: Time 1 behavior-based WIF

Step 1 —~ Core Self-evaluations .17 .17 .00

Step 2 — Time 1 Work-family Decision-making Self-efﬁcacy .19 .03 .07
DV: Time 1 behavior-based WIF

Step 1 — Core Self—evaluations .20 .20 .00

Step 2 — Time 1 Work-family Decision-making Self-efﬁcacy .22 .02 .07

 

Notes. DV = Dependent variable. WIF = Work-interference-with-fami1y. F IW =
F amily-interference-with-work.

Supplemental analyses were conducted to examine the relationship over time
between WFDMSE and the six types of AWFC. The ﬁgure representing this analysis is
presented in Figure 12. Note that this is a path analysis and that no latent variables are
modeled due to the limited sample size. Table 11 shows that while WFDMSE and
AWFC are signiﬁcantly negatively correlated at Time 1 (for all six types of AWF C), the
paths from WFDMSE to AWFC are not signiﬁcant over time. In other words, there is no
evidence that WFDMSE is causing AWF C over time (given the time frame examined
here). Note that for anticipated behavior-based WIF and F IW, there is signiﬁcant

relationship between Time 2 AWFC and Time 3 WFDMSE. A causal interpretation

87

would suggest that increases in anticipated behavior-based conﬂict are causing decreases
in WFDMSE over time. In sum, while there is evidence that WFDMSE is negatively
correlated with all six forms of AWFC (thus supporting Hypothesis 15), there is no

evidence that WFDMSE is causing AWF C over time.

 

 

 

T1 WFDMSE A T2 WFDMSE T3 WFDMSE

 

 

V
1

 

 

 

 

 

 

 

 

 

 

 

 

T1 AWF C T2 AWFC T3 AWFC

 

 

V

 

 

 

 

 

 

 

 

 

Figure 12. Path Analysis: Relationships Between Anticipated Work-family
Conﬂict and Work-family Decision-making Self-efﬁcacy Over Time

Notes. Tl= Time 1. T2 = Time 2. T3 = Time 3. WFDMSE = Work—family
decision-making self-efﬁcacy. AWF C = Anticipated work-family conﬂict. Each of
the six types of AWF C was examined in separate path analyses.

88

Table 11

Path Analysis for H15: Auto-correlated Cross-lagged Relationships Between Anticipated
Work-family Conflict and Work-family Decision-making Self-eﬂicacy Over Time

Type of AWFC N 12 Df P CFI RMSEA COV. at T1 SE (p)

 

 

 

 

 

Time-based 91 27.26 6 .00 .94 .20 -.28 (.01)

WIF .
TIWFDMSE 9 T2TWIF -.03 (.64)
TlTWIF 9 T2WFDMSE -.02 (.71)
T2WFDMSE 9 T3TWIF -.04 (.65)
T2TWIF9 T3WFDMSE .02 (.82)

Time-based 91 29.39 6 .00 .94 .21 -.24 (.00)

FIW
TIWFDMSE 9 T2TFIW .00 (.98)
TITFIW 9 T2WFDMSE -.03 (.65)
T2WFDMSE 9 T3TFIW -.15 (.07)
T2TF 1W9 T3WFDMSE .06 (.39)

Strain-based 91 36.56 6 .00 .94 .24 -.45 (.00)

WIF
TIWFDMSE 9 TZSWIF -.02 (.81)
TISWIF 9 T2WFDMSE .03 (.63)
T2WFDMSE 9 T3SWIF -.05 (.50)
T2SW1F 9 T3WFDMSE .00 (.16)

Strain-based 91 33.07 6 .00 .94 .22 -.35 (.00)

FIW
TIWFDMSE 9 TZSFIW .06 (.35)
TlSFIW 9 T2WFDMSE .08 (.20)
T2WFDMSE 9 T3SFIW -.10 (.22)
TZSFIW 9 T3WFDMSE .01 (.94)

Behavior-based 91 14.20 6 .03 .98 .12 -.28 (.00)

WIF
TIWFDMSE 9 TZBWIF -.03 (.66)
TIBWIF 9 T2WFDMSE -.03 (.59)
T2WFDMSE 9 T3BWIF -.04 (.66)
TZBWIF 9 T3WFDMSE -.17 (.OIL

Behavior-based 91 16.25 6 .01 .97 .14 -.21 (.00)

FIW
TIWFDMSE 9 TZBFIW .09 (.21)
TIBFIW 9 T2WFDMSE .02 (.77)
T2WFDMSE 9 T3BFIW -.10 (.18)
T2BFIW 9 T3WFDMSE -.17 (.01)

 

Notes. COV. = Covariance at T1. SE = Standardized estimate (direction of arrows
reﬂects direction of the path). T1 = Time 1. T2 = Time 2. T3 = Time 3. WFDMSE =
Work-family decision-making self-efﬁcacy. T = time-based, S = strain-based, B =
behavior-based, WIF = work-interference-with-family, F [W = family-interference-with-
work.

89

Effects of the Intervention on Career and Work-family Decision-making Self-efficacy

Hypothesis 16 predicted that individuals in the Work-family condition would
have higher WFDMSE than those in the Career condition following the intervention.
Table 12 shows the mean levels of WFDMSE at each time point for both conditions.
Although participants were randomly assigned to conditions, the results of a one-way
ANOVA indicates that initial WFDMSE (at Time 1) differed signiﬁcantly between the
two groups, with the Career condition having higher initial levels than the Work-family
condition (F = 4.51, p = .04). Between-condition differences in WFDMSE at Time 2 and
Time 3 were examined using an ANCOVA controlling for Time 1 differences between
conditions in WFDMSE. Results indicated that there were no signiﬁcant between-
condition differences at Time 2 (F = 1.13, p = .29) or at Time 3 (F = 2.75, p = .10).
Additionally, I compared the change scores over time in WFDMSE to examine whether
condition predicted degree of change. The results of a one-way ANOVA show no
differences across conditions in the change in WFDMSE from Time 1 to Time 2 (F = .56,
p = .46). Similarly, the results of a one-way ANOVA show no differences across
conditions in the change in WFDMSE from Time 1 to Time 3 (F = .65, p = .42).

Based on these analyses, Hypothesis 16 is not supported. In other words, the
intervention did not have the intended effect of increasing WFDMSE in the Work-family

condition more so than in the Career condition.

90

Table 12
Descriptive Statistics for H15 : WFDMSE by Condition

 

 

 

 

 

N Mean SD

Career condition .

WFDMSE Time 1 54 3.63 0.60
WFDMSE Time 2 54 3.87a 0.57
WFDMSE Time 3 47 3.81b 0.60
Work-family condition

WFDMSE Time 1 54 3.38 0.64
WFDMSE Time 2 54 3.568‘ 0.71
WFDMSE Time 3 44 3.55b 0.52

 

Notes. WFDMSE = Work-family decision-making self-efﬁcacy. aWhen controlling for
Time 1 levels of WFDMSE, these means are not signiﬁcantly different from one another.
b When controlling for Time 1 levels of WFDMSE, these means are not signiﬁcantly
different from one another.

It is also interesting to examine the pattern of change over time within each
condition. For the Career condition, there was a signiﬁcant increase in WFDMSE from
Time 1 to Time 2, as examined by a paired-samples t-test (t = 3.69, p = .00). From Time
2 to Time 3, there was no signiﬁcant change in the Career condition (t = 1.12, p = .27).
From Time 1 to Time 3, the change was signiﬁcant (t = -2.93, p = .01). In other words,
in the Career condition, WFDMSE increased and this increase persisted over time.

In the Work-family condition, there was also a signiﬁcant increase in WFDMSE
from Time 1 to Time 2 (t = -3.49, p = .00). From Time 2 to Time 3, there was no
signiﬁcant change (t = 1.42, p = .16). From Time 1 to Time 3, the change was signiﬁcant
(t = -2.22, p = .03). In other words, in the Work-family condition, WFDMSE increased

and this increase persisted over time. Thus, it appears that both interventions (both

Career and Work-family) boosted participant WFDMSE and that these increases

91

persisted over the two-week lag. No differential effectiveness of one intervention over

the other is discernable.

 

 

WFDMSE Over Time

 

+ Career

——I— Work-family

 

 

WFDMSE

 

 

Time

 

 

 

 

 

 

Figure 13. Work-family Decision-making Self-efﬁcacy Over Time
Notes. WFDMSE = Work-family decision-making self-efﬁcacy

Hypothesis 17 examined the inﬂuence of the interventions on CDSE. Table 13
shows the mean levels of CDSE at each time point for both conditions and Figure 14
shows a graph of CDSE at each time point for both conditions. Similar to the previous
hypothesis, it is important to control for pre-existing differences across groups when
comparing the impact of the intervention on CDSE. In this case, it is preferable to
control for pre-existing differences in Time 1 WFDMSE when looking at differences in
CDSE at Time 2 and Time 3. An ANCOVA was conducted examining differences
across conditions in Time 2 CDSE controlling for Time 1 WFDMSE. Results indicated a

signiﬁcant difference across groups (F = 5.97, p = .02). The Career condition had greater

CDSE at Time 2 than the Work-family condition, when controlling for Time 1
WFDMSE. A similar analysis was conducted to compare Time 3 CDSE across
conditions controlling for Time 1 WFDMSE. In this case, there was no signiﬁcant
difference across conditions (F = 3.49, p = .07). In other words, the raised CDSE in the
Career condition (as opposed to the Work-family condition) does not persist over time.

Additionally, I compared the change scores over time in CDSE to examine
whether condition predicted degree of change. The results of a one-way AN OVA show
no differences across conditions in the change in CDSE from Time 1 to Time 2 (F = 3.47,
p = .07). Similarly, the results of a one-way ANOVA show no differences across

conditions in the change in WFDMSE from Time 1 to Time 3 (F = .29, p = .59).

 

 

 

 

 

Table 13
Descriptive Statisticsfor H I 7: CDSE by Condition

N Mean SD
Career condition
CDSE Time 1 54 3.84 0.54
CDSE Time 2 54 4.013 0.48
CDSE Time 3 47 3.95c 0.49
Work-family condition
CDSE Time 1 54 3.60 0.57
CDSE Time 2 54 3.64b 0.65
CDSE Time 3 44 3.71c 0.50

 

Notes. CDSE = Career decision self-efﬁcacy. a’b When controlling for Time 1 WFDMSE,
Time 2 CDSE in the Career condition is signiﬁcantly greater than Time 2 CDSE in the
Work-family condition. c When controlling for Time 1 WFDMSE, these two means are
not signiﬁcantly different from one another.

The pattern of change in CDSE over time can also be examined. In the Career

condition, there was a signiﬁcant increase in CDSE from Time 1 to Time 2, as assessed

by a paired samples t-test (t = -3.24, p =.00). From Time 1 to Time 3, the change was not

93

signiﬁcant (t = -1.97, p = .06). In other words, in the Career condition, CDSE increased
but this change diminished over time.

In the Work-family condition, there was no change in CDSE from Time 1 to Time
2 (t = -0.84, p = .41) or from Time 1 to Time 3 (t = -1.57, p = .12). In other words, the
Work-family condition had no signiﬁcant inﬂuence on CDSE. Overall, Hypothesis 17
was partially supported in that participants in the Career condition had higher CDSE after

the intervention, but this difference did not persist over time.

 

CDMSE Over Time

 

 

 

 

 

 

LU
as) + Career
8 A,,,--’/’ .--...» Work-family
1 2 3
Time

 

 

 

Figure 14. Career Decision Self-efﬁcacy Over Time
Notes. CDSE = Career decision self-efﬁcacy

Effects ofthe Intervention on Anticipated Work-family Conﬂict

Hypothesis 18 focused on the differential impact of the two conditions on levels
of AWF C. Table 14 presents the mean scores for the six types of AWF C at each time
point for both conditions. A MANOVA was conducted to examine the differences across

conditions in the six types of AWFC at Time 1. The overall MANOVA was. not

94

signiﬁcant (F = 1.29, p = .26). However, results of an ANOVA indicate a pre-existing
difference (i.e., differences at Time 1) in strain-based FIW (F = 4.41, p = .04).

Based on these ﬁndings, a MANCOVA was conducted to examine the differences
between conditions in AWF C at Time 2 controlling for Time 1 strain-based F IW. The
overall MANCOVA was not signiﬁcant (F = 1.01, p = .43). However, the ANCOVAs
examining the effects of condition on each type of AWFC did indicate a signiﬁcant
difference between conditions in strain-based WIF and strain-based F [W at Time 2 (even
when controlling for Time 1 strain—based FIW; F = 4.78, p = .04). In other words,
individuals in the Work-family condition anticipated higher strain-based conﬂict in both
directions after the intervention than those in the Career condition (even when controlling
for Time 1 differences in this construct).

A second MANCOVA was conducted to examine the differences between
conditions in AWFC at Time 3 (again controlling for Time 1 differences in strain-based
F IW). The MANCOVA was signiﬁcant (F = 3.30, p = .01). DANCOVAs examining the
differences between conditions in each of the six types of AWF C indicate signiﬁcant
differences between conditions in strain-based WIF (F = 5.06, p = .03) and strain-based
F IW (F = 7.21, p = .01). Results indicate that the participants in the Work-family
condition were higher in both types of AWF C than the participants in the Career
condition, even when controlling for pre-existing differences across groups.

I also examined the patterns of change over time in each of the six types of
AWFC over time. Figures 15 through 20 show levels of AWFC graphed across the three
time points for each condition (for each type of AWF C). Table 14 presents t-tests of the

signiﬁcance in change over time. Participants in both conditions showed a decrease in

time-based WIF from Time 1 to Time 2, but the magnitude of change in both groups
decreased by Time 3. Similarly, participants in both conditions showed a decrease in
time-based F IW from Time 1 to Time 2, but this change did not persist over time in either
condition. Participants in the Career condition showed a decrease in strain-based WIF
that persisted over the two week time-lag, whereas participants in the Work-family
condition showed no change. There was no change in strain-based FIW, behavior-based
WIF, or behavior-based F IW for either condition.

Hypothesis 18, which predicted that participants in the Work—family condition
would have lower levels of AWF C after the intervention than those in the Career
condition, was not supported. Participants in both conditions showed lower levels of
time-based WIF and time-based FIW after the intervention, although these changes did
not persist over time. Only participants in the Career condition showed a decrease in

strain-based WIF.

96

Table 14

Descriptive Statistics and One-way ANO VAs for H18: A WF C Across Conditions

 

 

 

 

 

 

 

C C C WF WF WF F p
Mean SD N Mean SD N

Time 1

Time-based WIF 3.31 0.76 54 3.48 0.71 54 1.37 0.24
Time-based F 1W 3.01 0.74 54 3.12 0.72 54 0.59 0.44
Strain-based WIF 2.42 0.76 54 2.73 0.98 54 3.29 0.07
Strain-based F [W 2.07 0.69 54 2.40 0.92 54 4.41 0.04
Behavior-based WIF 2.34 0.69 54 2.52 0.97 54 1.27 0.26
Behavior-based FIW 2.28 0.69 54 2.33 0.79 54 0.09 0.76
Time 2

Time-based WIF 3.07 0.87 52 3.16 0.86 53 0.32 0.57
Time-based FIW 2.80 0.74 52 2.89 0.84 53 0.35 0.55
Strain-based WIF 2.27 0.77 52 2.68 0.93 53 6.13 0.01
Strain-based F [W 2.02 0.57 52 2.40 0.80 53 8.03 0.01
Behavior-based WIF 2.23 0.67 52 2.52 0.85 53 3.51 0.06
Behavior-based FIW 2.17 0.72 52 2.42 0.87 53 2.54 0.11
Time 3

Time-based WIF 3.16 0.90 47 3.23 0.84 44 0.17 0.68
Time-based FIW 2.99 0.86 47 2.94 0.91 44 0.06 0.80
Strain-based WIF 2.25 0.80 47 2.75 0.95 44 7.45 0.01
Strain-based F [W 2.01 0.68 47 2.52 0.92 44 8.84 0.00
Behavior-based WIF 2.26 0.53 47 2.55 0.90 44 3.75 0.06
Behavior-based FIW 2.31 0.60 47 2.35 0.85 44 0.06 0.81

 

Notes. C = Career condition. WF = Work-family condition. WIF = Work-interference-
with-family. FIW = F amily-interference-with-work.

Table 15

Paired-samples T—testsfor Differences in Anticipated Work-family Conﬂict

Over Time by Condition

 

Career condition

Work-family condition

 

 

T19T2 T19T3 T19T2 T19T3
(hp) (tie) (up) (up)
Time-based WIF 3.26 (.00) 1.92 (.06) 3.92 (.00) 2.04 (.05)
Time-based FIW 3.54 (.00) 0.65 (.52) 2.82 (.01) 1.49 (.14)
Strain-based WIF 3.02 (.00) 2.35 (.02) 0.61(.54) -0.51 (.61)
Strain-based FIW 1.37 (.17) 0.92 (.36) 0.15 (.88) -1 .67 (.10)
Behavior-based WIF 1.31 (.20) 1.26 (.22) -0.19 (.85) -0.76 (.45)
Behavior-based FIW 2.02 (.05) 0.23 (.78) -1.64 (.11) -0.87 (.39)

 

Notes. T = Time. WIF = Work—interference-with-family. F 1W = Family-

interference-with-work.

97

 

 

Time-based WIF Over Time

 

+ Career

 

 

—I— Work-family

 

Time-based WIF

 

1 2 3

Time

 

 

Figure 15. Time-based Work-interference-with-family Over Time
Notes. WIF = Work-interference-with-fami1y

 

 

 

 

 

 

 

 

 

 

 

 

 

Time-based FIW Over Time
3.2 4
a 3" "
E 3 9 \x
8 \\\ .4». +Career
g 2.9 ‘r-r‘” .
.9 V L... Work-family
d)
E 2.8 v
I‘ 2.7
2.6 i i .
1 2 3
Time

 

 

Figure 16. Time-based F amily-interference-with-work over Time
Notes. FIW = Family-interference-with-work

98

 

 

 

 

 

 

 

 

 

 

 

 

 

Strain-based WIF Over Time
3 _
W
e 2 5 kN—H
g 2
1:
g +Career
re 15
9 ~I-Work-fami|y
.5 1
E
w 0.5
0 I I I
1 2 3
Time

 

 

 

Figure 17. Strain-based Work-interference-with-family Over Time
Notes. WIF = Work-interference-with-family

 

Strain-based FIW Over Time

 

 

 

 

 

 

 

 

 

 

 

3 ..
E 2 5 I“ -4..-_____-___ i“ -...--WJ
IL A A
1: 2 v v
iii + Career
in 1 5 -
£3 ' Work-family
.E 1
I?
‘0 0.5
0 I 1 .
1 2 3

Time

 

 

Figure 18. Strain-based Family-interference-with-work Over Time
Notes. F IW = F amily-interference-with-work

99

 

 

 

 

 

 

 

 

 

 

 

 

 

Behavior-based WIF Over Time
2.6 —

I ---- -l~'//J
LL. 2.5
E:
g 2.4
"’ +Car r
a? 2.3 \ ee _
3 v —!- Work-family
'5 2.2
(U
8
m 2.1

2 1 1 .
1 2 3
Time

 

 

Figure 19. Behavior-based Work-interference-with-family Over Time
Notes. WIF = Work-interference-with-family

 

 

Behavior-based FIW Over Time

2.45

2.35
2.3
2.25
2.2
2.15
2.1
2.05

 

—+-— Career

a. a. Work-family

 

 

 

Behavior-based FIW

 

Time

 

 

Figure 20. Behavior-based F amily-interference-with-work Over Time
Notes. F IW = F amily-interference-with-work

100

Factors Influencing Degree of Change in WFDMSE

Hypotheses 19 through 24 examine the extent to which different variables
(gender, age, marital status, relationship status, family role importance, and core self-
evaluations) inﬂuence the extent of change in WFDMSE over time for participants in the
Work-family condition. Time 2 WFDMSE was treated as the dependent variable. Time
1 WFDMSE was entered as an independent variable in the ﬁrst step of the regression.
The variable of interest was entered in the second step of the regression. Further,
supplemental analyses tested the interaction between the variable of interest and Time 1
WFDMSE was entered in the third step of the regression (see Table 16). Note that the
impact of family role importance could not be examined because the variable is
extremely skewed and thus violates one of the assumptions of regression. Therefore,
Hypothesis 23 was not examined.

Of the ﬁve remaining variables examined, two explain signiﬁcant incremental
variance in the prediction of Time 2 WFDMSE when controlling for Time 1 WFDMSE:
relationship status and core self-evaluations. Being in a relationship is associated with
smaller increases in WFDMSE than being single (unstandardized beta = -.23, p = .03).
This is in the opposite direction of Hypothesis 22. More positive core self-evaluations
are associated with greater increases in WFDMSE than lower core self-evaluations
(unstandardized beta = .24, p = .02). This is in the predicted direction of Hypothesis 24.
Overall, Hypotheses 19 through 23 were rejected while Hypothesis 24 was supported.

As noted above, supplemental analyses investigated an interaction between the

variables of interest and Time 1 WFDMSE in the prediction of Time 2 W F DMSE. As

101

can be seen, there are two signiﬁcant interactions with WFDMSE: age and marital status.
Figures 21 and 22 depict these interactions.

For age, younger participants (one s.d. below the mean) reported higher levels of
WFDMSE at Time 2 than older participants (one s..d above the mean) at all levels of
Time 1 WFDMSE. The interaction effect is reﬂected in the slopes of the two lines in
Figure 21. Young participants who are high in Time 1 WFDMSE show the greatest
increases in Time 2 WFDMSE. In other words, the experiment is most impactful for
these participants.

The interaction between Time 1 WFDMSE and marital status is depicted in
Figure 22. Unmarried participants showed higher levels of Time 2 WFDMSE at all
levels of Time 1 WFDMSE (as compared to married participants). The signiﬁcant
interaction indicates that unmarried participants with high initial levels of WFDMSE
were the most responsive to the intervention. (i.e., their WFDMSE showed the greatest

increase after the intervention).

102

Table 16
Moderated Regressions: H19 — H24

 

Dependent variable = WFDMSE Time 2

 

 

 

 

 

 

R2 RZA p

Step 1: TIWFDMSE .72 .72 .00
Step 2: Gender .72 . .00 .62
Step 3: TIWFDMSE x Gender .73 .01 .12
Step 1: TIWFDMSE .72 .72 .00
Step 2: Age .72 .01 .37
Step 3: TIWFDMSE x Age .75 .03 .02
Step 1: TIWFDMSE .72 .72 .00
Step 2: Step 1: Marital status .73 .02 .07
Step 3: TIWFDMSE x Marital .76 .03 .02
status

Step 1: TIWFDMSE .72 .72 .00
Step 2: Relationship status .74 .03 .03
Step 3: TIWFDMSE x Relationship .75 .01 .29
status

Step 1: TIWFDSME .72 .72 .00
Step 2: CSE .74 .03 .02
Step 3: TIWFDMSE x CSE .75 .01 .14

 

Notes. TIWFDMSE = Time 1 Work-family decision-making self-efﬁcacy. R1 = Role
importance. CSE = Core self-evaluations.

 

WFDMSE by Age

 

 

, u —+— Young
n--«-—----"""' . .1 .. Old

 

 

 

Time 2 WFDMSE

 

1 2
Time 1 WFDMSE

 

 

 

Figure 21. Interaction Between Work-family Decision-making Self-
efﬁcacy and Age.
Notes. WFDMSE = Work—family decision-making self-efﬁcacy

103

 

WFDMSE by Marital Status

 

 

 

 

 

 

 

 

5 '1
LU 4 /
U) /
E
Lat. 3 +Not married
N 2 P,” —I——Mam'ed
GD
15'
1— 1
0 i .

 

 

1 2
Time 1 WFDMSE

 

 

 

Figure 22. Interaction Between Work-family Decision-making Self-
efﬁcacy and Marital Status.

Notes. WFDMSE = Work-family decision—making self-efﬁcacy

Outcomes of A WF C and WFDMSE

Hypotheses 25 through 45 examine the relationship between AWF C, WFDMSE,
and outcomes of interest (certainty regarding medical specialty plans, certainty regarding
family plans, requests for the electronic newsletter, viewing the website, and gathering
information during medical school). Over the time lag, I expected earlier levels of
AWF C and WFDMSE to inﬂuence later levels of the outcomes of interest. Each
outcome of interest is described in turn below.

Certainty of medical specialty plans. Table 17 presents the results of path
analyses examining the relationship between AWFC and certainty regarding medical
specialty plans over time. An examination ofthe ﬁndings in Table 17 indicates that for

two of the six types ofAWFC, the path from Time 2 AWFC to Time 3 certainty was

104

signiﬁcant: time-based WIF and strain-based WIF. Note, however, that the paths from
Time 1 AWFC to Time 2 certainty were not signiﬁcant for any of the types of AWF C.
Therefore, Hypothesis 25 was unsupported, by-and-large. Additionally, because most of
the paths across variables were not signiﬁcant, Hypothesis 26 (which suggests that the
direction of causality is from AWF C to certainty regarding medical specialty) is also not

supported.

105

Table 17

Path Analyses for H25 -26: Relationship between Anticipated Work-family Conﬂict and
Certainty Regarding Medical Specialty Over Time

 

 

 

 

 

 

 

Type of AWFC N x2 P CFI RMSEA COV. at T1 SE (p)
(p)

Time-based 91 21.24 6 .00 .97 . .17 -.15 (.15)

WIF
TlMEDCERT 9 T2TWIF .05 (.42)
TlTWIF 9 T2MEDCERT .01 (.73)
T2MEDCERT 9 T3TWIF .01 (.87)
T2TWIF 9 T3 MEDCERT -.12 (.03)

Time-based 91 22.6 .00 .96 .18 -.23 (.03)

F IW
TlMEDCERT 9 T2TFIW .10 (.19)
TITFIW 9 T2MEDCERT -.01 (.80)
T2MEDCERT 9 T3TFIW .14 (.09)
T2TFIW 9 T3 MEDCERT -.04 (.50)

Strain-based 91 23.60 .00 .97 .18 -.27 (.01)

WIF
TlMEDCERT 9 TZSWIF .05 (.35)
TISWIF 9 T2MEDCERT -.02 (.65)
T2MEDCERT 9 T3SWIF -.06 (.36)
TZSWIF 9 T3 MEDCERT -.12 (.03)

Strain-based 91 22.51 .00 .97 .18 -.16 (.13)

FIW
TlMEDCERT 9 T2SFIW -.01 (.93)
TlSFlW 9 T2MEDCERT -.05 (.17)
T2MEDCERT 9 T3SFIW .00 (.98)
T2SFIW 9 T3 MEDCERT -.08 (.14)

Behavior—based 91 20.30 .00 .97 .16 -.02 (.87)

WIF
TlMEDCERT 9 TZBWIF -.09 (.17)
TIBWIF 9 T2MEDCERT -.04 (.29)
T2MEDCERT 9 T3BWIF .01 (.85)
TZBWIF 9 T3 MEDCERT -.02 (.77)

Behavior-based 91 10.62 .10 .99 .09 .00 (.99)

F [W
T1 MEDCERT 9 TZBFIW .00 (.99)
TIBFIW9 T2MEDCERT -.06 (.10)
T2MEDCERT 9 T3BFIW -.10 (.91)
TZBFIW 9 T3 MEDCERT —.02 (.67)

 

Notes. COV. = Covariance at T1. SE = Standardized estimate (direction of arrows
reﬂects direction ofthe path). T1 = Time 1. T2 = Time 2. T3 = Time 3. T = time-based,
S = strain—based, B = behavior-based, WIF = work-interference-with-fami1y, FIW =
family-interference-with-work. MEDCERT = Certainty regarding medical specialty

plans.

106

Table 18 presents the results of a path analysis of the relationship between
WFDMSE and certainty regarding medical specialty plans over time. Results indicate
that there is a signiﬁcant correlation between these two variables at Time 1.
Additionally, two of the four paths from one variable to the other are signiﬁcant. One
signiﬁcant path is from T2 WFDMSE to Time 3 certainty. Another signiﬁcant path is
from Time 1 certainty to Time 2 WFDMSE. Overall, then, there is mixed evidence
regarding earlier levels of WFDMSE inﬂuencing certainty over time. Thus, Hypothesis
27 is partially supported. However, there is no evidence that the paths from WFDMSE to
certainty are stronger than the paths from certainty to WFDMSE. Therefore, Hypothesis
28 is not supported.

Table 18

Path Analyses for H2 7-28: Relationship between Work-family Decision-making Self-
efficacy and Certainty Regarding Medical Specialty Over Time

 

 

Variable N 16 Df P CFI RMSEA COV.atTl SE (p)
(p)
WFDMSE 91 17.44 6 .01 .98 .15 .31(.01)
TlMEDCERT 9 T2WFDMSE .17 (.01)
TIWFDMSE -> T2MEDCERT .03 (.51)
T2MEDCERT —> T3WFDMSE .13 (.06)
T2WFDMSE -> T3MEDCERT .13 (.02)

 

Notes. COV.— Covariance at T1. SE= Standardized estimate (direction of arrows
reﬂects direction of the path). MEDCERT: Certainty regarding medical specialty plans.
WFDMSE= Work- family decision- making self-efﬁcacy.

Certainty regarding family plans. Table 19 presents the results of path analyses
examining the relationship between the six types of AWFC and certainty regarding
family plans over time. The paths between AWF C and certainty regarding family plans
over time were largely not signiﬁcant. The two paths that were signiﬁcant were from

Time 1 AWFC to Time 2 family plans certainty. Higher time-based F [W at Time 1 was

associated with lower family plans certainty at Time 2. Higher strain-based FIW at Time

107

1 was associated with lower family plans certainty at Time 2. These ﬁndings are in the
direction predicted in Hypothesis 29. This data suggests that Time 1 time-based FIW and
strain-based F IW may be inﬂuencing certainty at Time 2. This is in line with Hypothesis
29. However, paths from Time 2 levels of conﬂict to Time 3 levels of certainty were not
signiﬁcant. Therefore, Hypothesis 29 is partially supported for time-based WIF and
strain-based FIW. Hypothesis 30 predicted that the strength of paths from AWFC to
certainty would be stronger then the reverse. Overall, the data presented here does not

support this conclusion, although the two signiﬁcant paths are in the direction predicted.

108

Table 19
Path A nalyses for H29-30: Relationship between Anticipated Work-family Conﬂict and
Certainty Regarding Family Plans Over Time

 

 

 

 

 

 

 

Type of AWFC N 76 Df P CFI RMSEA COV. at T1 SE (p)
(p)
Time-based WIF 91 18.17 6 .01 .97 - .15 .01 (.93)
TlFAMCERT 9 T2TWIF .01 (.92)
TITWIF 9 T2FAMCERT -.07 (.15)
T2FAMCERT 9 T3TWIF -.07 (.40)
T2TWIF 9 T3 FAMCERT -.05 (.36L
Time-based FIW 91 28.90 6 .00 .94 .21 .02 (.85)
TlFAMCERT 9 T2TFIW .08 (.25)
TITFIW 9 T2FAMCERT -.10 (.04)
T2FAMCERT 9 T3TFIW -.13 (.12)
T2TFIW 9 T3 FAMCERT -.05 (.38)
Strain-based WIF 91 28.10 6 .00 .95 .20 -/13 (.19)
TlFAMCERT 9 T2SWIF -.03 (.62)
TISWIF 9 T2FAMCERT —.09 (.08)
T2FAMCERT 9 T3SWIF -.09 (.18)
T2SWIF 9 T3 FAMCERT .00 (.95)
Strain-based FIW 91 28.57 6 .00 .95 .20 -.12 (.25)
TlFAMCERT 9 TZSFIW -.07 (.20)
TISFIW 9 T2FAMCERT -.10 (.03)
T2FAMCERT 9 T3SF 1W -.10 (.18)
TZSFIW 9 T3 FAMCERT -.03 (.55)
Behavior-based 91 12.06 6 .06 .98 .ll , -.15 (.15)
WIF
TlFAMCERT 9 T2BWIF -.07 (.29)
TIBWIF 9 T2FAMCERT -.01 (.88)
T2FAMCERT 9 T3BWIF -.02 (.81)
T2BWIF 9 T3 FAMCERT -.01 (.92)
Behavior-based 91 14.41 6 .03 .98 .13 -.15 (.16)
F IW
TlFAMCERT 9 TZBFIW .02 (.82)
TIBFIW 9 T2FAMCERT -.02 (.76)
T2FAMCERT 9 T3BFIW -.02 (.77)
T2BFIW 9 T3 FAMCERT .00 (.94)

 

Notes. COV. = Covariance at T1. SE = Standardized estimate (direction of arrows
reﬂects direction of the path). T1 = Time 1. T2 = Time 2. T3 = Time 3. T = time-based,
S = strain-based, B = behavior-based, WIF = work-interference-with-family, F [W =
family-interference-with-work. F AMCERT = Certainty regarding family plans.

Table 20 presents the results ofa path analysis ofthe relationship between

WFDMSE and certainty regarding family plans over time. Results indicate that there is a

109

signiﬁcant correlation between these two variables at Time 1. In this analysis, one path
across variables is close to reaching signiﬁcance. There is a positive, non-signiﬁcant
path from Time 1 WFDMSE to Time 2 certainty regarding family plans. This is in line
with Hypothesis 31 (given that WFDMSE is seen as inﬂuence certainty over time) and
Hypothesis 32 (given the relative strengths of paths between the variables) but does not
fully support it because the path does not reach the signiﬁcance level of p < .05.

Table 20

Path Analyses for H31-32: Relationship between Work-family Decision-making Self-
eﬂicacy and Certainty Regarding Family Plans Over Time

 

 

Variable N 76 Df P CF I RMSEA COV. at T1 SE (p)
(p)
WFDMSE 91 27.08 6 .00 .95 .20 .23 (.03)

TlFAMCERT 9 T2WFDMSE .ll (.09)
T1WFDMSE9 T2FAMCERT .10 (.06)
T2FAMCERT 9 T3WFDMSE .04 (.59)
T2WFDMSE9 T3FAMCERT .07 (.20)

 

Notes. COV. = Covariance at T1. SE = Standardized estimate (direction of arrows
reﬂects direction of the path). FAMCERT = Certainty regarding family plans. WFDMSE
= Work-family decision-making self-efﬁcacy.

Information-gathering during medical school. Correlations between the six types
of AWFC and information-gathering during medical school at Time 1 were examined.
Results indicate that none of the six types of AWFC were signiﬁcantly correlated with
this variable. Therefore, Hypothesis 33 is not supported for any of the six types of
AWFC.

Requests for electronic newsletter. For each of the six types of anticipated work-
family conﬂict, a path analysis was conducted in order to examine the relationship
between AWF C and requests for the electronic newsletter over time (see Figure 4). The
results from each of these path models are presented in Table 21. The path from Time 1

AWFC to requests to receive the newsletter was not signiﬁcant for any ofthe six types of

110

AWF C. The paths from requests to receive the newsletter to Time 3 AWF C were also
not signiﬁcant. The results of these analyses indicate that Hypotheses 35 and 36 were not
supported.

Table 21

Path Analyses for H3 5 -3 6: Relationship between Anticipated Work-family Conﬂict and
Requestsfor the Electronic Newsletter Over Time

 

Type of AWFC N x2 Df P CFI RMSEA SE (p) - Path SE (p) -
from T1 to Path from
Newsletter Newsletter

 

 

 

 

 

 

to T3
Time-based 91 6.92 2 .03 .97 .17 .12(.26) .05 (.51)
WIF
Time-based 91 15.75 2 .00 .89 .28 .07 (.50) .01(.95)
FIW
Strain-based 91 14.99 2 .00 .94 .27 -.10(.35) .07 (.29)
WIF
Strain-based 91 16.33 2 .00 .93 .28 -.04 (.74) .04 (.61)
FIW
Behavior-based 91 3.69 2 .16 .99 .10 -.10 (.34) .02 (.82)
WIF
Behavior-based 91 5.88 2 .05 .97 .15 -.06 (.54) -.15 (.05)
FIW

 

Notes. SE = Standardized estimate. WIF = Work-interference-with-family. F IW =
Family-interference-with-work.

Table 22 presents the results of a path analysis investigating the relationship
between WFDMSE and requests to receive the electronic newsletter (see Figure 4).
Neither of these paths was signiﬁcant. Therefore, Hypothesis 37 and 38 were also not

supported.

Ill

Table 22
Path Analyses for H3 7-3 8: Relationship between WFDMSE and Requests for the
Electronic Newsletter Over Time

 

 

Variable N x2 Df P CFI RMSEA SE (p) - Path SE (p) -
from T1 Path from
WFDMSE Newsletter
to to T1
Newsletter WFDMSE

WFDMSE 91 13.92 2 .00 .93 .26 -.02 (333) .03 (.65)

 

Notes. SE = Standardized estimate. WFDMSE = Work-family decision-making self-
efﬁcacy.

 

Viewing the website. Path analyses were conducted to examine the relationship
between the six types of AWF C and viewing the website. Table 23 shows that none of
the paths from AWF C to viewing the website were signiﬁcant. Therefore, Hypotheses 38
and 39 were not supported.

Table 23

Path Analyses for H3 9—4 0: Relationship between Anticipated Work-family Conﬂict and

Viewing the Website Over Time

Type of AWFC N x2 Df P CFI RMSEA SE (p) - Path SE (p) -
from AWF C Path from
to Website Website to

 

 

 

 

 

 

AWFC
Time-based 91 6.81 2 .03 .97 .16 .06 (.56) .15(.05)
WIF
.Time-based 91 15.50 2 .00 .90 .27 .11(.30) .08 (.32)
FIW
Strain-based 91 15.75 2 .00 .93 .28 -.06(.58) .02 (.80)
WIF
Strain-based 91 11.71 2 .03 .95 .23 —.09 (.49) .02 (.79)
FIW
Behavior-based 91 .99 2 .61 1.00 .00 .04 (.68) .13(.08)
WIF
Behavior-based 91 1.33 2 .52 1.00 .00 .00(.97) -.03 (.67)
FIW

 

Notes. SE = Standardized estimate. WIF = Work-interference-with-family. FIW =
Family-interference-with-work. Website = Viewing the website.

As can be seen in Table 24, the path from WFDMSE to viewing the website was
not signiﬁcant. Therefore, Hypothesis 41 and 42 are also not supported.

Table 24
Path Analyses for H41-42: Relationship between Work-family Decision-making Self-
efﬁcacy and Viewing the Website Over Time

 

Variable N 96 Df P CFI RMSEA SE (p) - SE (p) -
Path from Path from
WFDMSE Website to
to Website WFDMSE

WFDMSE 91 11.20 2 .00 .95 .23 -.02 (.82) .00 (.95)

Notes. SE = Standardized estimate. WFDMSE = Work-family decision-making self-

efﬁcacy. RPK = Role—planning knowledge.

Information-gathering during two week lag. Path analyses were conducted to
examine the relationship between the six types of AWF C and reports of gathering
information during the two-week time lag. Results indicate that none of the paths from

AWFC to gathering information. were signiﬁcant (see Table 25). Therefore, Hypotheses

43 and 44 were not supported.

113

Table 25
Path Analyses for H43-44: Relationship between Anticipated Work-family Conﬂict and
Gathering InformatioﬂOver Two Week Time Lag) Over Time

 

Type of AWFC N 36- Df P CF I RMSEA SE (p) - SE (p) -
Path from Path from
AWFC to Gathering
Gathering Information
information to AWF C

 

 

 

 

 

 

Time-based 91 9.33 2 .01 .95 .20 -.08 (.47) .04 (.59)
WIF
Time-based 91 14.03 2 .00 .90 .26 .00 (.97) .02 (.83)
FIW
Strain-based 91 14.75 2 .00 .94 .27 -.02 (.87) .03 (.71)
WIF
Strain-based 91 13.13 2 .00 .94 .25 -.10(.36) .04 (.62)
FIW
Behavior-based 91 1.34 2 .51 1.00 .00 .08 (.45) .68 (.70)
WIF
Behavior-based 91 1.07 2 .59 1.00 .00 .03 (.81) -.13 (.09)
FIW

 

Notes. SE = Standardized estimate. WIF = Work-interference-with-family. FIW =
F amily-interference-with-work.

As can be seen in Table 26, the path from WFDMSE to gathering information
over the two week time lag was not signiﬁcant. Therefore, Hypothesis 44 is not
supported.

Table 26 - -
Path Analyses for H4 I : Relationship between Work-family Decision-making Selflefﬂcacy
and Gathering Information (Over the Two Week Time Lag) Over Time
Variable N )6 Df P CFI RMSE SE (p) - SE (p) -
A Path from Path from
WFDMSE Gathering
to Gathering Information
Information to
WFDMSE
WFDMSE 91 10.09 2 .01 .95 .21 .07 (.50) .10 (.15)

 

Notes. SE = Standardized estimates. WFDMSE = Work-family decision-making self-
efﬁcacy.

114

Mediation of the Relationship between WFDMSE and Outcomes

Hypotheses 45 to 50 examine the whether the six forms of AWF C mediate the
relationship between WFDMSE and the outcomes of interest (certainty regarding medical
specialty plans, certainty regarding family plans, gathering information during medical
school, requests for the electronic newsletter, viewing the website, and gathering
information during the two week time-lag). According to Baron and Kenny (1986) there
are certain relationships that must be established for mediation to be tested. One critical
relationship is that the predictor variable must be signiﬁcantly related to the proposed
mediator. In the path analysis conducted examining the relationship between WFDMSE
and AWF C over time (see Table 11), one can see that the path from Time 1 WFDMSE to
Time 2 AWF C was not signiﬁcant for any of the six types of AWFC. Therefore, this
prerequisite for mediation is not met and mediation should not be tested. Therefore,
Hypotheses 45 — 50 are not supported.

DISCUSSION

The current study examined the construct of anticipated work-family conﬂict and
attempted to inﬂuence levels of AWFC through an intervention focused on increasing
work-family decision-making self-efﬁcacy. In the following sections, 1 will highlight
and discuss the major ﬁndings from the study, including directions for future research
and practice that emerge from these ﬁndings. 1 will then discuss some potential
limitations to the current research.
Predictors of A nticipated Work—family Conﬂict

I examined the relationship between several predictor variables and participants’

pre-existing levels of anticipated work-family conﬂict. By-and-large, demographics were

unrelated to AWF C. This is similar to the ﬁndings from F riede (2005) who did not
identify strong relationships between demographics and the six types of AWFC.
However, in the current study, I did ﬁnd that women anticipated higher levels of
both behavior-based WIF and behavior-based FIW. I expected that women would
anticipate higher levels of behavior-based WIF because they were expected to place a
stronger importance on the family role and were thus trying to protect it from spillover
from the work role. F riede (2005) examined the relationship between gender and AWF C
(using the same measure used here) and found that women anticipated less behavior-
based WIF and less time-based WIF (in line with the hypotheses from the current study
suggesting that women will anticipate less work-interference-with-family). The ﬁndings
from the current study suggest that women may be particularly sensitive to the potential
for behavior-based conﬂict. Whereas prior research on experiences of actual work-
family conﬂict has found mixed data regarding gender differences in work-family
conﬂict (Eby et al., 2005), this research points to gender differences in anticipated
behavior-based conﬂict. Future research should continue to examine this issue. In
particular, research should investigate the differences between male and female
professional students in their expectations regarding future work-family conﬂict to
identify mechanisms for supporting male and female students in their transitions into
work and family roles. A deeper understanding of gender-based differences in AWF C
will facilitate the custom—tailoring of interventions that will best meet the distinct needs
of male and female participants. The ﬁndings from the current study indicate that female

participants may particularly beneﬁt from interventions that address issues associated

116

with anticipated behavior-based conﬂict and teach strategies for dealing with the
differing behavioral patterns that are effective for women at work and at home.

The current study also examined the relationship between core self-evaluations
and anticipated work-family conﬂict and found that individuals with higher core self-
evaluations anticipated less strain-based and behavior-based conﬂict in both directions.

F riede (2005) found core self-evaluations to be negatively correlated with all six types of
AWFC. In general, these ﬁndings lend support to the notion that individuals with higher
core self-evaluations anticipate less work-family conﬂict. In both studies, the
relationships were particularly strong for strain-based conﬂict in both directions. Future
research should examine whether individuals with higher core self-evaluations actually
do experience less strain-based work-family conﬂict.

Friede and Ryan (2004) describe three possible reasons why core self-evaluations
may be linked to experiences of work-family conﬂict. They note that individuals with
higher core self-evaluations may make more effective choices about which work and
family roles to enter and how to manage those roles. They may have more positive
perceptions of those work and family roles and how they can be balanced (even when
controlling for objective features of the work and family environments). Finally,
individuals with more positive core self-evaluations may select different coping strategies
or more effectively implement the same coping strategies as individuals with more
negative core self-evaluations. Given these arguments, individuals with more negative
core self-evaluations may beneﬁt from greater support as they transition into work and
family roles. This support may take different fomis. Individuals with lower core self-

evaluations may beneﬁt from support that helps them to perceive their future work and

117

family environments as more positive and controllable. They may also beneﬁt from
additional support that helps them in the process of making decisions about which work
and family roles to enter. Additionally, they may beneﬁt from extra support choosing
and effectively implementing coping strategies. Future research should address the
effectiveness of these different types of interventions for providing supplemental support
to individuals with more negative core self-evaluations as they transition into work and
family roles.

Furthermore, in the current study, both work and family role-importance were not
signiﬁcantly correlated with any of the six types of AWF C. In general, F riede (2005) did
not ﬁnd strong correlations among these variables either. As mentioned in the Method
section of this paper, family role-importance was negatively skewed indicating that the
majority of students reported extremely high levels of family role-importance. While this
may be a function of social desirability, it is also quite possible that these students
genuinely place a high importance on family. Further, although there is more variability
with regard to work role-importance within the sample, one might expect medical
students to have relatively high levels of work role—importance when compared to
individuals who have pursued less demanding careers. In other words, future research
should investigate the relationship between work role-importance and AWF C with a
broader sample than medical students. However, given the data presented here, it may be
a challenge to secure a sample that has signiﬁcant variability on the family role.

In the current study, I also found that both role-planning knowledge and
WFDMSE were signiﬁcantly negatively correlated with all six types of AWFC.

Conﬁdence and knowledge about how to plan for multiple roles and make decisions

118

about how to balance them are associated with fewer concerns about work-family conﬂict
in the future. However, note that the cross-lagged path analyses did not support the
ﬁnding that earlier WFDMSE and role-planning knowledge were causing later levels of
AWFC. However, this may be due to the short time ﬂame and the fact that the Time 1
relationships among the variables were controlled for, as were participant prior levels on
each variable (i.e., autoregressive paths were included). This means that much of the
variability in Time 2 and Time 3 AWFC was already accounted for. For example, Time

1. strain-based FIW accounts for over 70% of the variance in Time 2 strain-based F IW.
Given these ﬁndings, the causal relationship over time between WFDMSE, role-planning
knowledge, and AWF C is not clear. Future research that examines changes in
WFDMSE, role-planning knowledge, and AWF C over a period of years that spans major
life transitions (e. g., entering medical school, starting a residency, making a decision
about a job) would be valuable for drawing conclusions regarding the causal relationships
between these constructs over time. Future research along these lines would also allow
an investigation of whether AWF C and role-planning knowledge are reciprocally related,
inﬂuencing one another throughout the lifespan.

At this point, it is also important to address the common theme of self-efﬁcacy
that underlies a number of predictor variables assessed in the study. Core self-
evaluations, CDSE, WFDMSE, and role-planning knowledge may all be driven by an
underlying generalized sense of self-efﬁcacy. Judge, Erez, Bono, & Thoresen (2002)
speciﬁcally discuss that generalized self—efﬁcacy is one of the four components of core
self-evaluations. Additionally, Chen, Gully, and Eden (2001) note that generalized self-

efﬁcacy is most likely a predictor of more speciﬁc types of self-efﬁcacy (i.e., CDSE and

119

WFDMSE). Furthermore, role-planning knowledge (a self-rated perception of
knowledge) may be inﬂuenced by generalized self—efﬁcacy given the conﬁdence that
underlies beliefs in one’s own knowledge (a type of knowledge that has not been
examined and evaluated in more traditional ways): Correlations among these constructs
range from r = .46 (core self-evaluations and CDSE) to r = .75 (CDSE and WFDMSE).
These correlations are certainly substantial. Nonetheless, factor analyses indicated that
the measures are empirically distinguishable (see Method section). In other words,
although generalized self-efﬁcacy may strongly inﬂuence each of these constructs, they
each have their own unique distinguishing features. In general, social-cognitive career
theorists have favored more speciﬁc measures of self-efﬁcacy that are directly matched to
the criterion of interest (Betz & Luzzo, 1996). Given the ﬁndings presented here and the
social-cognitive tradition, I decided to treat these scales as distinct despite their high
intercorrelations. In the next section, I will describe how this choice may have inﬂuenced
the ﬁndings in the current study. In the current section (which focuses on the predictors
of AWF C), it sufﬁces to say that AWF C is most likely inﬂuenced by generalized self-
efﬁcacy. However, it is not surprising that self-efﬁcacy constructs more speciﬁcally tied
to career and work-family notions are strongly related to AWF C. Similar to arguments
provided above that suggest that individuals with low core self-evaluations might beneﬁt
from supplemental training due to their higher levels of AWFC, the ﬁndings presented
here suggest that individuals with low generalized self-efﬁcacy (a component of core
self-evaluations) may beneﬁt from supplemental training that helps them to avoid the
higher levels of work-family conﬂict that they tend to anticipate.

Effects of Intervention

The next portion of the study examined the effects of the Career and Work-family
intervention on WFDMSE and AWFC. Interestingly, there were pre-existing differences
in WFDMSE across the two conditions, even though participants were randomly assigned
to condition. Therefore, Time 1 WFDMSE was controlled in my examination of the
effects of the interventions on Time 2 WFDMSE. As noted above, the Work-family
intervention did not signiﬁcantly increase WFDMSE more so than the Career
intervention. The desired effect of increasing WFDMSE was achieved in both groups
and this effect persisted over the two-week time lag. In other words, it is not that the
Work-family intervention was ineffective in increasing WFDMSE, it is that the Career
intervention was also effective in increasing WFMDSE. Both conditions were designed
to be very similar (the basic wording of the scripts is almost identical and only
differences are related to work—family versus career decision-making). Both conditions
included Bandura’s mechanisms for increasing self-efﬁcacy.

Therefore, one possible interpretation of the ﬁndings is that interventions
focusing on self-efﬁcacy are likely to increase WFDMSE regardless of the speciﬁc focus
of the intervention. In other words, participant overall level of efﬁcacy may matter more
so than their speciﬁc level of efﬁcacy towards a particular type of decision-making. This
would suggest that it is not critical to incorporate speciﬁc interventions focusing on
WFDMSE into the career counseling provided to professional students.

An alternative explanation is that WFDMSE can be conceptualized as a speciﬁc
type of CDSE (i.e., career choices associated with balancing work and family). Indeed,
the Work-family condition did focus on making specialty choices (i.e., career choices)

based on information about work and family. This interpretation would suggest that it is

not necessarily generalized self—efﬁcacy that is critical, rather it is career-related
decision-making self-efﬁcacy that is critical. In other words, because the two
interventions were both focused on career decision-making, they were both effective in
inﬂuencing WFDMSE. In this case, the developmental opportunities provided to
professional students should still focus on CDSE (and may or may not include a work—
family component) and generalized self—efﬁcacy may not be sufﬁcient.

A third possibility is that participants in the career condition chose to think about
work-family balance during the intervention as part of the “performance
accomplishments” component of increasing self-efﬁcacy (e. g., when asked to write about
the aspects of their career that were most important to them). In other words, although
the Career intervention did not explicitly mention work-family balance, students might
have chosen to consider issues of work-family balance when addressing issues of career
choice and the medical specialty that they plan to go into. Participants’ responses during
the intervention were available for examination. One question asked participants to
identify what features of their medical specialty were most important to them. 50% of
participants in the Career condition mentioned a characteristic related to work-family
balance as a feature of their specialty that is most important to them (e.g., ability to work
less than full-time, time for family, ability to leave work at the ofﬁce to spend time with
family). This suggests that many participants in the Career condition were thinking about
work-family issues, even though they were not explicitly told to do so. This may explain,
in part, why participants in the Career condition also showed increases in WFDMSE. In
other words, the psychological impact of the two conditions was similar because

participants in both groups were thinking about work-family issues. Future research

could include a control condition that explicitly discusses aspects of specialty choice that
are less directly associated with issues of work-family balance (e.g., working with one’s
hands). This would help to distinguish between alternative explanations for why
WFDMSE increased in both conditions in the current study.

It was also expected that participants in the Work-family condition would have
lower levels of AWF C than participants in the Career condition after the intervention.
Contrary to expectations, participants in the Work-family condition anticipated higher
strain-based FIW at Time 2 and higher strain-based WIF and F IW at Time 3 than those in
the Career condition. Perhaps more interesting is an examination of the patterns of
change within condition across the three time points. In both conditions, time-based WIF
decreased and this decrease persisted over the two week time lag. Additionally, time-
based WIF decreased after the intervention for both conditions but this decrease did not
persist over the two week lag. In the Career condition, participants reported decreased
strain—based WIF and this persisted over the time lag. In the. Work-family condition, no
changes in strain-based WIF were evident. For both conditions, there was no signiﬁcant
change in strain-based F IW or behavior-based WIF or F IW.

These ﬁndings suggest that the intervention was effective in reducing AWFC. In
particular, time-based conﬂict was the type of anticipated work-family conﬂict most
strongly affected by the interventions. One explanation for these ﬁndings is that time-
based conﬂict is perceived as the most controllable form of conﬂict (i.e., students can use
their decision-making skills to inﬂuence their levels of time-based AWFC). Time-based
conﬂict is a particularly concrete type of conﬂict that participants can readily understand.

Additionally, the types of decisions that one would need to make in order to reduce time—

123

based conﬂict are straight-forward. On the other hand, strain-based and behavior-based
conﬂict may seem more abstract to individuals. The types of decisions that would need
to be made in order to avoid these types of conﬂict may also be less clear. Enhancing
participant self-efﬁcacy may only inﬂuence their anticipation of conﬂict that they
perceive as under their control.

Future research could empirically assess the extent to which the three types of
conﬂicts are perceived as controllable. Furthermore, it would be valuable to investigate
the extent to which these three types of conﬂict are actually controllable. Interventions
should be focused on increasing participant conﬁdence and capabilities for making
decisions that have the possibility of helping them to avoid work-family conﬂict.
Interventions that encourage participants to try to control things that are actually out of
their control may be detrimental to their well—being. This is substantiated by research
that suggests that active attempts to problem-solve may be ineffective in less controllable
situations (Miller & Kirsch, 1987; Mak & Mueller, 2000; Parkes, 1990). Therefore, in
the creation of interventions, researchers should be cautious about creating illusory
control in situations where individuals have little or no control. Note, however, that there
is also evidence that mildly exaggerated perceptions of control may actually beneﬁt well-
being despite their inaccuracy (see Taylor & Brown, 1994; Taylor, Kemeny, Reed,
Bower, & Gruenewald, 2000, for a discussion of this issue). Given these ﬁndings,
researchers should carefully investigate the positive and negative outcomes associated
with enhancing participant perceptions of control over work-family conﬂict in cases
where control does and does not exist.

Factors Influencing the Degree of Change in WFDMSE

In addition to examining the impact of the interventions on WFDMSE and
AWFC, I examined the factors that inﬂuenced the degree of change in WFDMSE in the
Work-family condition. With regard to demographics, only one was related to the degree
of change in WFDMSE. Individuals in a relationship reported greater increases in
WFDMSE than single individuals. This was in line with our hypothesis which suggested
that individuals in relationships would be more concerned about Ieaming how to make
decisions about balancing work and family, and therefore would beneﬁt more from the
intervention. Note, however, that I did not directly examine the psychological
mechanisms that may have caused this effect (e. g., desire to learn more about how to
balance work and family). There were no pre-existing differences in anticipated work-
family conﬂict among individuals in relationships and those not in relationships.
Therefore, if future researchers are interested in examining how different individuals
react to the types of interventions conducted in this study, it is important to also
empirically investigate why such differences among individuals emerge, rather than
assuming the causes of the differences. Importantly, as research in this area progresses,
customizing interventions to the needs of speciﬁc demographic groups may become a
possibility and in-depth research would be necessary as part of the customization process.

Furthermore, as predicted, individuals with more positive core self-evaluations
showed greater increases in WFDMSE after the intervention. It was hypothesized that
these participants would have greater conﬁdence in their ability to implement the material
provided in the intervention and to inﬂuence their future work-family balance. This
ﬁnding may indicate that having high generalized self-efﬁcacy (a component of the core

self-evaluations construct) increases sensitivity to more speciﬁc self-efﬁcacy training.

As noted previously, there is expected to be a dynamic relationship between generalized
and speciﬁc self-efﬁcacy (Chen et al., 2001). The implications of this ﬁnding are
therefore complex. Perhaps self-efﬁcacy-based interventions should include a
component focused on enhancing generalized self-efficacy prior to focusing on more
speciﬁc types of self-efﬁcacy. There is evidence that high generalized self-efﬁcacy
primes individuals for motivation and learning during training (Salas & Cannon-Bowers,
2001). On the other hand, perhaps it is important to focus interventions on speciﬁc types
of self-efﬁcacy and the conﬁdence that participants gain will spillover to enhance a more
generalized sense of self-efﬁcacy. Research conducted by Schwoerer, May, Hollensbe,
and Mencl (2005) does provide evidence that speciﬁc self-efﬁcacy enhanced through
training can spillover to enhance generalized self-efﬁcacy. Future research can certainly
investigate this issue and, similar to future research on effects of being in a relationship,
may identify opportunities for customizing interventions for participants with low and
high levels of initial core self-evaluations.

Additionally, two interactions between demographics and Time 1 WFDMSE were
identiﬁed (predicting Time 2 WFDMSE) through supplemental analyses. Young
participants and unmarried participants who were high in Time 1 WFDMSE showed the,
greatest increases in Time 2 WFDMSE. These ﬁndings run counter to the argument that
individuals for whom work-family decisions are more quickly approaching will gain the
most from the intervention. Young participants and unmarried participants are farther
from making work-family decisions. This effect only occurred for young and unmarried
participants who were high in initial levels of WFDMSE. One potential explanation is

that these participants may have naive overconﬁdence in their ability to effectively make

126

decisions about balancing work and family. In other words, these participants believe
that they will be able to make effective decisions about balancing work and family and
they believed that strategies suggested in the intervention would help them to do so (more
so than other participants who may have a more realistic appraisal of the beneﬁts of the
intervention). Another possible explanation is that young/unmarried participants with
high initial WFDMSE may feel the most freedom and conﬁdence to successfully
implement the suggestions provided in the intervention whereas other participants are
more constrained by the choices that they have already made or their lack of conﬁdence.
Future research should see if these unhypothesized interactions emerge in replications of
the current study and, if so, investigate the psychological mechanisms that are driving
these relationships.

Outcomes of Anticipated Work-family Conﬂict

For the most part, there was little evidence of causality over time between levels
of AWFC and the outcomes that were assessed. AWFC was not associated with
information-gathering before medical school or during the two week time-lag.
Additionally, AWF C was not associated with requests for the electronic newsletter or
participant viewing of the website. It is important to consider why these relationships did
not emerge as predicted.

One possibility is that the interventions examined in this study affect participants’
perceptions and beliefs, but these cognitive changes do not impact behavioral outcomes.
Behavioral choices may be driven by other factors (such as interests, values, or
demographics). Note that other individual characteristics (e.g., gender, relationship

status, work importance, and family importance) were related to these outcomes.

Alternatively, it may be that the interventions utilized in the current study were not strong
enough to impact behavior but that behavior could potentially be affected through
improving the design and content of the interventions. The limitations associated with
intervention design are discussed in greater detailbelow.

It may also be that the outcomes assessed were not appropriate for the given
study. The theme relating these outcomes is the notion that individuals want to gather
more information about balancing work and family. Perhaps AWFC is not related to a
desire to gather more information in the short term but may be related to more long-term
behavioral outcomes that were not assessed in the current study (e. g., choice of medical
specialty, choice of residency program). Cohort research studies that examine the
relationship between AWFC and the behaviors and choices of professional students over
time (and include a broader range of behavioral outcomes) will be an important addition
to this stream of research.

With regard to the relationship between AWF C and certainty regarding medical
specialty plans, the path from Time 2 time-based WIF to Time 3 certainty was signiﬁcant
and the path from Time 2 strain-based WIF to Time 3 certainty was signiﬁcant. These
ﬁndings are in line with hypotheses. However, the majority of hypothesized paths were
not signiﬁcant.

For certainty regarding family plans, the path from Time 1 time-based F lW to
Time 2 certainty was signiﬁcant. The path from Time 1 strain-based F IW was also
signiﬁcantly related to Time 2 certainty. Again, these paths are in line with hypotheses
but the majority of hypothesized paths were not signiﬁcant. One possible explanation for

this ﬁnding is high stability of these constructs over the time period examined. In other

words, to the extent that there is little change in levels of a construct, it is difﬁcult to
predict change in that construct. Re-examining the relationship between these constructs
over a longer time-period would shed light onto the dynamic relationship between
anticipated conﬂict and certainty that is predicted’by the social-cognitive theory of career
development (Lent et al., 1994).

It is interesting to note the nature of the paths that did emerge as signiﬁcant. For
certainty regarding medical specialty plans, the two types of AWF C that yielded
signiﬁcant paths were both work-interference-with-family (time- and strain-based). For
certainty regarding family plans, the two types of AWFC that yielded signiﬁcant paths
were both fami1y-interference-with—work (time-based and strain-based). In other words,
changes in beliefs about future work-interfering-with-family may inﬂuence people to
adjust their certainty about their work whereas changes in beliefs about future family-
interfering-with-work may inﬂuence people to adjust their certainty about their family
plans. This parallels existing research which shows that stressors in the work domain
primarily impact work-interference-with-family whereas stressors in the family domain
primarily impact family-interference-with-work (e.g., Frone, Yardley, Markel, 1997).
The nature of cross-domain spillover regarding AWF C and certainty merits future
research. In particular, longitudinal, non-experimental research that examines the
relationship between stressors from the work and family domains, conﬂict perceived as
spilling-over from the work and family domains, and certainty regarding plans in the
work and family domains would be extremely valuable to shed further light on these
ﬁndings.

Potential Methodological Limitations

In addition to the suggestions for future research provided above, it is important to
address the potential methodological limitations of the current study and propose
improvements for future research. One main limitation of the study is the sample size.
This limitation is most evident in the fact that path analyses had to be used instead of
structural equation modeling. Measurement error modeled in the path analyses may have
obscured interesting and important relationships between latent constructs. Future
research should focus on attaining a larger sample size in order to facilitate statistical
analysis.

Another potential limitation is the fact that the data was almost entirely self-
report. Relationships across self-reported data may be inﬂated (Podsakoff et al., 2003).
This concern is mitigated to some extent by ﬁndings in the current study that show
differential relationships among self-reported variables and differentiation between
similar constructs. Additionally, this concern is mitigated by the fact that data was
collected over different time points and many relationships examined were among
variables assessed at different time points. To further reduce concerns about common
rater effects, additional raters could provide insight into some of the constructs assessed.
For example, career counselors could provide data regarding student certainty regarding
professional plans. Partners could provide data regarding levels of AWFC and certainty
regarding family plans.

An additional concern is that participants were inﬂuenced by social desirability to
report increased WFDMSE and decreased AWF C. The fact that WFDMSE and AWFC
were assessed multiple times may have suggested to participants in both conditions that

this was the focus of the current study. Note, however, that some of the changes that

130

emerged after the intervention persisted over the two week time-lag. While it is possible
that the participants were inﬂuenced by social desirability at all time points, one might
expect the social desirability effects to decrease over the time lag (farther in time from
the intervention). If possible, future research that'more effectively hides the intention of
the research (perhaps by including additional, “distractor” constructs) would be
beneﬁcial. Additionally, framing the intervention as part of the career training provided
by the medical school and making it appear as though the assessments were unrelated to
the intervention might decrease socially-desirable responding. In general, the
methodology of the study would be improved if a greater time-lag between Time 1, the
intervention, Time 2, and Time 3 were possible.

Another methodological concern in this study is the design of the interventions
themselves. They were both extremely brief, administered over the intemet, and did not
include opportunities for group discussions, questions, or feedback. There is meta-
analytic evidence that key features of the intervention designs for the current study have
been associated with effective training in past research (e. g., lectures, behavioral-
modeling, practice, web-based administration; Burke & Day, 1986 Sitzmann, Kraiger,
Stewart & Wisher, 2006). However, Campbell (1988) and Tannenbaum and Yuk] (1992)
both noted that it is most critical that the method of training match the objective of the
training and participant characteristics. The current interventions were designed based on
well-established mechanisms of enhancing self-efﬁcacy (Bandura & Adams, 1977).
However, future research should address whether training focused on work-family
decision-making self-efﬁcacy and anticipated work-family conﬂict should incorporate

speciﬁc design features. Additionally, understanding the particular training needs of

I31

medical students (and sub-populations within that group, such as female medical
students) will enhance the alignment between the design of the training and the needs of
participants. To the extent that the training content and design can be better matched to
the objectives of the training and the needs of the participants, it may have a much greater
impact.

Furthermore, a related concern is the impact of the training over longer time
periods and outside of the laboratory. Training effects that were found to persist over a
two—week time-lag in the current study may disappear over a longer time-lag.
Additionally, the lack of impact on behavioral outcomes may suggest a lack of transfer of
the training to real-world decision-making. For intervention effects to persist over time
and transfer outside of the training environment, the design of the intervention may need
to be altered. Again, the alignment between the design and content of the training and the
situations in which it is intended to be implemented is critical. The transfer of training
literature suggests that intervention effects are more likely to be maintained outside of the
training environment when the training environment matches the application
environment, participants learn the underlying reasoning behind the information imparted
in training, they see the implications of the training for a variety of situations that they
may later face, practice is distributed over time, and feedback is provided (Baldwin &
Ford, 1988). Certainly, the content and design of the current interventions should be
reﬁned so that it more closely meets the conditions associated with high transfer of

training.

CONCLUSIONS

The ﬁndings presented here add to my correlational research on the antecedents
and consequences of AWFC. Additionally, this research contributes to existing research
being conducted on work-family conﬂict by emphasizing the importance of considering
the thoughts, feelings, and decisions of professional students regarding balancing work
and family prior to entering those roles. Further, this research contributes to the existing
research being conducted on career decision-making by addressing the role of family in
the career decision-making process. Future research should build upon the ﬁndings
presented here to understand the process by which professional students transition into
work and family roles. This research will be critical for designing interventions that help
professional students make decisions that will allow them to be successful and happy in

their multiple life roles.

133

APPENDICES

134

Appendix A
Email Invitation from Dean to Participate in Study

Dear COM/CHM medical students,

I am writing to invite you to participate in an important research study. This study will
allow us a greater understanding of how our medical students select and transition into
their medical specialties and how we can best support you in this process. We will
receive a detailed report based on the ﬁndings from this study. However, we will not
receive your individual responses.

This is a two-part research study that will take approximately 40 minutes of your time
total. Everyone who completes the study will receive a $10 check in the mail.

Again, your participation is greatly appreciated and provides us with data on an important
topic. Thank you in advance for your participation.

Sincerely,

<Insert Dean name>.

Study information:

Study name: Professional Development for Medical Students

Study details: The study has two parts. The ﬁrst part takes approximately 30 minutes and
will involve watching brief online presentations and completing a web-based survey.
The second part of the study will take place two weeks after the ﬁrst part. You will
receive an email asking you to take another web-based survey that will take
approximately 10 minutes to complete. All responses will remain conﬁdential. This
research is completely voluntary and there is no penalty for those who do not participate.
Compensation: All participants who complete both parts of the study will receive a $10
check in the mail. Those who only complete the ﬁrst part of the study will receive a $5

check in the mail.

To participate visit: psychology.msu.edu/balance (Note: there’s no “”www in the URL)
Enter Access ID: medicine

Contact: Dr. Ann Marie Ryan (ryanan@msu.edu) with questions.

Your time and energy are greatly appreciated. They will help your medical school
understand your needs as well as contribute to an important research project.

135

Appendix B
Consent Form Time I

Medical Student Professional Development
Please read the information below:

We will be asking you to respond to a series of questions about how you think about and
plan for your future medical career. You will be asked to view two brief presentations
that are focused on your professional career and answer questions based on what you
learn in the presentations. We are also asking you to respond to some commonly used
personality and demographic questions that will help us interpret the meaning of your
responses to the questionnaire.

This is a two-part research study. We expect that it will take about 30 minutes for you to
complete the ﬁrst part of the study. In two weeks, you will be asked to ﬁll out a second
short survey that will take approximately 10 minutes to complete. Participants who
complete both parts of the research study will receive a $10 check in the mail.
Participants who only complete the ﬁrst part of the research study will receive a $5 check
in the mail.

There are no foreseeable risks associated with participating in this study. Your name and
information will remain conﬁdential. Your privacy will be protected to the maximum
extent allowable by law. The data will be saved for at least ﬁve years after it is collected
and will only be accessible by the primary investigator and one graduate student. By
typing your name below you indicate that you are free to refuse to participate in this
project or any part of this project. You may refuse to participate in certain procedures or
answer certain questions. Participation is completely voluntary. You may choose not to
participate at all and may discontinue your participation at any time without penalty or
loss of beneﬁts.

If you have any questions or concerns about your participation in this project, you can
reach Ann Marie Ryan by phone: (517)353-8855, fax: (517)353-4873, email:
ryanan@msu.edu, or regular mail: 333 Psychology Building, East Lansing, MI 48824. If
you have questions or concerns regarding your rights as a study participant, or are
dissatisﬁed at any time with any aspect of this study, you may contact - anonymously if
you wish - Peter ‘Vasilenko, Ph.D. Chair of the Human Research Protection Program by
phone: (517)355-2180, fax: (517)432-4503, email: irb@msu.edu, or regular mail: 202
Olds Hall, East Lansing, MI 48824.

Please mark the box that says "I agree to give my consent to participate" if you agree to
participate in this study. Mark "I do not want to participate" if you do not agree to
participate in this study. If you agree to participate, enter your name below and you will

be taken to the survey once this step is completed.

[:1 1 agree to give my consent to participate

136

1:1 I do not want to participate
First name Middle initial Last name

 

137

 

Appendix C
Consent F arm Time 3

Medical Student Professional Development
Please read the information below:

We will be asking you to respond to a series of questions about how you think about and
plan for your future medical career. You may be asked to view brief presentations about
career decision-making and answer questions based on what you learn in the
presentations.

This is a research study. We expect that it will take about 10 minutes for you to complete
this study. Because you already completed the ﬁrst part of the study two weeks ago, you
will receive a $10 check in the mail if you complete this second part of the study. You
will still receive a $5 check in the mail for your participation in the ﬁrst part of the study
if you choose not to complete this part. There are no foreseeable risks associated with
participating in this study. Your name and information will remain conﬁdential. Your
privacy will be protected to the maximum extent allowable by law. The data will be
saved for at least ﬁve years after it is collected and will only be accessible by the primary
investigator and one graduate student. By typing your name below you indicate that you
are free to refuse to participate in this project or any part of this project. You may refuse
to participate in certain procedures or answer certain questions. Participation is
completely voluntary. You may choose not to participate at all and may discontinue your
participation at any time without penalty or loss of beneﬁts.

If you have any questions or concerns about your participation in this project, you can
reach Ann Marie Ryan by phone: (517)353-8855, fax: (517)353-4873, email:
ryanan@msu.edu, or regular mail: 333 Psychology Building, East Lansing, MI 48824. If
you have questions or concerns regarding your rights as a study participant, or are
dissatisﬁed at any time with any aspect of this study, you may contact - anonymously if
you wish - Peter Vasilenko, Ph.D. Chair of the Human Research Protection Program by
phone: (517)355-2180, fax: (517)432-4503, email: irb@msu.edu, or regular mail: 202
Olds Hall, East Lansing, MI 48824.

Please mark the box that says "I agree to give my consent to participate" if you agree to
participate in this study. Mark "I do not want to participate" if you do not agree to
participate in this study. If you agree to participate, enter your name below and you will
be taken to the survey once this step is completed.

1:1 I agree to give my consent to participate
1:1 I do not want to participate

First name Middle initial Last name

 

 

138

Appendix D
Email invitation to participate in Time 3

Hello,

Thank you for your participation in the research study, “Medical Student Professional
Development.” We greatly appreciate your time and effort.

When you completed the ﬁrst data collection, you agreed to be re-contacted to complete a
follow-up survey. If you complete the follow-up survey, you will receive a check for $10

in the mail. The survey should take approximately 10 to 15 minutes.

To participate visit: psychology.msu.edu/balance (note: no “www” in the URL)
Enter Access ID: doctor

Please contact Ann Marie Ryan at ryanan@msu.edu with any questions.

Thank you very much

139

Appendix E
Debrieﬁng Form

Thank you for your participation in this study. The purpose of this study was to reduce
the anticipated work-family conﬂict experiencedby medical students. To do this, we
designed a web-based intervention that consisted of series of Powerpoint presentations
focused on building self-conﬁdence in your ability to make decisions that will allow you
to successﬁrlly balance your medical career with your family life. Because we wanted to
examine the effectiveness of this intervention, we used an experimental study with three
different conditions. Some participants participated in the intervention focused on work-
family decision-making. Some participants participated in the intervention focused on
career decision-making (with no discussion of work-family issues). Some participants
did not participate in an intervention at all. You were randomly assigned to one of these
three experimental conditions. There are no foreseeable risks associated with
participation in any of these conditions. Participants who completed both parts of the
survey received a $10 check in the mail.

Additionally, for those of you who were interested in receiving additional information,
we sent you an email with additional information. That email included a link to a
website. We tracked your time and actions on the website for research purposes. We
were interested in the extent to which medical students were interested in obtaining more
information about work-family issues. This information will be linked with your survey
data and then all identifying information will be removed ﬁom the data.

We believe that the intervention focused on the work-family balance of medical students
may be helpful to students who are thinking about their futUre work and family plans. If
you would like the opportunity to view the work-family intervention, you may do so at
your convenience.

Visit psychology.msu.edu/balance (Note: there’s no “www” in the URL)
Access ID: debrief

If you have any questions or concerns about your participation in this project, you can
reach Ann Marie Ryan by phone: (517)353-8855, fax: (517)353-4873, email:
ryanan@msu.edu, or regular mail: 333 Psychology Building, East Lansing, MI 48824. If
you have questions or concerns regarding your rights as a study participant, or are
dissatisﬁed at any time with any aspect of this study, you may contact - anonymously if
you wish - Peter Vasilenko, Ph.D. Chair of the Human Research Protection Program by
phone: (517)355-2180, fax: (517)432-4503, email: irb@msu.edu, or regular mail: 202
Olds Hall, East Lansing, MI 48824.

140

Appendix F
Electronic Newsletter Email

Work-Family Balance
A Newsletter for Medical Students

Overview:

This email provides you with an overview of some of the latest research and information
that can help medical students think about choosing a medical specialty and balancing
work and family.

To get more information about the topics provided below, click on the link below.
http://psychology.msu.edu/Balance/TheBIDLink.asp?BID=93

If you are not automatically directed to the website, copy and paste the link into your
browser.

Below are brief summaries of the topics that you can read about on the website.

9 Strategies for Balancing Work and Family Research has identiﬁed nine strategies for
achieving better balance of work and family that you can use as a doctor. You can read
about those nine strategies in this article.

3 Characteristics of Residency Programs to Consider This article summarizes three
important characteristics of residency programs to consider when deciding which
program to enter. Read this article to learn about the importance of considering program
stability, support, and institutional climate in your decision-making process.

Work—Family Demands of Different Specialties Medical specialties differ in the work-
family challenges that they pose for doctors. A recent study asked practicing doctors to
rate the work-family demands of different medical specialties. The results of that study
are reported in this article.

Average yearly earnings for doctors

The average salary of doctors differs by medical specialty. This article summarizes the
average salaries of doctors in different specialties, as provided by the Association of
American Medical Colleges.

Again, to read these different articles, click on the following link:
http:l/psychology.msu.edu/Balance/TheBlDLink.asp?BID=93
If you are not automatically directed to the website, copy and paste the link into your
browser.

141

Appendix G
Powerpoint Slides for Career Condition Including Tasks

INTRODUCTION AND TUTORIAL 1
Welcome to “Career Flaming: An Online Tutorial for Medical Students”

Please make sure that the volume on your computer is turned up so that you can hear the
dialogue that accompanies these slides before proceeding. If you have trouble running
the slide show or hearing the volume, please call (517) 432-7069. Someone will be in
touch with you shortly. Do not proceed if you cannot hear the dialogue.

The following presentations are focused on an issue that is critical to many medical
students: how to plan for your medical career. When we talk about career planning,
we’re talking about ﬁnding a career that you are happy with — including your choice of
medical specialty, work setting, and speciﬁc job. We’ll be focusing on your choice of
medical specialty and your plans for entering that specialty. Your choice of medical
specialty will to be a major component of your future life. There is evidence that when
people enter a medical specialty that doesn’t match their unique preferences, they
experience dissatisfaction and stress. However, there are steps that you can take now, in
medical school, that may help you to decide upon and prepare for entering your medical
specialty. This online tutorial will help you with that process. You’ll be thinking about
your ideal medical specialty and Ieaming about different aspects of making a decision
about which specialty to enter. You’ll be asked to think about different aspects of medical
specialties and how those will affect your level of satisfaction with the specialty you
decide upon. “I am Dr. Ann Marie Ryan, a professor of Psychology with an expertise in
career planning. I’ll be leading this online tutorial.

In the following tutorials, 1 want to provide you with some important information on
strategies that you can use to help you prepare for choosing and entering your medical
specialty. I hope to help you ﬁnd and enter a medical specialty that matches your own
personal goals, desires, and values. There will be three short tutorials. After each
tutorial, you will complete a task that will help you to think about the information
provided in the tutorial. Although each tutorial is brief, we are hoping to help you think
about how you will select a medical specialty that works for you. Remember that this
tutorial is the starting point for thinking about these issues, and that you aren’t expected
to ﬁgure everything out in one day. Even if you have already decided on a medical
specialty, this presentation can help you think about that choice in new ways. I hope to
show you that you do have the ability to make well-informed, well-thought-out decisions
and that there are techniques, tools, and resources available to help you. The three
tutorials are on the following steps that you can take to prepare you for selecting a
medical specialty that’s right for you:

Step 1: Self-appraisal: Understanding and assessing your personal ideal medical
specialty. Step 2: Gathering information: Using resources and strategies to gather
information about the characteristics of different medical specialties. Step 3: specialty
selection: Identifying a medical specialty that ﬁts with your unique preferences.

Please watch each presentation carefully and complete the tasks provided. Each tutorial
will begin with a discussion of one of the steps Ijust described . Then, you’ll hear an
example of a medical student who has completed this step. Then, I’ll talk a little more
about that step. After that, you’ll complete a brief task to help you think about the
tutorial. After you complete the task, you can click on the next tutorial. Thank you.

Let’s begin Tutorial 1: Self-appraisal.
The ﬁrst step in preparing to select a medical specialty is called “self appraisal”.

Self-appraisal refers to assessing yourself to understand your own ideal medical specialty.
This means taking the time to think about and analyze your preferred job tasks, work
environment, and job rewards and challenges. Different individuals, based on their own
unique preferences, will look for different things in a medical specialty. By
understanding your unique preferences, you can begin to understand what your ideal
medical specialty looks like.

Existing research can guide your thinking about ﬁnding a specialty that suits your
preferences. Researchers have found three major ways in which medical specialties
differ: job tasks, work environment, and rewards and challenges. You can appraise your
own preferences for a medical specialty based on these characteristics. With regard to
job tasks, research suggests that individuals who have the opportunity to perform their
preferred tasks on-the-job are more satisﬁed. Job tasks include anything you actually
“do” on the job. Including performing complicated procedures, talking with patients, or
supervising others. Additionally, when individuals work in an environment that matches
their preferences, they ﬁnd their jobs more fulﬁlling. For example, characteristics of the
work environment include whether the environment is faced-paced, competitive, or
noisy. Finally, jobs offer different rewards and pose different challenges for doctors.
Research also shows that when the rewards of one’s specialty matches what individuals
desire and has challenges that are tolerable, they will tend to be happier on the job.
Examples of rewards include a high salary or the opportunity to help others, while a
challenge would be having to face losing patients.

The ﬁrst step in the self-appraisal process is to appraise your preferences. Here, we focus
on the example of job rewards. Do you want to earn a high salary? Do you want to help
the less fortunate? Do you want to build relationships with your colleagues? Do you
want job security? Different individuals will have different preferences regarding the
rewards of the job that they want. The second step is to prioritize those preferences. Of
course, in an ideal world, you would get everything you wanted in terms of job rewards.

143

Since that’s not realistic, it’s critical to pick just a few characteristics that are your top
priorities. After you identify all of your preferences, select a few preferences that are the
most important to you. You will use these preferences when you think about which
specialty matches your preferences.

There are several tools that can assist you in the self-appraisal process.

You could record and track your preferences in a journal and evaluate their importance to
you. You could ﬁll out questionnaires or surveys that assess your preferences. These
surveys are often available online or through career services and can be scored to provide
feedback to you. Additionally, you can have a conversation with a career counselor in
which you discuss your preferences and identify your priorities. He or she may be able to
ask thought-provoking questions that help you identify your preferences.

Finally, there may be friends or family members that you can discuss these issues with.
Since they know you best, they may be able to help you think through what you want and
what is really important to you. Now, I’d like you to listen to a medical student talk
about her own self-appraisal process and how this inﬂuenced what she decided to look
for in a medical specialty. Then, I’ll describe a brief task that will help you to begin your
own self-appraisal process.

Hi, my name is Stephanie, and I ﬁnished medical school in May of 2006. When I was
deciding what medical specialty to enter, I decided to go through the self-appraisal
process. Basically, I sat down at my computer and just started writing about my ideal
job. I wrote down all the things I could think of -— what I wanted to do, what mattered to
me most, what negatives I could tolerate, the type of setting I wanted to work in. Then, I
began to prioritize that list. I realized that one thing that matters a lot to me is to have a
variety in my life. Thinking about this made me realize that I want a medical specialty
where I can have a lot of variety in my work and work on unique and interesting cases.
That way, I can stay excited about my job and always look forward to doing something
unique or exciting at work. I think it would be hard for me to have a more repetitive job.
So, I considered work variety to be one of my top priorities for what I want in a medical
specialty. When I was deciding among medical specialties, I paid attention to whether
the jobs tended to have a lot of variety or whether the doctors tended to repeat the same
tasks primarily. I learned that jobs like Obstetrics have less variety because you are
primarily doing the same work with each patient, but jobs in Internal Medicine have more
variety since you see all kinds of cases. So, I deﬁnitely weighed that factor when
deciding to go into Internal Medicine. Of course, it wasn’t the only factor that I
considered, but it was good to know that I was entering a job that matched my
preferences.

Stephanie provided a great example about how she thought about her personal
preferences for her medical specialty. She thought about how her ideal medical specialty
would allow her to have variety. So you see, there are deﬁnitely ways to think about
what you want in a medical specialty. Many medical students feel that they don’t know

144

how to make decisions about their medical specialty. Self-appraisal is the ﬁrst step to
making good decisions. Once you understand your own preferences, then, you can think
about what career ﬁts with your preferences. Self-appraisal is something that any
medical student can do. You probably do some self-appraisal already. You should feel
conﬁdent that you are capable of engaging in self-appraisal about your ideal medical
specialty. Even if you’re not certain how you want things to be, you still know yourself
and have some idea about your preferences. Just by taking out a little bit of time to think
about these things, you can start the process of making decisions that will help you select
a medical specialty that works for you.

Your next step is to close the presentation and complete the task provided. This task
should start you on the road to engaging in more deep self-appraisal. The purpose of this
task is to help you engage in self-appraisal by thinking about your preferences: your
preferred job tasks, work environment, rewards, and challenges. It is important for you to
understand how your unique preferences inﬂuence the type of job that will work for you.
Now, please close this presentation and complete the task. When you are ﬁnished, you
can begin Tutorial 2.

Tutorial 1 Tasks:

1. In the space provided, spend a minute or two brain-storming about your preferences
for characteristics of your medical specialty. Write down brief phrases that come to
mind. It could be something like “work with my hands” or “pays well”.

2. From the list you just completed, write down the one preference that is the most
important to you.

3. Write a sentence or two about why it is important to you.

145

TUTORIAL 2
Let’s begin Tutorial 2: Gathering information.

Gathering information refers to gathering information about different medical specialties
to learn about whether they will ﬁt your preferences. This means taking the time to
gather the necessary information from relevant sources about different medical
specialties. As discussed in the previous tutorial, different medical specialties involve
different job tasks. Different specialties also involve different work environments.
Finally, different medical specialties offer different rewards and different challenges. The
ways in which medical specialties differ will mean that some will match with your
preferences whereas others will not. There is lots of different information that you can
gather about medical specialties. This information will eventually help you decide
whether the specialty is right for you.

Research suggests that you can beneﬁt from gathering information about different
medical specialties.Here are some examples of the types of information you can gather
about the job tasks, work environment, rewards, and challenges of different medical
specialties. For example, with regard to job tasks, you can ﬁnd out whether the specialty
requires you to perform long medical procedures. You can also gather information about
work environment in the specialty, because medical specialties take place in different
work environments. For example, you can ﬁnd out whether the specialty is faced-paced.
Finally, you can learn about the rewards and challenges of different medical specialty.
For example, you could Ieam about the salaries in different? medical specialties. By
gathering information about the speciﬁc demands of different medical specialties, you
can Ieam more about how they will ﬁt with your preferences.

There are a number of different ways to gather information.

For example, you could use the intemet or books to gather information about different
medical specialties and their jobs tasks, work environment, rewards and challenges. You
could also talk to doctors about the characteristics of their medical specialty.
Additionally, when you are on rotation in different medical specialties during medical
school, you can think about how the characteristics of the different medical specialties
and whether they will match your preferences. Additionally, you can talk to a career
counselor who may help you gain insight into the characteristics of different medical
specialties. Now, I’d like you to listen to a medical student talk about his own
information gathering process. Then, I’ll describe a brief task that will help you to begin
your own information gathering process.

Hey, I’m Tony. I’m doing my residency in general surgery. I can tell you a little bit
about the process that I went through to Ieam about different medical specialties while I

146

was in med school. I did a lot of different things to get the information that I needed. I
read a lot of articles that I found online about different medical specialties. There were
tons of helpful statistics about the work hours and salary of different specialties. So, that
was a good ﬁrst step. But, then I realized that I needed more information. So, when I
was doing my rotations in med school, I tried to observe different doctors and what their
work life was like. I looked at the different job duties that they had, the kind of situation
that they worked in, and how rewarding they found their jobs. I also talked to the doctors
in different specialties and speciﬁcally asked them about the different aspects of their job.
This helped me even more. After listening to them, I realized that different specialties
are so different from one another. I was able to weigh the information that I gathered
when making my decision to enter surgery. I found that the characteristics of surgery
really matched my preferences.

Tony provided an example of how he was able to gather information about different
medical specialties. By doing research, observing doctors, and speaking with doctors, he
was able to gather valuable information. There are very straightforward ways to gather
information that will help you Ieam about different medical specialties. Medical students
may feel that the information that they need about different specialties isn’t available or
are shy about asking questions of doctors, professors, and career counselors. However,
you really do have the ability to gather lots of information that will help you make an
informed decision. You can gather tons of information from a number of different
sources. You’ve probably already begun to gather information about different medical
specialties. You should feel conﬁdent that you are capable of engaging in information
gathering. By using some or all of the resources available to you, you can start to gather
information that will help you pick a specialty that works for you.

Your next step is to close this presentation and complete the task provided. This should
start you on the road to engaging in a more thorough information gathering process. The
purpose of this task is to help you start thinking about the type of information that you
would like to gather and the sources that you can get that information from. Now, please
close this presentation and complete the task. When you are ﬁnished, you can begin
Tutorial 3.

Tutorial 2 Tasks:
1. Write down two questions that you would want to ask about the characteristics of
different medical specialties.
2. For each of those questions, write down one source that you could use to gather that
information. -

TUTORIAL 3

Let’s begin the last tutorial: Identifying specialties that “ﬁt”

147

Identifying ﬁt refers to identifying a specialty that will ﬁt with your preferences. This
means using the knowledge that you gained through self appraisal and the information
you gathered about different specialties to identify a few specialties that provide the best
ﬁt to your preferences. Identifying a specialty that ﬁts with your preferences is an
important step in preparing for your future medical career. After engaging in self-
appraisal, you know what you’re looking for. By gathering information, you know what
the specialties have to offer. When you identify ﬁt, you see which specialties offer what
you’re looking for.

Research suggests that when an individual selects a career that ﬁts with their preferences,
they experience greater satisfaction and higher performance. You can examine whether a
specialty matches your preferences. The example provided here examines whether
emergency medicine and obstetrics are a good ﬁt with the preferences of a medical
student. This medical student wants to have the opportunity to do a number of different
unique procedures and to be able to become familiar with his patients. Emergency
medicine offers the opportunity to perform unique procedures, so that makes it a good ﬁt.
However, it does not offer the opportunity for doctors to become very familiar with their
patients, so this bad ﬁt. On the other hand, Obstetrics requires the doctor to perform
mainly routine procedures, so this is a bad ﬁt, but it has medium levels of patient
familiarity (since patients become familiar over a 9 month period, not usually a lifetime),
so this is an okay ﬁt. Overall, it looks like Emergency Medicine may offer a slightly
better ﬁt to the preferences of the medical student in this example. This example should
help you see how you can assess the ﬁt of different medical specialties with your
preferences. Probably no specialty will be perfect, but some may ﬁt your preferences and
goals better than others.

I suggest that when you are evaluating whether a specialty ﬁts your preferences, you
create a table like the one on the previous slide. Write your top preferences and what the
specialty you are considering offers in terms of that preference. Then, write down
whether it ﬁts or doesn’t ﬁt with each preference. This will help you to determine
whether a particular specialty ﬁts with your preferences. You can use this table to
compare multiple specialties to one another, similar to the previous slide. Many medical
students have found that creating a table such as this is worthwhile. It will help you
select a medical specialty that matches your preferences. Now, I’d like you to listen to a
medical student talk about her own process of identifying a specialty that ﬁt her
preferences. Then, I’ll describe a brief task that will help you to begin your own process
of thinking about a specialty that ﬁts your preferences.

Hi, my name is Ruchi. I’m doing my residency in Pediatrics. I want to tell you a little bit
about how I decided that Pediatrics was a good ﬁt for me, in terms of how it matches my
preferences. When I was deciding what specialty to enter, I created a table like the one
suggested in this tutorial. I decided to focus on the two things that I decided were most
important to me in a medical specialty. I wanted to feel like I was really helping people

148

and I wanted a specialty that utilized my communication skills. So, I picked my top three
choices of specialty and analyzed how well they each matched my preferences. I found
that pediatrics was the best ﬁt for me. It ended up going into Pediatrics partially for this
reason. It also ﬁts well with my other interests. I deﬁnitely considered which specialty
matched my preferences most closely when I was deciding on a specialty.

Ruchi provided an example of how she was able to identify a specialty that ﬁt with her
preferences. By matching the preferences that you identify through self-appraisal to the
characteristics of the specialties that you identify through information gathering, you can
ﬁnd a specialty that will work for you. Finding a good match is something that every
medical student can do. In fact, once you’ve thought about your own preferences and
learned about the different specialties, you may already begin to have an idea of what
specialties will ﬁt best for you. Once you’ve engaged in self-appraisal and gathered
information, ﬁguring out the specialty that ﬁts best is straight-forward. You have the
capacity to identify a medical specialty that ﬁts your goals and preferences.

Your next step is to close this presentation and complete the task provided. This task is
just to help you start thinking about identifying ﬁt. You’ll need more time for self-
appraisal and gathering information before you’ll really be able to identify a good ﬁt.
Now, please close this presentation and complete the task. When you are ﬁnished, you
will be asked to complete the remainder of the survey.

Tutorial 3 Tasks:
1. Write down one preference that you have for characteristics of your medical specialty.
2. Write down two medical specialties that you are interested in.
3. Write down how well you think each of those specialties matches your preference.
4. Identify which of those specialties you think provides the better ﬁt to that preference.

149

Appendix H
Powerpoint slides for work-family condition including tasks

INTRODUCTION AND TUTORIAL 1

Welcome to “Balancing Work and Family: An Online Tutorial for Medical Students”

Please make sure that the volume on your computer is turned up so that you can hear the
dialogue that accompanies these slides before proceeding. If you have trouble running
the slide show or hearing the volume, please call (517) 432-7069. Someone will be in
touch with you shortly. Do not proceed if you cannot hear the dialogue.

The following presentations are focused on an issue that is critical to many medical
students: how to balance your career as a doctor with your family life. When we talk
about work-family balance, we’re talking about ﬁnding a balance that you are happy with
between the medical career that you decide on and your family life. Your work obviously
places demands on your time and energy. So does your family life. When we talk about
your family, we’re talking about your spouse or partner, your children, and sometimes
caring for elderly parents. There is evidence that when people don’t feel like they have a
good balance between their work lives and their family lives they experience
dissatisfaction and stress. However, there are steps that you can take now, in medical
school, that may help you to prepare for balancing work and family after medical school.
This online tutorial will help you with that process. You’ll be thinking about your ideal
work-family balance and how to pick a medical specialty that will help you balance work
and family. You’ll be asked to think about different aspects of work-family balance and
how those will affect which specialty you decide upon. “I am Dr. Ann Marie Ryan, a
professor of Psychology with an expertise in work-family balance. I’ll be leading this
online tutorial.

In the following tutorials, I want to provide you with some important information on
strategies that you can use to help you prepare for choosing and entering your medical
specialty in a way that will allow you to balance your work life with your family life. I
hope to help you ﬁnd a balance between work and family that meets your own personal
goals, desires, and values. There will be three short tutorials. After each tutorial, you
will complete a task that will help you to think about the information provided in the
tutorial. Although each tutorial is brief, we are hoping to help you think about how you
will balance your medical career with your family life. Remember that this tutorial is the
starting point for thinking about these issues, and that you aren’t expected to ﬁgure
everything out in one day. I hope to show you that you do have the ability to make well-
informed, well-thought—out decisions and that there are techniques, tools, and resources
available to help you. The three tutorials are on the following steps that you can take to
prepare you for selecting a medical specialty that’s right for you:

150

Step I: Self-appraisal: Understanding and assessing your personal ideal balance between
your medical career and your family life.

Step 2: Gathering information: Using resources and strategies to gather information about
the characteristics of different medical specialties and how they affect work-family
balance.

Step 3: specialty selection: Identifying a medical specialty that ﬁts with your unique
preferences for balancing work and family.

Please watch each presentation carefully and complete the tasks provided. Each tutorial
will begin with a discussion of one of the steps I just described . Then, you’ll hear an
example of a medical student who has completed this step. Then, I’ll talk a little more
about that step. Aﬁer that, you’ll complete a brief task to help you think about the
tutorial. After you complete the task, you can click on the next tutorial. Thank you.

Let’s begin Tutorial 1: Self-appraisal.
The ﬁrst step in preparing to balance work and family is called “self appraisal”.

Self-appraisal refers to assessing yourself to understand your own ideal balance between
your medical career and your family life. This means taking the time to think about and
analyze how you envision your work-life, your family-life and your balance between
work and family. Different individuals, based on their own unique preferences, will
envision different lifestyles. By understanding your unique preferences for balancing

work and family, you can begin to understand what your ideal medical specialty looks
like.

Existing research can guide your thinking about ﬁnding a specialty that suits your
preferences for how you want to balance work and family. Researchers have found three
major ways in which work interferes with family-life: Time, Stress, and Behaviors. You
can appraise your own preferences for a medical specialty based on how it inﬂuences
your time, stress, and behaviors. With regard to time, research suggests the time
demands of work can interfere with time available for family. For example, hours of
work and hours on call both inﬂuence time available for family. Additionally, when you
can’t predict or control your work hours, it creates challenges for your family life.

With regard to stress, research suggests that the stress of work can interfere with family-
life because people often bring the stress of their work home with them, which negatively
inﬂuence their family life. For example, if you work in an extremely fast-paced
environment or have a lot of work overload, you may come home physically and
psychologically exhausted. If you are making life-or-death decisions at work, you might
bring the weight of those decisions home with you. Finally, with regard to behaviors,
research suggests that when people commonly act in ways that are effective and
necessary at work, they may have difﬁculty “shutting off” those behaviors at home. For
example, someone who has to act very authoritative at work may have a greater tendency
to act authoritative at home — which may create challenges for family life. Additionally,

151

behaviors such as following strict protocols or maintaining professionalism at work may
inﬂuence family-life.

The ﬁrst step in the self-appraisal process is to appraise your preferences. Here, we focus
on the example of the time demands of your medical specialty. What are your
preferences for the time-demands of your job? How many hours a week do you want to
work? Do you want a predictable schedule? Do you mind being on call at night? Do you
want a ﬂexible schedule? Different individuals will have different preferences for the
time demands of their job. The second step is to prioritize those preferences. Of course,
in an ideal world, you would get everything you wanted in terms of the time-demands of
your job. Since that’s not realistic, it’s critical to pick just a few characteristics of your
job that are your top priorities in terms of balancing work and family. After you identify
all of your preferences, select a few preferences that are the most important to you. You
will use these preferences when you think about which specialty matches your
preferences.

There are several tools that can assist you in the self-appraisal process.

You could record and track your preferences in a journal and evaluate their importance to
you.

You could ﬁll out questionnaires or surveys that assess your preferences. These surveys
are often available online or through career services and can be scored to provide
feedback to you. Additionally, you can have a conversation with a career counselor in
which you discuss your preferences and identify your priorities. He or she may be able to
ask thought-provoking questions that help you identify your preferences.

Finally, there may be friends or family members that you can discuss these issues with.
Since they know you best, they may be able to help you think through what you want and
what is really important to you. Now, I’d like you to listen to a medical student talk
about her own self-appraisal process and how this inﬂuenced what she decided to look
for in a medical specialty. Then, I’ll describe a brief task that will help you to begin your
own self-appraisal process.

Hi, my name is Stephanie, and I ﬁnished medical school in May of 2006. When I was
deciding what medical specialty to enter, I decided to go through the self-appraisal
process. Basically, I sat down at my computer and just started writing about my ideal job
in terms of how it would inﬂuence my work-family balance. I wrote down all the things I
could think of — the work hours, the stress, everything. Then, I began to prioritize that
list. I realized that one thing that matters a lot to me is to be is to not get too stressed out
on the job because 1 always tend to take my stress home with me and let it affect other
aspects of my life. Thinking about this made me realize that I want a medical specialty
where there are fewer emergencies and less on-the-job stress and time pressure. I think
that it would be really stressful for me to have to deal with life-or-death situations on a
daily basis and then to try to go home to be a family member. So, I considered lower job
stress as one of my top priorities for how I want to balance work and family. When I was

152

deciding among medical specialties, I paid attention to whether the specialties had really
high on—the-job stress. I learned that jobs like Emergency Medicine have really high on-
the-job stress, but jobs in Internal Medicine have less on-the-job stress because most
patients aren’t facing an emergency. So, I deﬁnitely weighed that factor when deciding
to go into Internal Medicine. Of course, it wasn’t the only factor that I considered, but it
was good to know that I was entering a job that matched my preferred way of balancing
work and family.

Stephanie provided a great example about how she thought about her personal
preferences for balancing work and family. She thought about how her ideal medical
specialty would allow her to plan ahead for balancing work and family. So you see, you
can engage in the self-appraisal process now to begin to understand how you want to
balance work and family. Many medical students feel that they don’t know how to make
decisions about balancing work and family. Self-appraisal is the ﬁrst step to making
good decisions. Once you understand your own preferences, then, you can think about
what career and family lifestyles ﬁt with your preferences. Self-appraisal is something
that any medical student can do. You probably do some self-appraisal already. You
should feel conﬁdent that you are capable of engaging in self-appraisal about your ideal
balance between work and family. Even if you’re not certain how you want things to be,
you still know yourself and have some idea about your preferences. Just by taking out a
little bit of time to think about these things, you can start the process of making decisions
that will help you balance work and family in the future.

Your next step is to close the presentation and complete the task provided. This task
should start you on the road to engaging in more deep self-appraisal. The purpose of this
task is to help you engage in self-appraisal by thinking about how you want to balance
work and family. It is important for you to understand how your unique preferences
inﬂuence the type of job that will help you to balance work and family in a way that
meets your needs. Now, please close this presentation and complete the task.

When you are ﬁnished, you can begin Tutorial 2.

Tutorial 1 Tasks:

I. In the space provided, spend a minute or two brain-storming about your preferences
for balancing work and family. Write down brief phrases that come to mind. It could
be something like “home for dinner” or “not exhausted at the end of the day”.

2. From the list you just completed, write down the one preference that is the most
important to you in terms of how you want to balance work and family.

3. Write a sentence or two about why it is important to you.

153

TUTORIAL 2
Let’s begin Tutorial 2: Gathering information.

Gathering information refers to gathering information about different medical specialties
to Ieam about the demands that different medical specialties place on your ability to
balance work and family. This means taking the time to gather the necessary information
from relevant sources about different medical specialties.You can gather information
about different medical specialties and how they inﬂuence work-family balancing by
learning about the time, stress, and behavioral demands of different medical specialties.
As discussed in the previous tutorial, different medical specialties place different
demands on your time. They also place different amounts and types of stress on doctors.
Finally, different medical specialties also have require different types of behaviors for
effective performance. The ways in which medical specialties differ will mean that some
will match with your preferences for balancing work and family whereas others will not
match with your ideal balance between work and family. There is lots of different
information that you can gather about medical specialties. This information will help you
decide which specialty is right for you.

Research suggests that you can beneﬁt from gathering information about different
medical specialties. Here are some examples of the types of information you can gather
about the time, strain, and behavioral demands of different medical specialties. With
regard to time-based issues, you can ﬁnd out what the average work hours are for
different specialties or how predictable the hours are. You can Ieam about the stress that
different specialties place on doctors. For example, you could learn how often you have
to face medical emergencies in a particular specialty. Finally, you can also gather
information about the behavioral demands of different specialties. For example, you can
ﬁnd out whether different specialties require you to act in speciﬁc ways such as
dispassionate, authoritative, or decisive. By gathering information about the speciﬁc
demands of different medical specialties, you can Ieam more about how they will affect
your work-family balance and whether they will work for you.

There are a number of different ways to gather relevant information.

For example, you could use the intemet or books to gather information about different
medical specialties. You could also talk to doctors about the demands of their medical
specialty and how those demands inﬂuence their ability to balance work and family.
Additionally, when you are on rotation in different medical specialties during medical
school, you can think about how the demands of the specialty would allow you to balance
work and family if you were to enter that ﬁeld full time. Additionally, you can talk to a
career counselor who may help you gain insight into the demands of different medical
specialties. Now, I’d like you to listen to a medical student talk about his own
information gathering process. Then, I’ll describe a brief task that will help you to begin
your own information gathering process.

154

Hey, I’m Tony. I’m doing my residency in general surgery. I can tell you a little bit
about the process that I went through to Ieam about how different medical specialties
would inﬂuence my work-family balance while I was in med school. I did a lot of
different things to get the information that I needed. I read a lot of articles that I found
online about different medical specialties and ordered a book about medical specialties.
There were tons of helpful statistics about the time and salary of different specialties. So,
that was a good ﬁrst step. But, then I realized that I needed more information about how
doctors in different specialties manage to have a medical career and a family life. So,
when I was doing my rotations in med school, I tried to observe different doctors and the
daily lives that they led. I looked at how stressed out they were, how tired they seemed,
whether they were on call on the weekends. I also talked to the doctors in different
specialties and speciﬁcally asked them about their family life. This helped me even
more. After listening to them, I realized that different specialties are so different in terms
of the lifestyle that you get to lead outside of work. I was able to weigh the information
that I gathered when making my decision to enter surgery. I found that the characteristics
of surgery really matched my preferences for how I want to balance work and family.

Tony provided an example of how he was able to gather information about different
medical specialties and how they inﬂuence work-family balance. By doing research,
observing doctors, and speaking with doctors, he was able to gather valuable information.
There are very straightforward ways to gather information that will help you Ieam about
balancing work and family in different medical specialties. Medical students may feel
that the information that they need about balancing work and family isn’t available or are
shy about asking questions of doctors, professors, and career counselors. However, you
really do have the ability to gather lots of information that will help you make an
informed decision. You can gather tons of information from a number of different
sources. You’ve probably already begun to gather information about how different
medical specialties will affect your personal life. You should feel conﬁdent that you are
capable of engaging in information gathering. By using some or all of the resources
available to you, you can start to gather information that will help you balance work and
family in the future.

Your next step is to close this presentation and complete the task provided. This should
start you on the road to engaging in a more thorough information gathering process. The
purpose of this task is to help you start thinking about the type of information that you
would like to gather and the sources that you can get that information from. Now, please
close this presentation and complete the task. When you are ﬁnished, you can begin
Tutorial 3.

Tutorial 2 Tasks:

1. Write down two questions that you would want to ask about the work-family
demands of different medical specialties.

155

2. For each of those questions, write down one source that you could use to gather that
information.

TUTORIAL 3
Let’s begin the last tutorial: Identifying specialties that “ﬁt”

Identifying ﬁt refers to identifying a specialty that will ﬁt with your work and family
goals and preferences. This means using the knowledge that you gained in self appraisal
and the information you gathered about different specialties to identify a few specialties
that provide the best ﬁt to your work-family preferences. Identifying a specialty that ﬁts
with your work-family goals is an important step in preparing for your future medical
career and family life. After engaging in self-appraisal, you know what you’re looking
for. By gathering information, you know what the specialties have to offer. When you
identify ﬁt, you see which specialties offer what you’re looking for.

Research suggests that when an individual selects a career that ﬁts with their preferences,
they experience greater satisfaction and higher performance. You can examine whether a
specialty matches your preferences for how you want to balance work and family. The
example provided here examines whether emergency medicine and obstetrics are a good
ﬁt with the work-family preferences of a medical student. This medical student wants
predictable hours and low jobs stress. Emergency medicine offers predictable hours, so
that makes it a good ﬁt. However, it has high stress on-the-job, so that makes it a bad ﬁt.
On the other hand, Obstetrics has unpredictable hours, so this is a bad ﬁt, but it has
medium levels of on-the-job stress, so this is an okay ﬁt. Overall, it looks like
Emergency Medicine may offer a slightly better ﬁt to the work-family preferences of this
medical student in this example. This example should help you see how you can assess
the ﬁt of different medical specialties with your work-family preferences. Probably no
specialty will be perfect, but some may ﬁt your preferences and goals better than others.

I suggest that when you are evaluating whether a specialty ﬁts your work and family
preferences, you create a table like the one on the previous slide. Write your preferences
for balancing work and family and what the specialty you are considering offers in terms
of that preference. Then, write down whether it ﬁts or doesn’t ﬁt with each preference.
This will help you to determine whether a particular specialty ﬁts with your preferences.
You can use this table to compare multiple specialties to one another, similar to the
previous slide. Many medical students have found that creating a table such as this is
worthwhile. Although you may not enter the specialty that has the best ﬁt with your
work and family goals because of other reasons, it is still helpful to understand which
specialty ﬁts your work-family preferences best so that you can weigh that information
when making your decision. Now, I’d like you to listen to a medical student talk about
her own process of identifying a specialty that ﬁt her work-family preferences. Then, 1’11

156

describe a brief task that will help you to begin your own process of thinking about a
specialty that ﬁts your work and family preferences.

Hi, my name is Ruchi. I’m doing my residency in Pediatrics. I want to tell you a little bit
about how I decided that Pediatrics was a good ﬁtfor me, in terms of how it will allow
me to have my career and my family life at the same time. When I was deciding what
specialty to enter, I created a table like the one suggested in this tutorial. I decided to
focus on the two things that I decided were most important to me in terms of how I
wanted to balance work and family. I wanted a ﬂexible working schedule and a job that
wasn’t too stressful on a daily basis. So, I picked my top three choices of specialty and
analyzed how well they each matched my preferences. I found that pediatrics was the
best ﬁt for me in terms of work-family balance. It ended up going into Pediatrics
partially for this reason. It also ﬁts well with my interests. I deﬁnitely considered which
specialty was the best ﬁt for my preferred work and family situation when I made my
decision.

Ruchi provided an example of how she was able to identify a specialty that ﬁt with her
work-family preferences. By matching the work-family preferences that you identify
through self-appraisal to the characteristics of the specialties that you identify through
information gathering, you can ﬁnd a specialty that will work for you. Finding a good
match is something that every medical student can do. In fact, once you’ve thought about
your own preferences and learned about the different specialties, you may already begin
to have an idea of what specialties will ﬁt best for you. Once you’ve engaged in self-
appraisal and gathered information, ﬁguring out the specialty that ﬁts best is straight-
forward. You have the capacity to identify a medical specialty that ﬁts your goals and
preferences for balancing work and family.

Your next step is to close this presentation and complete the task provided. This task is
just to help you start thinking about identifying ﬁt. You’ll need more time for self-
appraisal and gathering information before you’ll really be able to identify a good ﬁt.
Now, please close this presentation and complete the task. When you are ﬁnished, you
will be asked to complete the remainder of the survey.

Tutorial 3 Tasks:
1. Write down one preference that you have for how you want to balance work and
family.
2. Write down two medical specialties that you are interested in.
3. Write down how well you think each of those specialties matches your preference.
Identify which of those specialties you think provides the better ﬁt to that preference.

157

Appendix I
Anticipated Work-family Conﬂict Items

Instructions: Thinking about your future job in medicine and the family that you plan to
have in the future, please indicate whether you agree or disagree with the following
statements. Your ratings should range from 1 (showing that you strongly disagree with
the statement) to 5 (showing that you strongly agree with the statement).

1 = Strong disagree

2 = Disagree
3 = Neither agree nor disagree
4=Agw

5 = Strongly agree

Time-based work-interference-with-family

1. My work will keep me from my family activities more than I would like.

2. The time I will devote to my job will keep me from participating equally in household
responsibilities and activities.

3. I will have to miss family activities due to the amount of time I will have to spend on
work responsibilities.

Time—based family-interference-with-work

1. The time I will spend on family responsibilities will often interfere with my work
responsibilities.

2. The time I will spend with my family will often cause me not to spend time in
activities at work that could be helpful to my career. ‘

3. I will have to miss work activities due to the amount of time I will have to spend on
family responsibilities.

Strain-based work-interference-with-family

1. I think that when I get home from work I will often be too frazzled to participate in
family activities/responsibilities.

2. 1 Will often be so emotionally drained when I get home from work that it will prevent
me from contributing to my family.

3. Due to all the pressures I will have at work, sometimes when I get home I will be too
stressed to do the things I enjoy.

Strain-based family—interference-with—work

1. Due to stress at home, I will often be too preoccupied with family matters at work.

2. Because I will often be stressed from my family responsibilities, 1 will have a hard
time concentrating on my work.

3. Tension and anxiety from my family life will often weaken my ability to do my job.

Behavior-based work-interference-witlr-family

1.

2.

3.

The problem-solving behaviors I will use in my job will not be effective in resolving
problems at home.

Behavior that is effective and necessary for me at work will be counterproductive at
home.

The behaviors that I will perform that will make me effective at work will not help
me to be a better parent and spouse/partner.

Behavior-based family-interference-with-work

1.
2.

3.

The behaviors that will work for me at home will not be effective at work.

Behavior that is effective and necessary for me at home will be counterproductive at
work.

The problem-solving behavior that will work for me at home will not be as useﬁrl at
work.

159

Appendix J
Anticipated Work-Family Conﬂict Item Intercorrelations

Table J1
Anticipated Work-Family Conﬂict Item Intercorrelations for Time I

 

Mean sd 1 2 3 4 5 6 7 8

 

TWIF Item 1 3.41 0.88

TWIF Item 2 3.28 0.96 .60

TWIF Item 3 3.50 0.85 .57 .42

TFIW Item 1 2.88 0.87 .24 .24 .37

TFIW Item 2 3.16 0.93 .21 .17 .26 .41

TFIW Item 3 3.12 0.87 .16 .07 .21 .42 .62

swrr Item 1 2.35 0.92 .25 .22 .31 .24 .27 .16
SWIFItem2 2.42 1.01 .28 .23 .31 .30 .32 .17 .69
SWIFItem3 2.94 1.15 .28 .11 .28 .21 .30 .24 .52 .63
SFIWIteml 2.27 0.89 ‘.14 .03 .18 .35 .19 .26 .54 .64
SFIWItem2 2.25 0.88 .08 .02 .11 .34 .22 .33 .57 .59
SFIWItem3 2.18 0.89 .00 .07 .06 .27 .22 .23 .53 .65
BWIFIteml 2.43 1.03 .09 .10 .02 -.04 -.01 .04 .24 .27

BWIF Item 2 2.44 0.96 .05 .16 .14 .09 .07 .15 .35 .38
BWIF Item 3 2.40 0.99 .11 .19 -.05 .02 .01 .05 .27 .29
BFIW Item 1 2.33 0.80 .04 .22 .01 .13 .03 .16 .34 .39
BFIW Item 2 2.31 0.78 .10 .28 .18 .18 .10 .19 .36 .44
BFIW Item 3 2.27 0.81 06 22 .06 .17 .07 21 30 .42

 

Notes. N = 104 to 108. Correlations with r 21.19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01. TWIF = time-based work-
interference-with-family. TFIW = time-based family-interference-with-work.
SWIF = strain-based work-interference—with-family. SF IW = strain-based
family-interference-with-work. BWIF = behavior-based work-interference—with-
family. BFIW = behavior-based family-interference-with-work.

160

Table J1, continued

Anticipated Work—Family Conﬂict Item Intercorrelations for Time I
9 10 l 1 12 l3 14 15 16 17

 

TWIF Iteml

TWIF Item2

TWIFItem3

TFIW Iteml

TFIW Item2

TFIW Item3

SWIF Iteml

SWIFItem2

SWIF Item3

SFIW Iteml .59

SFIW Item 2 .58 .85

SFIWItem3 .53 .78 .83

BWIFIteml .17 .24 .30 .32

BWIFItem2 .29 .43 .45 .46 .65

BWIF Item3 .22 .32 .33 .34 .49 .53
BFIWIteml .32 .38 .46 .49 .57 .69 .62
BFIWItem2 .26 .36 .43 .46 .67 .75 .56 .84
BFIWItem3 .31 .34 .40 .44 .59 .69 .54 .76 .83

 

161

Table J2
Anticipated Work-Family Conﬂict Item Intercorrelationsfor Time 2

 

Mean sd 1 2 3 4 5 6 7 8
TWIF Item 1 3.22 0.98
TWIF Item 2 2.98 0.98 .64
TWIF Item 3 3.15 0.98 .74 .61
TFIW Item 1 2.71 0.91 .41 .37 .46
TFIW Item 2 3.03 0.98 .48 .44 .49 .48
TFIW Item 3 2.80 0.92 .45 .26 .41 .60 .64
SWIF Item 1 2.35 0.95 .30 .36 .37 .31 .24 .31
SWIFItem2 2.33 0.92 .34 .35 .39 .37 .25 .32 .90
SWIFItem3 2.76 1.04 .39 .42 .47 .23 .27 .26 .67 .65
SFIWIteml 2.26 0.78 .17 .21 .24 .38 .09 .33 .65 .61
SFIWItem2 2.24 0.77 .26 .27 .28 .42 .11 .31 .73 .65
SFIWItem3 2.14 0.80 .16 .11 .18 .33 .11 .22 .64 .63
BWIFIteml 2.32 0.85 .19 .24 .15 .27 ..14 .25 .38 .48

BWIF Item 2 2.35 0.88 .12 .25 .11 .25 .15 .24 .37 .38
BWIF Item 3 2.48 0.94 .14 .30 .07 .16 .28 .22 .25 .30
BFIW Item 1 2.30 0.85 .19 .32 .17 .20 .22 .21 .34 .39
BFIW Item 2 2.29 0.88 .20 .33 .15 .20 .21 .23 .34 .38
BF .IW Item 3 2.29 0.86 18 34 18 28 18 24 41 .44

 

Notes. N = 104 to 108. Correlations with r 2 F19 l are signiﬁcant at p < .05.
Correlations with r 2 | .25 I are signiﬁcant at p < .01. TWIF = time-based work-
interference-with-family. TFIW = time-based family-interference-with-work.
SWIF = strain-based work-interference-with-fami1y. SFIW = strain-based
family-interference-with-work. BWIF = behavior-based work-interference-with-
family. BFIW = behavior-based family-interference-with-work.

162

Table J2, continued
Anticipated Work-Family Conﬂict Item Intercorrelations for Time 2

 

9 10 11 12 l3 14 15 I6 17

 

TWIF Item 1

TWIF Item 2

TWIF Item 3

TFIW Item 1

TFIW Item 2

TFIW Item 3

SWIF Item 1

SWIF Item 2

SWIF Item 3

SFIW Item 1 .48

SF 1W Item 2 .52 .84

SFIW Item 3 .45 .72 .79

BWIF Item 1 .34 .40 .43 .43

BWIF Item 2 .33 .42 .45 .37 .81

BWIF Item 3 .29 .33 .33 .27 .54 .60

BFIW Item 1 .30 .41 .41 .40 .74 .75 .57
BFIW Item 2 .27 .42 .41 .33 .71 .86 .60 .86
BFIW Item 3 .29 .40 .47 .41 .87 .83 .55 .79 .77

 

I63

Table J3
Anticipated Work-Family Conﬂict Item Intercorrelations for Time 3

 

Mean sd 1 2 3 4 5 6 7 8
TWIF Item 1 3.31 0.95
TWIF Item 2 3.04 1.02 .68
TWIF Item 3 3.21 0.99 .68 .63
TFIW Item 1 2.76 0.92 .48 .58 .48
TFIW Item 2 3.11 1.06 .40 .43 .37 .69
TFIW Item 3 3.02 0.97 .28 .33 .35 .64 .76
SWIF Item 1 2.31 0.89 .14 .28 .26 .40 .31 .28
SWIFItem2 2.31 0.91 .15 .24 .22 .33 .25 .25 .86
SWIFItem3 2.83 1.11 .26 .23 .25 .27 .22 .15 .71 .71
SFIWIteml 2.33 0.90 .18 .30 .26 .42 .28 .25 .62 .68
SFIWItem2 2.25 0.85 .12 .26 .19 .39 .24 .21 .64 .69
SFIWItem3 2.19 0.91 .06 .26 .18 .40 .23 .22 .62 .67
BWIFIteml 2.38 0.89 .20 .30 .19 .36 .17 .25 .37 .35

BWIF Item 2 2.32 0.82 .10 .21 .18 .33 .14 .22 .38 .36
BWIF Item 3 2.48 0.94 .06 .17 .09 .20 .19 .11 .28 .16
BFIW Item 1 2.28 0.81 .15 .27 .20 .29 .10 .15 .42 .35
BFIW Item 2 2.30 0.80 .10 .21 .17 .31 .07 .15 .31 .31
BFIW Item 3 2.41 0.83 05 22 10 31 18 18 .24 .15

 

Notes. N = 1.04 to 108. Correlations with r 2| .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01. TWIF = time-based work-
interference—with-family. TFIW = time-based family-interference-with-work.
SWIF = strain-based work-interference-with-family. SF 1W = strain-based
family-interference-with-work. BWIF = behavior-based work-interference-with-
family. BF [W = behavior-based family-interference-with-work.

164

Table J3, continued
Anticipated Work-Family Conﬂict Item Intercorrelations for Time 3

 

9 10 II 12 l3 14 15 16 I7

 

TWIF Item 1

TWIF Item 2

TWIF Item 3

TFIW Item 1

TFIW Item 2

TF 1W Item 3

SWIF Item 1

SWIF Item 2

SWIF Item 3

SFIW Item 1 .59

SFIW Item 2 .58 .92

SFIW Item 3 .57 .80 .82

BWIF Item I .32 .33 .35 .38

BWIF Item 2 .25 .38 .42 .42 .74

BWIF Item 3 .23 .26 .30 .31 .51 .47

BFIW Item 1 .37 .36 .44 .45 .72 .73 .50

BFIW Item 2 .22 .42 .50 .45 .67 .75 .40 .80
BFIW Item 3 .12 .25 .30 .27 .55 .60 .57 .69 .66

 

165

Appendix K
Anticipated Work-family Conﬂict CFAs at Each Administration

 

 

 

 

 

 

 

 

 

Table K1
Estimates of F it Indices for Anticipated Work-Family Conﬂict at Time I
36 Df P Comparative Fit Root mean square
Index (CF I) error of
approximation
(RMSEA)
6-dimensional model 185.15 126 .00 .94 .07
3-dimensional model 267.73 132 .00 .87 .10
2-dimensional model 635.04 134 .00 .52 .19
Unidimensional model 623.88 135 .00 .53 .19
N = 100
Table K2
Estimates of F it Indices for Anticipated Work-Family Conﬂict at Time 2
x2 Df P Comparative F it Root mean square
Index (CF I) error of
approximation
(RMSEA)
6-dimensional model 301.18 126 .00 .87 .12
3-dimensional model 386.89 132 .00 .82 .15
2-dimensional model 837.67 134 .00 .49 .24
Unidimensional model 707.60 135 .00 .48 .23
N = 93
Table K3
Estimates ofFit Indicesfor Anticipated Work-Family Conﬂict at Time 3
x2 Df P Comparative Fit Root mean square
Index (CFI) error of
approximation
(RMSEA)
6-dimensional model 180.14 126 .00 .95 .07
3-dimensional model 307.76 132 .00 .84 .13
2-dimensional model 643.40 134 .00 .54 .21
Unidimensional model 707.60 135 .00 .48 .23
N = 85

166

Appendix L
Certainty Regarding Medical Specialty Plans Items

1. Please indicate which medical specialty you feel that you are MOST likely to enter.
Anesthesiology
Dermatology
Emergency medicine
Family practice
Internal medicine
Medical genetics
Neurological surgery
Neurology
Nuclear medicine
Obstetrics and gynecology
Opthalmology
Orthopedic surgery
. Otolaryngology
Pathology — anatomical and clinical
My ISt choice is not listed here (see next item)

.0 a a Ferrari ewe-.0 9‘1»

2. Please indicate which specialty you feel that you are MOST LIKELY to enter.
Pediatrics

Physical medicine and rehabilitation

Plastic surgery

Preventive medicine

Psychiatry

Radiation oncology

Radiology — diagnostic

Surgery — general

Urology

Unsure

Other

I selected my ﬁrst choice in the previous question

 

Fri-retro rm .o-o 9‘1»

3. How certain are you that your medical specialty plans will actually happen?
a. Very unsure
b. Unsure
c. Neither sure nor unsure
(1. Sure

e. Very sure

167

4. How likely is it that you will change your mind about your medical specialty plans?
a. Very unlikely
b. Unlikely
c. Neither likely nor unlikely
d. Likely
e. Very likely

5. How much doubt do you have about your medical specialty plans?
a. Very little doubt
b. Little doubt
0. Neither a lot of doubt nor a little doubt
(I. Some doubt
e. A lot of doubt

168

Appendix M
Certainty Regarding Medical Specialty Plans Item Intercorrelations

 

Mean sd 1 2 3 4 5 6

 

Time 1

 

1. Item 1 3.47 1.16
2. Item2 3.14 1.16 .81

 

 

 

 

3. Item 3 2.94 1.27 .76 .74

Time 2

4. Item 1 3.49 1.03 .82 .67 .68

5. Item 2 3.08 1.16 .79 .90 .75 .69

6. Item 3 3.02 1.19 .74 .72 .89 .69 .79

Time 3

7. Item 1 3.56 1.04 .80 .70 .64 .84 .68 .66
8. Item 2 2.97 1.11 .76 .75 .68 .70 .64 .67
9. Item 3 3.03 1.21 65 61 71 .70 .61 .65

 

Notes. N ranges from 91 to 108. Correlations with r 2 I .21 I are signiﬁcant at p < .05.
Correlations with r 2 I .27 I are signiﬁcant at p < .01.

 

7 8

 

Time 1

 

1. Item 1
2. Item 2
3Jmm3

 

Time 2

 

4. Item 1
5. Item 2
6. Item 3

 

Time 3

 

7. Item 1
8. Item 2 .68
9. Item 3 .72 ' .76

 

169

Appendix N
Certainty Regarding Family Plans Items

. Do you plan to be married or have a life-long committed partnership at some
point in your life?

a. Yes

b. No

c. Maybe

(I I haven’t thought about it

. At what age do you plan to marry or enter a life-long committed partnership?
I do not plan to be married/partnered

I am already married/partnered

22 — 25 years old

25 — 30 years old

30 — 35 years old

35 years old or older

Unsure

I haven’t thought about it

P‘CP we 9.0 9‘9»

. Do you plan to have children at some point in your life?

a. Yes

b. No

c. Maybe

(I I haven’t thought about it

. How many children do you plan to have (include in this number any children that
you already have)?

0

l

2

3

4

5 or more

I haven’t thought about it

Unsure

War-m 999‘s»

. At what age do you plan to start having children?

a. I already have children
b. 21 — 23

c. 24 — 26

d. 27 — 29

e. 30 — 32

f. 33 — 35

g. 35+

h. I haven’t thought about it

170

i. Unsure
j. I do not plan to have children

6. How certain are you that your family plans will actually happen?
a. Very unsure
b. Unsure
c. Neither sure nor unsure
(1. Sure
e. Very sure

7. How likely is it that you will change your mind about your family plans?
a. Very unlikely
b. Unlikely
0. Neither likely nor unlikely
d. Likely
e. Very likely

8. How much doubt do you have about your family plans?
a. Very little doubt
b. Little doubt
c. Neither a lot of doubt nor a little doubt
d. Some doubt
e. A lot of doubt

l7l

Appendix 0
Certainty Regarding Family Plans Item Intercorrelations

 

Mean sd 1 2 3 4 5 6

 

Time 1

 

1. Item 1 3.36 1.18
2. Item 2 3.69 1.18 .81

 

 

 

 

3. Item 3 3.36 1.27 .76 .75

Time 2

4. Item 1 3.40 1.27 .82 .67 .68

5. Item 2 3.61 1.11 .79 .90 .75 .69

6. Item 3 3.38 1.25 .74 .72 .89 .69 .79

Time 3

7. Item 1 3.53 1.20 .80 .70 .64 .84 .68 .66
8. Item 2 3.63 1.18 .76 .75 .68 .70 .76 .67
9. Item 3 3.44 1.29 65 61 71 70 61 .65

 

Note. N ranges from 91 to 108. Correlations with r BI .21 I are signiﬁcant at p < .05.
Correlations with r 2 I .27 I are signiﬁcant at p < .01.

 

7 8

 

Time 1

 

1. Item I
2. Item 2
3. Item 3

 

Time 2

 

4. Item 1
5Jmm2
6. Item 3

 

Time 3

 

7. Item 1
8. Item 2 .68
9. Item 3 .72 .76

 

I72

Appendix P
Career Decision Self-efﬁcacy Items

Instructions: For each statement below, please think about your future career and
indicate how much conﬁdence you have that you could accomplish each of these tasks.

1 = No conﬁdence at all
2 = Very little conﬁdence
3 = Moderate conﬁdence
4 = Much conﬁdence

5 = Complete conﬁdence

Self-appraisal
1. Accurately assess your abilities in the medical ﬁeld.
2. Determine what your ideal medical specialty should be.
3. Decide what you value most in a medical specialty.
4. Figure out what you are and are not ready to sacriﬁce to achieve your career
goals.
5. Deﬁne the type of lifestyle you would like to live.
Gathering information
I. Use the intemet to ﬁnd information about the medical specialties that interest you.
2. Find out the employment trends for different medical specialties over the next ten
years.
3. Find out about the average yearly earnings of people in a medical specialty.
4. Talk with a person already employed in a medical specialty that you are interested
in. ‘
5. Find information about residencies in your preferred medical specialty.
Goal selection
1. Select one medical specialty from a list of potential specialties that you are
considering.
2. Select one residency program from a list of potential residency programs that you
are considering.
3. Choose a medical specialty that will ﬁt your preferred lifestyle.
4. Make a decision about which medical specialty to pursue and then not worry
about whether it was right or wrong.
5. Choose a medical specialty that will ﬁt your interests.

I73

Appendix Q
Career Decision Self-efﬁcacy Item Intercorrelations

Table Q1
Career Decision Self-Eﬂicacy Item Intercorrelations for Time I

 

Mean sd 1 2 3 4 5 6 7

 

Item 1 - SA 3.63 0.82

Item 2 — SA 3.58 0.90 .41

Item 3 — SA 3.76 0.78 .47 .68

Item 4 - SA 3.81 0.76 .41 .40 .52

Item 5 — SA 4.03 0.76 .43 .38 .53 .51

Item 6 — GI 3.97 0.80 .20 .14 .18 .17 .20

Item 7 — G1 3.29 0.94 .24 .33 .31 .21 .20 .33

Item 8 — G1 3.79 0.93 .19 .18 .28 .18 .17 .42 .64
Item 9 — GI 4.29 0.68 .29 .36 .30 .23 .20 .27 .21
Item 10 — G1 4.05 0.80 .37 .34 .32 .32 .29 .32 .39
Item 11 — GS 3.59 0.91 .27 .64 .44 .36 .14 .16 .42
Item 12 — GS 3.41 0.94 .48 .45 .40 .29 .28 .10 .42
Item 13 — GS 3.65 0.86 .41 .67 .64 .46 .51 .27 .51

Item 14— GS 3.12 1.10 .38 .56 .51 .37 .31 .06 .37
Item 15 — GS 3.86 0.87 .53 .63 .58 .53 .45 .24 .39
Notes. N = 106 to 108. Correlations with r Z [.19 Iare signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01. SA = self-appraisal, G1 =
gathering information, and GS = goal selection.

I74

Table Q1, continued
Career Decision Self-Eﬂicacy Item Intercorrelations for Time I

 

 

 

 

 

8 9 10 11 12 13 14
Item 1 — SA
Ime—SA
Iwm3—SA
Imm4—SA
Item 5 — SA
Item 6 — GI
Item 7 — GI
Item 8 — GI
Item 9 — GI .36
Item 10 — GI .47 .47
Item 11 — GS .31 .36 .45
Item 12 — GS .41 .30 .42 .67
Item 13 — GS 0.42 0.40 0.51 0.60 0.53
Item 14 — GS 0.26 0.39 0.38 0.60 0.53 0.66
Item 15 — GS 0.37 0.49 0.49 0.56 0.53 0.72 0.62
Table Q2
Career Decision Self-Eﬂicacy Item Intercorrelations for Time 2
Mean sd 1 2 3 4 5 6 7
Item 1 — SA 3.75 0.78 '
Item 2 — SA 3.76 0.77 .69
Item 3 - SA 3.86 0.70 .66 .73
Item 4 — SA 3.74 0.78 .61 .66 .65
Item 5 — SA 3.98 0.82 .51 .55 .56 .67
Item 6 — G1 3.90 0.85 .23 .29 .24 .34 .42
Item 7 — G1 3.48 1.00 .30 .38 .34 .40 .48 .66
Item 8 — G1 3.83 0.94 .26 .30 .28 .43 .54 .69 .72
Item 9 — G1 4.24 0.72 .35 .44 .36 .36 .39 .42 .45
Item 10 — G1 4.13 0.75 .39 .43 .44 .51 .51 .49 .55
Item 11 — GS 3.75 0.86 .51 .61 .56 .63 .56 .23 .32
Item 12 — GS 3.64 0.81 .58 .55 .55 .60 .53 .25 .33
Item 13 - GS 3.81 0.81 .59 .68 .63 .72 .60 .31 .41
Item 14 — GS 3.51 1.02 .52 .62 .49 .62 .45 .15 .31
Item 15 — GS 3.98 0.79 55 69 72 71 58 30 .32

 

Notes. N = 106 to 108. Correlations with r 2 I .19 I are signiﬁcant at p < .05.

Correlations with r 2 I .25 I are signiﬁcant atp < .01. SA = self-appraisal, G1 =

gathering information, and GS = goal selection.

Table Q2, continued
Career Decision Self-Efﬁcacy Item Intercorrelations for Time 2

 

 

 

 

 

8 9 10 11 12 13 14
Item I — SA
Item 2 — SA
Item 3 — SA
Imm4—SA
Item 5 — SA
Item 6 — GI
Item 7 — GI
Item 8 — GI
Item 9 — GI .36
Item 10 — GI .61 .58
Item 11 — GS .40 .41 .48
Item 12 - GS .39 .36 .54 .77
Item 13 - GS .36 .45 .53 .72 .69
Item 14 — GS .32 .34 .40 .78 .60 .66
Item 15 — GS .30 .44 .56 .60 .55 .74 .61
Table Q3
Career Decision Self-Efﬁcacy Item Intercorrelations for Time 3
Mean sd 1 2 3 4 5 6 7
. Item 1 — SA 3.63 0.69 i
Item 2 — SA 3.62 0.73 .48
Item 3 — SA 3.95 0.70 .55 .65
Item 4 — SA 3.85 0.71 .44 .48 .56
Item 5 — SA 4.03 0.75 .41 .47 .46 .53
Item 6 — GI 4.02 0.79 . .28 .05 .16 .15 .21
Item 7 - G1 3.53 0.74 .21 .13 .13 .26 .20 .51
Item 8 — G1 3.88 0.83 .17 .l l .16 .20 .26 .63 .63
Item 9 — G1 4.29 0.70 .43 .35 .48 .29 .30 .43 .28
Item 10 — G1 4.16 0.73 .38 .27 .38 .28 .33 .57 .39
Item 11 — GS 3.70 0.85 .53 .66 .57 .64 .38 .21 .23
Item 12 — GS 3.62 0.84 .47 .53 .47 .56 .31 .21 .17
Item 13 - GS 3.71 0.72 .47 .57 .54 .58 .47 .33 .30
Item 14 — GS 3.58 0.87 .43 .60 .39 .49 .37 .13 .16
Item 15 — GS 4.03 0.72 58 66 59 42 41 .23 .17

 

Notes. N = 89 to 91. Correlations with r 2 I .21 I are signiﬁcant atp < .05.

Correlations with r 2 I .27 I are signiﬁcant atp < .01. SA = self-appraisal, GI =

gathering information, and GS = goal selection.

I76

Table Q3, continued

 

Career Decision Self-Eﬂicacy Item Intercorrelations for Time 3

 

8 9 10 ll 12 13 14

 

Item I — SA

Item 2 — SA

Imm3—SA

Item 4 — SA

Item 5 — SA

Item 6 — GI

Item 7 — GI

Item 8 — GI

Item 9 — GI .44

Item 10 — GI .56 .66

Item 11 — GS .22 .44 .35

Item 12 — GS .23 .25 .42 .69
Item 13 - GS .39 .45 .51 .66 .62

Item 14 — GS .18 .20 .28 .72 .56 .53
Item 15 — GS .19 .33 .39 .63 .51 .64 .64

 

I77

Appendix R
Results of CFA for Career Decision Self-efﬁcacy Items

 

 

N 76 Df P Comparative Root mean
Fit Index (CF 1) square error of
approximation
(RMSEA)
Time 1 106 153.89 80 .00 .91 .09
Time 2 103 117.13 81 .01 .97 .07
Time 3 88 135.32 83 .00 93 .09

 

I78

Appendix S
W ork-family Decision-making Self-efﬁcacy Items

Instructions: For each statement below, please think about how you want to balance
your family life and your work in the future and indicate how much conﬁdence you have
that you could accomplish each of these tasks.

1 = No conﬁdence at all
2 = Very little conﬁdence
3 = Moderate conﬁdence
4 = Much conﬁdence

5 = Complete conﬁdence

Self-appraisal

l.
2.
3.
4.

5.

Accurately assess your preferences for how you want to balance your family life
with your 335% in medicine.

Determine what your ideal balance between family and w_o_rk would be.

Decide how you want to balance your family life with your w_013 in medicine.
Figure out what you are and are not willing to sacriﬁce in your family life to
achieve your lor_k_ goals.

Deﬁne how you would ideally like to balance family and m in the future.

Gathering information

I.

2.

3.
4.

5.

Gather information about the challenges for balancing family and log]; in
different medical specialties.

Find out about the demands that different medical specialties place on doctors for
balancing family and work.

Find out about the work-family balance of doctors in different medical specialties.
Talk with doctors about the work-family balance they experience in their medical
specialties.

Find information about the challenges and beneﬁts of balancing family and m
in different medical specialties.

 

Goal selection

1.

2.

3.

4.

Select a medical specialty that will put you on the right path for balancing family
and m as a doctor.

Select one medical specialty that matches your preferences for how you want to
balance your family life with your £913 in medicine.

Choose a medical specialty that will allow you to balance family and m in the
ways that you want.

Stay committed to pursuing a medical specialty that will allow you to balance
family and work.

5. Choose a medical specialty that ﬁts with your goals for how you want to balance

family and work.

179

Appendix T
Work-family Decision-making Self-eﬂicacy Item Intercorrelations

Table Tl
Work-family Decision-making Self-eﬂicacy Item Intercorrelations for Time I

 

Mean sd 1 2 ' 3 4 5 6 7

 

Item I — SA 3.48 0.81

Item 2 — SA 3.48 0.86 .76

Item 3 -- SA 3.36 0.85 .63 .72

Item 4 — SA 3.66 0.83 .54 .69 .52

Item 5 — SA 3.53 0.80 .62 .67 .62 .55

Item 6 -— G1 3.41 0.79 .32 .31 .36 .29 .26

Item 7 — G1 3.51 0.80 .49 .47 .41 .43 .32 .73

Item 8 — G1 3.54 0.84 .44 .42 .38 .40 .35 .74 .84
Item 9 — G1 3.74 0.84 .28 .28 .27 .36 .33 .56 .56
Item 10 — G1 3.56 0.79 .38 .36 .42 .45 .37 .69 .73
Item 11 — GS 3.40 0.89 .56 .48 .52 .51 .49 .49 .54
Item 12 - GS 3.36 0.89 .56 .45 .54 .50 .42 .52 .52
Item 13 — GS 3.41 0.83 .55 .49 .50 .55 .47 .49 .50

Item 14 — GS 3.64 0.86 .58 .49 .56 .51 .46 .45 .46
Item 15 —GS 3.50 0.88 60 53 47 51 50 .50 .54

 

Notes. N = 107 to 108. Correlations with r 2 I .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01. SA = self-appraisal. GI =
gathering information. GS = goal selection.

180

Table T1, continued

Work-family Decision-making Self-eﬁcacy Item Intercorrelations for
Time I

 

8 9 10 ll l2 13 14

 

Item 1 — SA

Item 2 - SA

Item 3 — SA

Item 4 — SA

Item 5 — SA

Item 6 — GI

Item 7 —- GI

Item 8 - GI

Item 9 — GI .72

Item 10 - GI .76 .74

Item 11 — GS .54 .46 .54

Item 12 — GS .51 .44 .51 .89

Item 13 — GS .50 .45 .52 .78 .87

Item 14 — GS .56 .44 .52 .69 .71 0.65
Item 15 — GS .58 .44 .52 .76 .79 .84 .75

 

181

Table T2
Work-family Decision-making Self-efﬁcacy Item Intercorrelations for Time 2

 

Mean sd 1 2 3 4 5 6 7

 

Item 1 — SA 3.72 0.77

Item 2 - SA 3.74 0.79 .82

Item 3 — SA 3.62 0.79 .76 .77

Item 4 — SA 3.74 0.79 .73 .69 .77

Item 5 — SA 3.66 0.79 .83 .77 .81 .76

Item 6 — GI 3.74 0.80 .56 .54 .60 .57 .57

Item 7 — G1 3.76 0.83 .51 .56 .60 .59 .57 .82

Item 8 — G1 3.78 0.80 .58 .61 .65 .60 .67 .77 .89
Item 9 — GI 4.00 0.79 .53 .54 .60 .54 .56 .73 .76
Item 10 — GI 3.74 0.79 .52 .55 .65 .57 .57 .74 .79
Item 11 — GS 3.60 0.83 .62 .67 .72 .64 .68 .42 .52
Item 12 — GS 3.65 0.80 .59 .67 .71 .65 .69 .50 .57
Item 13 — GS 3.64 0.78 .61 .66 .70 .65 .69 .51 .57

Item 14— GS 3.66 0.82 .62 .69 .71 .64 .60 .52 .59
Item 15 — GS 3.68 0.82 65 66 69 66 69 50 .57

 

Notes. N = 106 to 108. Correlations with r 2 I .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01. SA = self-appraisal. G1 =
gathering information. GS = goal selection.

[82

Table T2, continued
Work-family Decision-making Self—efﬁcacy Item Intercorrelations
for Time 2

 

8 9 10 11 12 I3 14

 

Item 1 — SA

Item 2 — SA

lam3—SA

Item 4 — SA

Item 5 —— SA

Item 6 — GI

Item 7 — GI

Item 8 - GI

Item 9 — GI .80

Item 10 — GI .82 .75

Item 11 -— GS .54 .50 .57

Item 12 — GS .62 .58 .59 .80
Item 13 — GS .53 .54 .61 .81 .86

Item 14 — GS .60 .52 .56 .77 .76 .78
Item 15 —- GS .58 .55 .59 .80 .79 .86 .87

 

Table T3

 

Work-family Decision—making Self-eﬂicacy Item Intercorrelations for Time 3

 

Mean sd 1 2 3 4 ' 5 6 7

 

Item I — SA 3.59 0.70

Item 2 — SA 3.70 0.74 .69

Item 3 — SA 3.63 0.68 .66 .76

Item 4 — SA 3.81 0.73 .44 .38 .56

Item 5 — SA 3.75 0.71 .67 .62 .54 .43

Item 6 — G1 3.64 0.72 .42 .39 .40 .36 .43

Item 7 — G1 3.73 0.73 .41 .42 .44 .49 .49 .82

Item 8 — G1 3.76 0.74 .48 .53 .51 .53 .53 .72 .79
Item 9 — G1 3.96 0.82 .35 .23 .33 .44 .44 .56 .64
Item 10 — G1 3.67 0.78 .38 .43 .43 .54 .53 .72 .79
Item 11 — GS 3.58 0.78 .48 .41 .50 .53 .49 .46 .63
Item 12 — GS 3.54 0.78 .42 .37 .55 .60 .51 .51 .57
Item 13 —- GS 3.57 0.81 .46 .31 .55 .58 .48 .36 .47

Item 14 — GS 3.81 0.77 .50 .43 .54 .59 .44 .46 .48
Item 15 —GS 3.58 0.78 46 40 56 57 51 45 .52

 

Notes. N = 107 to 108. Correlations with r BI .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01. SA = self-appraisal. G1 =
gathering information. GS = goal selection.

183

Table T3, continued
Work-family Decision-making Self-efﬁcacy Item Intercorrelations for Time 3

 

 

8 9 10 11 12 13 14
Item 1 — SA '
Item 2 — SA
Item 3 — SA
Item 4 — SA
ImmS-SA
Item 6 — GI
Item 7 — GI
Item 8 — GI
Item 9 — GI .63
Item 10 — GI .74 .70
Item 11 — GS .59 .57 .58
Item 12 — GS .61 .56 .65 .79
Item 13 — GS .50 .50 .49 .81 .87
Item 14 - GS .47 .50 .51 .65 .68 .67
Item 15 — GS .55 .55 .58 .81 .88 .86 .78

 

I84

Appendix U

Results of CFA for Work-family Decision-making Self-eﬂicacy Items

 

 

N 36 Df P Comparative Root mean
Fit Index (CFI) square error of
approximation
(RMSEA)
Time 1 105 128.09 79 .00 .97 .08
Time 2 101 107.78 79 .02 .98 .06
Time 3 83 137.10 82 .00 .95 .09

 

Appendix V
Role-planning Knowledge Items

Instructions: Please answer the following questions about your career and family plans.

I = Strongly disagree

2:.-
3:
4:
5:

l.

Disagree

Neither agree nor disagree
Agree

Strongly agree

I don’t know how to plan for combining my medical career and my family.

2. Figuring out how to balance my medical career and my family confuses me because I

10.

don’t feel I know enough about myself or about the stresses involved in balancing
these roles.

I can’t understand how some people can be so certain about how to successfully
manage career and family responsibilities.

When it comes to combining my medical career with my family, I can’t seem to make
up my mind how to do it successfully.

It’s easy to be certain how to manage my future medical career and family obligations
in ways that are realistic for me (reverse-coded).

I have little or no idea of what being both a doctor and a parent will be like.

I don’t know whether my plans for combining my medical career and my family will
allow me to be the kind of person I want to be.

I’m very clear on how to plan for combining my medical career and family
responsibilities (reverse-coded).

I don’t know whether my plans for combining my medical career with my family are
realistic.

I know a lot of strategies for combining a family with a career in a way that
minimizes the stress involved (reverse-coded)

186

Appendix W
Role-planning Knowledge Item Intercorrelations

 

 

Table Wl
Role-Llanning Knowledge Item Intercorrelations for Time 1

Mean sd 1 2 . 3 4 5
I. Item I 2.97 1.06
2. Item 2 2.85 1.17 .64
3. Item 3 2.87 1.15 .68 .63
4. Item 4 2.75 1.06 .64 .60 .72
5. Item 5 3.32 0.86 .40 .37 .52 .51
6. Item 6 3.35 1.15 .47 .45 .35 .29 .33
7. Item 7 3.05 1.16 .51 .45 .50 .59 .45
8. Item 8 3.47 0.96 .62 .53 .59 .52 .55
9. Item 9 3.32 0.98 .33 .28 .36 .41 .34

10. Item 10 3.32 0.99 59 42 46 42 39

 

Note. N = 106 to 108. Correlations with r 2 I .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01.

 

 

Table WI , continued

Role-planning Knowledge Item Intercorrelations for Time 1
6 7 8 9

I. Item 1

ZJmm2

3. Item 3

4Jmm4

5. Item 5

6. Item 6

7. Item 7 .32

8. Item 8 .52 .51

9. Item 9 .24 .58 .35

10. Item 10 .41 .33 .53 .27

 

187

Table W2
Role-planningKnowledge Item Intercorrelations for Time 2

Mean sd 1 2 3 4 5
. Item I 2.49 0.91
. Item 2 2.66 1.07 .65 -
. Item 3 2.78 1.04 .55 .57
. Item 4 2.70 1.05 .66 .62 .71
. Item 5 3.22 0.88 .41 .30 .32 .40
. Item 6 3.13 1.14 .48 .55 .43 .53 .26
. Item 7 2.90 1.12 .55 .48 .58 .65 .39
. Item 8 3.31 0.88 .43 .32 .23 .31 .51
. Item 9 3.26 1.03 .45 .37 .49 .60 .40
10. Item 10 3.09 1.01 .45 .45 .31 .40 .38
Note. N = 106 to 108.Correlations with r 2 I .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01.

\DOOQONUIAUJNI—a

Table W2, continued
Role-planning Knowledge Item Intercorrelations for Time 2

 

 

6 7 8 9
1. Item I
2. Item 2
31mm3
4. Item 4
Sme5
6. Item 6
7. Item 7 .43
8. Item 8 .25 .39
9. Item 9 .46 .63 .30
10. Item 10 .48 .28 .40 .36

 

I88

Table W3
RolﬂlanninLKnowledge Item Intercorrelations for Time 3

 

Mean sd 1 2 3 4 5

 

. Item I 2.43 0.94

. Item 2 2.49 1.02 .74

. Item 3 2.56 1.10 .60 .51

. Item 4 2.59 1.01 .71 .63 .67

. Item 5 3.07 0.88 .43 .39 .51 .43

. Item 6 3.09 1.23 .48 .47 .39 .38 .36
. Item 7 2.88 1.05 .60 .42 .54 .63 .31
. Item 8 3.13 0.91 .52 .46 .55 .54 .71
. Item 9 3.23 0.99 .45 .34 .52 .63 .42
10. Item 10 2.96 0.93 53 52 40 44 .49

\OOO\10\M-&WNi—n

 

Note. N = 89 to 91. Correlations with r 2 I .21 Iare signiﬁcant at p < .05.
Correlations with r 2 I .27 I are signiﬁcant at p < .01.

Table W3, continued
Role-planning Knowledge Item Intercorrelations for Time 3

 

6 7 8 9

 

I. Item 1
ZmeZ
3. Item 3
4Jmm4
5. Item 5
6me6
7. Item 7 .46

8. Item 8 .49 .40

9. Item 9 .40 .65 .56

10. Item 10 .55 .45 .53 .45

 

189

Appendix X
Factor Loadings for Role-Planning Knowledge Items

 

 

Table X1
Factor Loadings for Time 1 Role-Planning
Knowledge Items

Factor 1 Factor 2
Item I 0.79 0.31
Item 2 0.68
Item 3 0.69 0.41
Item 4 0.58 0.53
Item 5 0.47 0.41
Item 6 0.55
Item 7 0.32 0.78
Item 8 0.70 0.35
Item 9 0.65
Item 10 0.64

 

Notes. Two factors were extracted using
principal axis factor analysis with a varimax
rotation. Factor loadings less than I .30 I are

 

 

not shown.
Table X2
Factor Loadings for Time 2 Role-Planning
Knowledge Items

Factor 1 Factor 2
Item 1 0.65 0.42
Item 2 0.67
Item 3 0.76
Item 4 0.85
Item 5 0.60
Item 6 0.57
Item 7 0.68 0.31
Item 8 0.76
Item 9 0.61
Item 10 0.35 0.46

 

Notes. Two factors were extracted using
principal axis factor analysis with a varimax
rotation. Factor loadings less than I .30 I are
not shown.

190

Appendix Y
Demographics items
I. What is your gender
a. Male
b. Female

2. How old are you?
a. 22 or younger

b. 23—24
c. 25—26
(I. 27—28

e. 29 or older

3. What is your current marital status?

Never married

Currently married

Divorced or separated

Widowed

Long-term live-in relationship (for example, common law marriage, same
sex partnership)

{DP-953‘!”

4. Are you currently in a committed, romantic relationship?
a. Yes
b. No

5. Which medical school do you attend and what year are you?
a. COM lSt year
b. COM 2"d year

COM 3rd year

COM 4m year

CHM ISt year

CHM 2"d year

CHM 3rd year

CHM 4‘h year

P‘Q° rm .o-e

6. What is your current parental status?
a. I have never had any children
b. I have children that live with me
c. l have children that do not live with me (for example, other parent has full
custody, gave up child for adoption, children are grown)

I91

7. How old is your youngest child?

.09-99".”

I do not have children
0 — 5 years old

6 — 10 years old

11 — 15 years old

16 or older

8. What ethnicity do you consider yourself to be?

P‘QQ ran 999‘!»

American Indian or Alaskan native

Asian

Black / African American

Hispanic

Native Hawaiian or other Paciﬁc Islander
White/Caucasian/Not of Hispanic origin
Multi-racial

Other:

 

I92

Appendix Z
Core Self-evaluations Items

Instructions: Please indicate whether you agree or disagree with the following
statements.

1 = Strongly disagree

2 = Disagree

3 = Neither agree nor disagree

4 = Agree

5 = Strongly agree

1. I am conﬁdent I get the success I deserve in life.

2. Sometimes I feel depressed (reverse-coded).

3. When I try, I generally succeed.

4. Sometimes when I fail I feel worthless (reverse-coded).

5. I complete tasks successfully.

6. Sometimes, I do not feel in control of my school work (reverse-coded).
7. Overall, I am satisﬁed with myself.

8. I am ﬁlled with doubts about my competence (reverse-coded).

9. I determine what will happen in my life.

10. I do not feel in control of my success in medical school (reverse-coded).
l 1. I am capable of coping with most of my problems.

12. There are times when things look pretty bleak and hopeless to me (reverse-coded).

I93

Appendix AA
Core Self-evaluations Item Intercorrelations

 

 

Mean sd 1 2 3 4 5 6
I. Item 1 4.01 0.67
2. Item 2 2.56 0.97 .18
3. Item 3 4.27 0.57 .29 .18
4. Item 4 3.01 1.08 .23 .47 .35
5. Item 5 4.08 0.55 .36 .16 .40 .40
6. Item 6 2.73 1.12 .12 .39 .26 .52 .34
7. Item 7 4.05 0.82 .42 .26 .21 .44 .54 .32
8. Item 8 3.23 1.09 .17 .30 .32 .47 .45 .39
9. Item 9 3.79 0.79 .35 .14 .15 .22 .28 .14
10. Item 10 3.86 0.91 .30 .15 .14 .22 .47
11. Item 11 4.00 0.74 .40 .29 .31 .42 .42
12. Item 12 3.52 1.08 32 50 26 59 40

 

Notes. N = 106 to 108. Correlations with r 2 I .19 I are signiﬁcant at p < .05.
Correlations with r 2 I .25 I are signiﬁcant at p < .01.

 

7 8 9 10 II

 

. Item 1

. Item 2

. Item 3

me4

JEmS

. Item 6

. Item 7

. Item 8 .48

. Item 9 .40 .26

10. Item 10 .50 .31 .26
II. Item 11 .55 .28 .44
12. item 12 .43 .38 .20

\DOOﬂan-bwwu—n

.47
.43 .47

 

I94

Appendix BB
CFA for Core Self-evaluation Items

 

 

36‘ Df P Comparative Root mean
Fit Index (CF 1) square error of
approximation
(RMSEA)
CSE 104 153.89 80 .00 .91 .09

I95

Appendix CC
Role-importance Items

Work role-importance

1.

.U'PP’!"

Having work / a career that is interesting and exciting to me is my most important life
goal .

I expect my job / career to give me more real satisfaction than anything else I do.

My life would seem empty if I didn’t have a job / career.

If I chose not to work, I would regret it.

Although having a job / career requires many sacriﬁces, the beneﬁts of working are
worth it all.

Family role-importance

1.

U)

Although having a family requires many sacriﬁces, the love and enjoyment of family
of one’s own are worth it.

If I chose not to have a family, I would regret it.

It is important to me that I will be an effective family member.

The whole I idea of having a family and caring for them is not attractive to me
(reverse-coded).

My life would be empty if I never had a family.

196

Appendix DD
Role-importance Item Intercorrelations

 

 

Mean sd 1 2 3 4 5
l.WIteml 3.13 1.22
2. W Item 2 2.60 1.04 .57 -
3. Wltem3 3.30 1.11 .28 .41
4. W Item 4 3.64 1.03 .07 .19 .50
5. W Item 5 3.82 0.78 .09 .38 .29 .26
6. F Item 1 4.50 0.69 -.25 -.31 -.09 -.12 -.08
7.Fltem2 4.11 1.16 -.10 -.19 .06 .10 -.ll
8. F Item 3 4.57 0.59 -.09 -.24 .03 .07 .03
9. F Item 4 4.79 0.73 -.18 -.15 .09 .08 .13
10. F Item 5 3.89 1.28 -.06 -.15 .20 .12 -.08

 

Notes. N = 105 to 108. Correlations with r 2 I .19 I are signiﬁcant at p <
.05. Correlations with r 2 I .25 I are signiﬁcant at p < .01. W = work. F =
family.

 

 

6 7 8 9
l. W Item 1
2. W Item 2
3. W Item 3
4. W Item 4
5. W Item 5
6. F Item 1
7. F Item 2 .49
8. F Item 3 .63 .48
9. F Item 4 .56 .47 .50
10. F Item 5 .48 .54 .37 .36

 

I97

Appendix EE
Factor Loadings for Role-importance Items

Table EEI
Factor Loadings for Role-importance Items

 

Factor 1 Factor 2 Factor 3 Factor 4

 

l. W Item 1 0.65

2. W Item 2 0.81 0.34
3. W Item 3 0.35 0.63

4. W Item 4 0.74

5. W Item 5 0.65

6. F Item 1 0.82
7. F Item 2 0.68
8. F Item 3 0.70
9. F Item 4 0.68
10. F Item 5 0.63

 

Notes. Three factors were extracted using principal axis factor
analysis with a varimax rotation. Factor loadings less than
I .30 I are not shown. W = work. F = family.

Table E132
Factor Loadings for Role-importance Items
Forcing Two Factors

 

 

Factor 1 Factor 2
l. W Item 1 0.45
2. W Item 2 -0.30 0.70
3. W Item 3 0.71
4. W Item 4 0.46
5. W Item 5 0.46
6. F Item 1 0.78
7. F Item 2 0.69
8. F Item 3 0.70
9. F Item 4 0.68
10. F Item 5 0.63

 

Notes. Two factors were forced using
principal axis factor analysis with a varimax
rotation. Factor loadings less than I .30 I are
not shown. . W = work. F = family.

198

Appendix FF
Work-family Information Gathering Time 2 Items

Instructions: Please indicate how frequently you have engaged in the following
behaviors using the scale provided below.

1 = Never

2 2' Rarely

3 = Occasionally
4 = Often

5 = Frequently

1.

Since you entered medical school, how frequently have you researched
information related to balancing your medical career with your family life on the
intemet?

Since you entered medical school, how frequently have you spent time talking to
your friends or family about balancing your medical career with your family life?

Since you entered medical school, how frequently have you spent time talking
with doctors, professors, or administrators about balancing your medical career
with your family life?

Since you entered medical school, how frequently have you spent time thinking
about balancing your medical career with your family life?

Since you entered medical school, how frequently have you used additional

resources (not mentioned here) to gather information about balancing your
medical career with your family life (e. g., books, magazines?)

199

Appendix GG
Work-family Information Gathering Time 2 Item Intercorrelations

 

 

Mean sd 1 2 3 4
1. Item 1 2.09 1.08
2. Item 2 3.41 1.07 .42 .
3. Item 3 2.81 1.11 .34 .56
4. Item 4 3.69 1.11 .43 .66 .49
5. Item 5 2.17 1.08 .81 .42 .34 .43

 

Notes. N = 108. Correlations with r 2 I .19 I are signiﬁcant at p <
.05. Correlations with r 2 I .25 I are signiﬁcant at p < .01.

200

Appendix HH
Work-family Information Gathering Time 3 Items

. Did you receive an email about work-family issues facing medical students?
a. Yes

b. No

c. I don’t know

. Did you read the email about work-family issues facing medical students?
a. No

b. Yes, but I didn’t read it

c. Yes, and I read some of it

d. Yes, and I read all of it

. If you received an email, the email also included a link to ﬁnd out more
information on a website. Did you visit that website?
a. Yes

b. No
c. I did not receive an email with a link

. Approximately how much time did you spend visiting the website?
I did not visit the website
Less than 5 minutes
5 — 10 minutes
10 — 15 minutes
15 — 20 minutes
More than 20 minutes

{"79 9.057!”

. Over the past two weeks, how much time did you spend research information
related to balancing your medical career with your family life on the intemet?

a. None

b. Less than 1 hour
c. 1 — 2 hours

(I. 3 — 4 hours

c. More than 5 hours

. Over the past two weeks, how much time did you spend talking to your friends or
family about balancing your medical career with your family life?

a. None

b. Less than 1 hour
c. 1 — 2 hours

d. 3 — 4 hours

e. More than 5 hours

201

7. Over the past two weeks, how much time did you spend talking with doctors,
professors, or administrators about balancing your medical career with your

family life?
a. None
b. Less than 1 hour
0. l — 2 hours
(I. 3 -— 4 hours
e. More than 5 hours

8. Over the past two weeks, how much time did you spend thinking about balancing
your medical career with your family life?

a. None

b. Less than 1 hour
c. l — 2 hours

(I. 3 — 4 hours

c. More than 5 hours

9. Over the past two weeks, how much time did you spend using additional
resources (not mentioned here) to gather information about balancing your
medical career with your family life (e.g., books, magazines)?

a. None

b. Less than 1 hour
c. l — 2 hours

(I. 3 — 4 hours

c. More than 5 hours

10. What additional information did you use?

202

Appendix II

Work—family Information Gathering Time 3 Item Intercorrelations

 

 

Mean sd 1 2 3 4
1. Item 5 1.64 0.78
2. Item 6 2.30 1.14 .25
3. Item 7 1.77 0.86 .30 .47
4. Item 8 2.80 1.34 .26 .71 .43
5. Item 9 1.33 0.56 30 .18 .19 .25

 

Note. N = 90 to 91. Correlations with r 2 I .21 I are signiﬁcant at p
< .05. Correlations with r 2 I .27 I are signiﬁcant at p < .01.

203

Appendix J]

Factor Loadings for Work-family Information Gathering Time 3

 

 

Factor 1 Factor 2
Item 5 0.68
Item 6 0.90
Item 7 0.46 0.35
Item 8 0.76
Item 9 0.41

 

Notes. Two factors were forced using
principal axis factor analysis with a varimax
rotation. Factor loadings less than I .30 I are
not shown. .

204

Appendix KK
Discrimination Among Self-eﬂ‘icacy—related Scales

Table KKl

 

Factor loadings of WFRPK, CDSE, and WFDMSE Items

 

Factor

 

1 2 i 3 4 5

 

T1wﬁpk3 .771

lefrka .730

lefrpk4 .720

T1wﬁpk2 .705

lefrpk8 .630

leﬁka .541

T l wfrpk7 .513

lefrkaO .462

lefdmse8 .862

lefdmse l 0 .760

lefdmse7 .760

lefdmse6 .739

lefdmse9 .672

T1 cdsel 1 .814

TI cdse2 .690

T1cdse14 .620

Tlcdse12 .588

Tlcdsel3 .565

Tlcdse15 .520 .483
T l wfdmse2 ' .827

lefdmsel . .726
T1wfdmse5 .646

T l wfdmse3 .571

lefdmse4 .476

TlcdseS .690
Tlcdse4 . .597
Tlcdse3 .421 .516
Tlcdsel .499
lefrpk9

 

Notes. Principal axis factor analysis with a varimax rotation was used. Factor loadings
less than .40 are suppressed. WFRPK = Work-family role-planning knowledge. CDSE
= Career decision self-efﬁcacy. WFDMSE = Work-family decision-making self-
efﬁcacy. Continued on next page.

205

Table KKI , continued
Factor loadings of WFRPK, CDSE, and WFDMSE Items

 

 

 

Factor

6 7 8 _ 9
lefdmsel2 .699
lefdmsel3 .679
lefdmsel 1 .635
lefdmselS .602
T1wfdmse14 .427 .460
T l cdse8 .801
T1cdse6 .506
T 1 cdse7 .477
T1cdse10
T l wfrpk6 .468
T1cdse9

 

Notes. Principal axis factor analysis with a varimax rotation
was used. Factor loadings less than .40 are suppressed.
WFRPK = Work-family role—planning knowledge. CDSE
= Career decision self-efﬁcacy. WFDMSE = Work-family
decision-making self-efﬁcacy.

206

Table KK2

 

Factor loadings of CSE, CDSE, and WFDMSE Items

 

Factor

 

l 2 3 4 5 6 7 8 9

 

Tlcdsell .840

Tlcdsel4 .646

T1cdse2 .635

Tlcdse12 .559

Tlcdsel3 .538

Tlcdse15 .514 .444

Tlcdse9 .433

T1wfdmse8 .841

lefdmseIO .780

T1wfdmse9 .756

lefdmse6 .716

T l wfdmse7 .715

T1wfdmse2 .759

lefdmse5 .754

lefdmsel .695

lefdmse3 .669

lefdmse4 .524

T1cse9

Tlcse5 .639

Tlcse7 .612

TI cse10 .612

Tlcsell .548

Tlcsel .539

Tlcse3

lefdmse12 .782

T l wfdmse l 3 .732

T l wfdmsel 1 .668

lefdmseIS .643

lefdmse14 .440

Notes. Principal axis factor analysis with a varimax rotation was used.
Factor loadings less than .40 are suppressed. WFRPK = Work-family role-
planning knowledge. CDSE = Career decision self-efﬁcacy. WFDMSE =
Work-family decision-making self-efﬁcacy. Continued on next page.

207

 

Table KK2, continued
Factor loadings ofCSE, CDSE, and WFDMSE Items

 

 

 

Factor

1 2 3 4 5 6 7 8 9
T1cse4 .764
T1 cseZ ' .696
Tlcse12 .610
Tlcse6 .591 .427
T1cse8 .462
T1 cdse5 .710
T1 cdse3 .459 .590
TI cdse4 .548
Tlcdsel .440
T l cdse8 .753
T 1 cdse7 .548
Tlcdse10 .509
T1 cdse6 .506

 

Notes. Principal axis factor analysis with a varimax rotation was used. Factor loadings
less than .40 are suppressed. WFRPK = Work-family role-planning knowledge. CDSE
= Career decision self-efﬁcacy. WFDMSE = Work-family decision-making self-
efﬁcacy.

208

Appendix LL
Manipulation Check Items

Instructions: For each of the following topics, indicate whether you believe that this
topic was or was not included in the presentations that you just viewed.

1. Determining what your ideal medical specialty will be.
a. Deﬁnitely not included

b. Probably not included

c. No idea whether it was included or not
d. Probably included

e. Deﬁnitely included

2. Determining what your ideal balance between work and family would be.

a. Deﬁnitely not included

b. Probably not included

c. No idea whether it was included or not
(I. Probably included

e. Deﬁnitely included

3. Gathering information about a medical specialty you are interested in.
a. Deﬁnitely not included
b. Probably not included

0. No idea whether it was included or not
(I. Probably included
e. Deﬁnitely included

4. Gathering information about the work-family balance of a medical specialty you
are interested in.

a. Deﬁnitely not included

b. Probably not included

c. No idea whether it was included or not
(I. Probably included

e. Deﬁnitely included

5. Selecting one medical specialty from a list of potential specialties that you are
considering.

Deﬁnitely not included

Probably not included .

No idea whether it was included or not

Probably included

Deﬁnitely included

{DP-99"!”

6. Selecting one medical specialty that will help you balance work and family from a
list of potential specialties that you are considering.
a. Deﬁnitely not included

209

.09-99‘

Probably not included

No idea whether it was included or not
Probably included

Deﬁnitely included

210

REFERENCES

Amatea, B, Cross, G., Clark, J. & Bobby, C. (1986). Assessing the work and family role
expectations of career-oriented men and women: The Life Role Salience Scales.
Journal of Marriage and Family, 48, 831-838.

Aryee, S., Luk, V., Leung, A., & Lo, S. (1999). Role stressors, interrole conﬂict, and
well-being: The moderating inﬂuence of spousal support and coping behaviors

among employed parents in Hong Kong. Journal of Vocational Behavior, 54,
259-278.

Baldwin, T. T. & Ford, J. K. (1988). Transfer of training: A review and directions for
future research. Personnel Psychology, 41, 63-105.

Bandura, A., & Adams, N. E. (1977). Analysis of self—efﬁcacy theory of behavioral
change. Cognitive Therapy and Research, 1, 287—3 10.

Barnett, R., Gareis, K., James, J. & Steele, J. (2003). Planning ahead: College seniors'

concerns about career-marriage conﬂict. Journal of Vocational Behavior, 62,
305-319.

Baron, R. M., & Kenny, D. A. (1986). The moderator—mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51, 1173—1 182.

Behson, S. (2002). Coping with family-to-work conﬂict: The role of informal work
accommodations to family. Journal of Occupational Health Psychology, 7, 324-
341.

Bergeron, L. M., & Romano, J. L. (1994). The relationships among career decision-
making self-efﬁcacy, educational indecision, vocational indecision, and gender.
Journal of College Student Development, 35, 19-35.

Betz, N. E. (1992). Counseling uses of career self-efﬁcacyitheory. Career Development
Quarterly, 41, 22-27.

Betz, N. and Hackett, G. (1981). The relationship of career related self efﬁcacy
expectations to perceived career options in college women and men. Journal of

Counseling Psychologv, 28, 399 - 410.

Betz, N. E., & Luzzo, D. A. (1996). Career assessment and the Career Decision-
Making Self-Efﬁcacy scale. Journal of Career Assessment, 4, 413-428.

Bond, 1., Thompson, C ., Galinsky, E., & Prottas, D. (2002). Highlights ofthe national
studv of the changing workforce. New York: Families and Work Institute.

211

Bruck, C. S., & Allen, T. D. (2003). The relationship between big ﬁve personality
traits, negative affectivity, type A behavior, and work-family conﬂict. Journal of
Vocational Behavior, 63, 457-472.

Bu, N., & McKeen, C. (2000). Work and family expectations of the future managers and
professionals of Canada and China. Journal of Managerial Psychology, 15, 77 1 -
794.

Burke, R. (1988). Some antecedents and consequences of work-family conﬂict. Journal
of Social Behavior and personality, 3, 287-302.

Burke, M. J. & Day, R. D. (1986). A cumulative study of the effectiveness of managerial
training. Journal of Applied Psychology, 71, 232-245.

Burley, K. (1994). Gender differences and similarities in coping responses to anticipated
work-family conﬂict. Psychological Reports, 74, 115-123.

Campbell, J. P. (1988). Training design for performance improvement. In J. P. Campbell
& R. J. Campbell (Eds), Productivity in Organizations, (pp. 177-216). San
Francisco: Jossey-Bass.

Carlson, D. S. (1999). Personality and role variables as predictors of three fomis of work-
family conﬂict. Journal of Vocational Behavior, 55, 236-253.

Carlson, D., Kacmar, M., & Williams, L. (2000). Construction and initial validation of a
multidemensional measure of work-family conﬂict. Journal of Vocational
Behavior, 56, 249-276. ‘

Chen, G., Gully, S. M., & Eden, D. (2001). Validation of a new general self-efﬁcacy
scale. Organizational Research Methods, 4, 62-83.

Church, A. T., Teresa, J. S., Rosebrook, R., & Szendre, D. (1992). Self-efﬁcacy for
careers and occupational considerations in minority high school equivalency
students. Journal of Counseling Psychology, 39, 498-508.

Cohen, .I. (1969). Statistical Power Analysis for the Behavioral Sciences. Hillsdale,
NJ: Lawrence Erlbaum Associates.

Crites, JD. (1961). A model for the measurement of vocational maturity. Journal of
Counseling Psychology, 8, 255-259.

Diener, E., Larsen, R. J., & Emmons, R. A. (1984). Person-situation interactions: Choice

of situations and congruence of coping models. Journal of Personality and Social
Psychology), 4 7, 580-592.

212

Donaldson, S.I., & Christie, CA. (2006). Emerging career opportunities in the
transdiscipline of evaluation science._In S. I. Donaldson, D.E. Berger, &

K.Pezdek Eds., Applied psychology: New frontiers and rewarding careers.
Mahwah, NJ: Erlbaum.

Duxbury, L., & Higgins, C. (1991). Gender differences in work-family conﬂict. Journal
of Applied Psychology, 76, 60-73.

Eby, L., Casper, W., Lockwood, A., Bordeaux, C., & Brinley, A. (2005). Work and
family research in IO/OB: Content analysis and review of the literature (1980-
2002). Journal of Vocational Behavior, 66, 124-197.

F ogarty, G. J., Machin, M. A., Albion, M. J ., Sutherland, L. F ., Lalor, G. 1., & Revitt, S.
(1999). Predicting occupational strain and job satisfaction: the role of stress,

coping, personality, and affectivity variables. Journal of Vocational Behavior, 54,
429-452.

Friede (2005). Anticipated work-family conﬂict: The construct, its antecedents, and
consequences: Unpublished master ’s thesis, Michigan State University, East
Lansing, MI.

F riede, A., & Ryan, A. M. (2004). The Importance of the Individual: How Self-
Evaluations Inﬂuence the Work-Family Interface. In Kossek, E. & Lambert, S.
(Eds), Work and Life Integration: Organizational, Cultural and Individual
Perspectives (pp. 193-210). Mahwah, NJ : LEA Press.

Friedman, S. D., Christensen, P. & DeGroot, J. Work and life: The end of the zero-sum
game. Harvard Business Review, 1998, November-December, 1 19-129.

Frone, M., Russell, M. & Cooper, L. (1993). Relationship of work-family conﬂict,

gender, and alcohol expectancies to alcohol use/abuse. Journal of Organizational
Behavior, 14, 545-558.

F rone, M. R., Yardley, J. K., & Markel, K. S. (1997). Developing and testing an
integrative model of the work-family interface. Journal of Vocational Behavior,

50, 147-157.

Gianakos I (1999). Career counseling with battered women. Journal of mental health
counseling, 2], 1-14.

Gilbert, L., Dancer, S., Rossman, K., & Thorn, B. (1991). Assessing perceptions of
occupational-family integration. Sex Roles, 24, 107-119.

Greenhaus, J. & Beutell, N. (1985). Sources and conﬂict between work and family roles.
Academy of Management Review, 10, 76-88.

213

Hallet, M. & Gilbert, L. (1997). Variables differentiating university women considering

role-sharing and conventional dual-career marriages. Journal of Vocational
Behavior, 50, 302-322.

Hammer, L., Brockwood, K. & Neal, M. (2001). A systems theory approach to work-
family research: crossover effects on positive work-family spillover. Paper
presented at the Annual Conference of the Society for Industrial and
Organizational Psychology, San Diego, CA.

Judge, T., Erez, A., & Bono, J. (1998). The power of being positive: The relation
between positive self-concept and job performance. Human Performance, 11,
167-188.

Judge, T., Erez, A., Bono, J. & Thoresen, C. (2003). The core self-evaluations scale:
Development of a measure. Personnel Psychology, 56, 303-331.

Judge, T., Erez, A., Bono, J. & Thoresen, C. (2002). Are measures of self-esteem,
neuroticism, locus of control, and generalized self-efﬁcacy indicators of a

common core construct? Journal of Personality and Social Psychology, 83, 693-
710.

Kelloway, K., Gottlieb, B. & Barhman, L. (1999). The source, nature, and direction of
work and family conﬂict: A longitudinal investigation. Journal of Occupational
Health Psychology, 4, 337-346.

Kerpelman, J. & Schvaneveldt, P. (1999). Young adults' anticipated identity importance
of career, marital, and parental roles: Comparisons of men and women with
different role balance orientations. Sex Roles, 41 , 189-217.

Killian, J ., Casper, W., Sitzman, T., Ellsbury, S., Burtrum, M. & Stephens, K. (2003).
Effects of self-efficacy on views of future work/family roles. Poster
presented at the annual meeting for the Society of Industrial and Organizational
Psychology, Orlando, FL.

Kirkpatrick, D. L. (1994) Evaluating training programs: the four levels (San Francisco,
CA, Berrett-Koehler).

Kopelman, R., Greenhaus, J. & Connolly, T. (1983). A model of work, family, and
interrole conﬂict: A construct validation study. Organizational Behavior and
Human Decision Processes, 32, 198-215.

Kossek, E. E., F riede, A. (2006). The Business Case: Managerial Perspectives on Work
and the Family, In Pitt-Catsouphes, M., Kossek, E. E., Sweet, S. (Editors).T he
Handbook of Work-Family: Multi-Disciplinary Perspectives, Methods, And
Approaches. Mahwah, NJ: Lawrence Erlbaum Associates (LEA) Press, 611-627.

214

 

 

 

Larsen, R. J. (1992). Neuroticism and selective encoding and recall of symptoms:
evidence from a combined concurrent-retrospective study. Journal of Personality
and Social Psychology, 62, 480-488.

Lent, R. W., & Hackett, G. (1987). Career self-efﬁcacy: Empirical status and future
directions. Journal of Vocational Behavior, 30, 347-382.

Lent, R., Brown, S. & Hackett, G. (1994). Toward a unifying social cognitive theory of
career and academic interest, choice, and performance. Journal of Vocational
Behavior, 45, 79-122.

Livingston, M., Burley, K. & Springer, T. (1996). The importance of being feminine:
Gender, sex role, occupational and marital role commitment, and their
relationship to anticipated work-family conﬂict. Journal of Social Behavior and
Personality, I I, 179-192.

Lockwood, A. L., Casper, W. J ., Eby, L. T., & Bordeaux, C. (2002). A review of work-
family literature: Where we’ve been and where we need to go. Paper presented in
W. J. Casper (Chair) and J. Barling (Discussant) Symposium, Emerging
Directions in Work and Family Research at the annual American Psychological
Association Conference, Chicago, IL.

Luzzo, D. A., Funk, D. P., & Strang, J. (1996). Attributional retraining increases

career decision-making self-efﬁcacy. The Career Development Quarterly, 44,
378-386.

Mak, A. S., & Mueller, J. (2000). Job insecurity, coping resources and personality
dispositions in occupational strain. Work and Stress, 14, 312-328.

McAuliffe, G. J .(199 I ). Assessing and Treating Barriers to Decision Making in Career
Classes. Career Development Quarterly, 40, 82-92.

McCracken, R. & Weitzman, L. (1997). Relationship of personal agency, problem-
solving appraisal, and traditionality of career choice to women's attitudes toward
multiple role planning. Journal of Counseling Psychology, 44, 149-159.

Mihal, W.L., Sorce, P.A., & Comte, TE. (1984). A process model of individual career
decision making. Academy of Management Review, 9(1), 95-103.

Miller, S. M. & Kirsch, N. (1987). Sex differences in cognitive coping with stress. In R.
C. Barnett, L. Biener, & G. K. Baruch (Eds), Gender and stress (pp. 278-307).

New York: Free Press.

Moyle, P. (1995). The role of negative affectivity in the stress process: Tests of
alternative models. Journal ofOrganizational Behavior, 16, 647-668.

215

Noe, RA. (1986). Trainees’ attributes and attitudes: Neglected inﬂuences on training
effectiveness. Academy of Management Review, I I (4), 736-749.

Parasurman, S., & Simmers, C. A. (2001). Type of employment, work-family conﬂict,
and well-being: A comparative study. Journal of Organizational Behavior, 22,
551-568.

Parkes, K. R. (1990). Coping, negative affectivity, and the work environment: Additive
and interactive predictors of mental health. Journal of Applied Psychology, 75,
399-409.

Paulsen, P. (2001). New era trends and technologies in foreign language Ieaming: An
annotated bibliography. Interactive multimedia electronic journal of Computer-
Enhanced Learning.

Peterson, D. B. (1993). Measuring change: A psychometric approach to evaluating
individual coaching outcomes. Paper presented at the annual conference of the
Society for Industrial and Organizational Psychology, San Francisco.

Phillips, P. C. 1998. Designing experiments to maximize the power of detecting
correlations. Evolution, 52, 251—255.

Phillips, S.D & Imhoff, AR. (1997). Women and career development: A decade of
research. Annual Review of Psychology, 45, 31-59.

Podsakoff, P., MacKenzie, S., Lee, J. & Podsakoff, N. (2003). Common method biases in
behavioral research: A critical review of the literature and recommended
remedies. Journal of Applied Psycho/0g), 88, 879-903.

Rajadhyasksha, U., & Bhatnagar, D. (2000). Life role salience: A study of dual-career
couples in the Indian context. Human Relations, 53, 489-511.

Robbins, S. (1985). Validity estimates for the Career Decision-Making Self-Efﬁcacy
Scale. Measurement and Evaluation in Counseling and Development, 18, 64-71.

Russell, J. E. and Rush, M. C. (1987). A comparative study of age-related variation in
women's views of a career in management. Journal of Vocational Behavior, 30,
280-294.

Salas, E. & Cannon-Bowers, J. A. (2001). The science of training: A decade of progress.
Annual Review of Psychology, 52, 471-499.

Shadish, Cook & Campbell (2001). Experimental and Quasi-Experimental Designs/or
Generalized Causal Inference. Houghton Mifﬂin.

216

 

Schwoerer, C. E., May, D. R., Hollensbe, E. C., & Mencl, J. (2005). General and speciﬁc
self-efﬁcacy in the context of a training intervention to enhance performance
expectancy. Human Resource Development Quarterly, I 6, 111-129.

Sitzmann, T., Kraiger, K., Stewart, D., & Wisher, R. (2006). The comparative
effectiveness of web-based and classroom instruction: A meta-analysis. Personnel
Psychology, 59, 623-664. ’

Stoeva, A. Z., Chiu, R. K., & Greenhaus, J. H. (2002). Negative affectivity, role stress,
and work-family conﬂict. Journal of Vocational Behavior, 60, 1-16.

Tannenbaum, S. I. & Yuk], G. (1992). Training and development in work organizations.
Annual Review of Psychology, 43, 399-441.

Taylor, K.M., and Betz, NE. (1983). Applications of self - efﬁcacy theory to the
understanding and treatment of career indecision. Journal of Vocational Behavior,
37, 17-31.

Taylor, S. E. & Brown, J. D. (1994). Positive illusions and well-being revisited:
Separating fact from ﬁction. Psychological Bulletin, 116, 21-27.

Taylor, S. E., Kemeny, M. E., Reed, G. M., Bower, J. E., Gruenewald, T. M. (2000).
Psychological resources, positive illusions, and health. American Psychologist,
55, 99-109.

Weitzman, L. (1994). Multiple-role realism: A theoretical framework for the process of
planning to combine career and family roles. Applied and Preventive Psychology,
3, 15-25.

Weitzman, L. & Fitzgerald, L. (1996). The development and initial validation of scales
to assess attitudes toward multiple role planning. Journal of Career Assessment,

4, 269-284.

Women in US. academic medicine. (n.d.). Retrieved February 6, 2004, from
http://www.aamc.org/members/wim/statisticslstatsOO/starthtm.

217

 

 

SI ATI

MICHIGAN

(a
E
R
n
R
B
L
Y.
H.
S
R
E
V
m
U

1111111111

lllllll

1293 02956 7074

3