. t K 2
Jr.) .

l D
Evan...

I. E
«a
ﬁfth“?

in. ..

..a:u.n..wa2..m...u

12!»:‘1. 7n.L§?
1&3... 2.

 

.: .1. v.. f
3»..-
‘2 7|!
.‘ .I’AH:

in]!

 

 

}
.t. .
, 4.. c 3.:-
it
.5. .1
.n 5.2! v» “5‘
22.9...

y...

"a

9:}!-

9 C. .21.
11:]! (

.. 3.53:. .;-1.

2a.. 3.9: .v0. H:
I.

36:16.11 31...;

a 2%.”. ..
.uYH . 2.:

2 222%? . . . , £5
. .4;

4.] x...
.2 1

 

mm

coil

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

USING SELF-ENVIRONMENT PERCEPTIONS TO PREDICT
DYNAMIC GOAL COMMITMENT: PERCEIVED MET
EXPECTATIONS AND PERCEIVED P-O FIT

presented by

STEPHANIE M. MERRITT

has been accepted towards fulﬁllment
of the requirements for the

PH.D. degree in PSYCHOLOGY

 

 

/

  

V

Major Professor’s Signature

%/a 7

Date

MSU is an aﬁinnative-action, equal-opportunity employer

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5!08 K:IProiIAoc&Pres/ClRC/DateDm.indd

USING SELF -ENVIRONMENT PERCEPTIONS TO PREDICT DYNAMIC GOAL
COMMITMENT: PERCEIVED MET EXPECTATIONS AND PERCEIVED P-O FIT

By

Stephanie M. Merritt

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

2007

ABSTRACT

USING SELF-ENVIRONMENT PERCEPTIONS TO PREDICT DYNAMIC GOAL
COMMITMENT: PERCEIVED MET EXPECTATIONS AND PERCEIVED P-O FIT

By
Stephanie M. Merritt

In order to understand motivated behavior, many researchers have used goal
setting theory. One of the most persistent ﬁndings in the goal setting literature is that
difficult goals lead to higher performance than do easy goals. However, this ﬁnding is
based on the assumption that goal commitment exists. The present study examines goal
commitment from a person-environment interaction perspective. Existing models of goal
commitment have incorporated mechanical forms of person-environment interaction but
have lacked attention to the dynamic forms of person-environment interaction (Terborg,
1981), including reciprocal interactions and cognitive sensemaking effects. The present
study approaches the prediction of goal commitment from a theoretical perspective
focusing on reciprocal and cognitive sensemaking forms of person-environment
interaction.

Two variables, perceived met expectations and perceived person-organization (P-
0) ﬁt, are employed as new predictors of goal commitment. The focus on perceptions of
met expectations and P-O ﬁt is consistent with the cognitive sensemaking perspective,
which recognizes that different individuals will have different perceptions of the same
environment (i.e., organization). Consistent with the reciprocal interactions approach, I
detail the ways in which the person and the environment may affect each other over time,

thereby resulting in changes in perceptions of met expectations, P~O ﬁt, and goal

commitment. Also consistent with this approach, hypotheses are presented for both
concurrent and longitudinal relationships among variables.

The hypotheses were tested with a three-wave longitudinal data collection.
Respondents were 566 incoming college freshmen at 10 US. colleges and universities. A
mediated latent growth model (LGM) was constructed relating both initial construct
levels and construct change trajectories. The results indicated that Wave 1 levels of
perceived met expectations, perceived P-O ﬁt, and goal commitment were signiﬁcantly
and positively associated; and perceptions of P-O ﬁt fully mediated the relationship
between perceptions of met expectations and goal commitment. For change trajectories,
a different picture emerged. Changes in perceptions of met expectations were
signiﬁcantly and positively associated with changes in P-O ﬁt and changes in goal
commitment; however, there was no evidence for a mediation effect. Overall, the model
accounted for 19% of the variance in Wave 1 goal commitment and 77% of the variance
in changes in goal commitment.

The results suggested that perceived met expectations and perceived P-O ﬁt are
useful predictors of goal commitment. Implications are that goal commitment models
might beneﬁt from additional consideration of reciprocal and dynamic forms of person-
environment interaction. Researchers should also devote increased attention to the
differences between concurrent and longitudinal relationships of predictors with goal

commitment.

To my parents, for all of your love and support

iv

ACKNOWLEDGMENTS

First and foremost, I would like to acknowledge the support and guidance of my
committee chair, Dr. Neal Schmitt, whose expertise and patience has been and continues
to be invaluable. I also thank my committee members, Drs. Dan Ilgen, Fred Oswald, Fred
Leong, and Jim Dulebohn, for challenging me to make this effort the best it can be. It has
been a fantastic journey from ﬁrst draft to ﬁnal product -— thank you for keeping me on
the right path.

I would also like to thank The College Board for its ﬁnancial support in making
this amazing dataset a possibility. To Claire, John, and Joe — thanks for all your hard
work helping us keep the ship aﬂoat. To Alyssa, Anna, Brian, Smriti, and Nancy -
thanks for all the months you spent with me reading, generating measures, building the
surveys, answering participant e-mails, writing tech reports, and of course. . .MERGING
the data. This dissertation could never have been a reality without all of your efforts!

Finally, I would like to thank all of the faculty, students, and staff in the MSU [/0
program for all of the gifts I've received just from being in your presence. I have learned
much from every single one of you, and I feel honored to have had the privilege of

studying with you.

TABLE OF CONTENTS

List of Tables ................................................................................................................... viii
List of Figures .................................................................................................................... ix
Introduction ......................................................................................................................... 1
Person-Environment Interactions .................................................................................... 3
Forms of Person-Environment Interactions ................................................................ 5
Purpose of the Study ....................................................................................................... 7
Study Overview and Roadmap ....................................................................................... 8
Goal Commitment ......................................................................................................... 10
Goal Commitment Measurement .............................................................................. 15
Perceived Met Expectations .......................................................................................... 18
Perceived Met Expectations Measurement ............................................................... 22
Perceived Person-Organization Fit ............................................................................... 25
Perceived P-O Fit Measurement Issues .................................................................... 32
Generalizability to Student Samples ............................................................................. 35
General Study Rationale ............................................................................................... 37
Hypotheses .................................................................................................................... 39
Method .............................................................................................................................. 50
Initial Data Collection and Sample ............................................................................... 50
Wave 1, 2, and 3 Data Collections ................................................................................ 51
Measures ....................................................................................................................... 55
Results ............................................................................................................................... 58
Descriptive Statistics ..................................................................................................... 58
Data Analysis Steps ...................................................................................................... 61
Establishing Measurement Invariance ...................................................................... 61
Univariate LGMs ...................................................................................................... 71
Mediated LGM .......................................................................................................... 73
Overall Model Fit .......................................................................................................... 76
Hypotheses 1 and 2 ....................................................................................................... 76
Hypotheses and & 4 ...................................................................................................... 79
Hypotheses 5 and 6 ....................................................................................................... 79
Variance Accounted For ............................................................................................... 80
Partial Mediation Tests ................................................................................................. 80
Results Summary .......................................................................................................... 82
Relationships with Distal Outcomes and Supplemental Analyses ............................... 83
Discussion ......................................................................................................................... 89
Potential Limitations ..................................................................................................... 97
Suggestions for Future Research ................................................................................ 101
Practical Applications ................................................................................................. 103
Conclusion .................................................................................................................. l 07

vi

Appendix A: Study Information ..................................................................................... 110

Appendix B: Perceived Met Expectations Scale Items ................................................... 112
Appendix C: Perceived P-O Fit Scale Items ................................................................... 113
Appendix D: Goal Commitment Scale Items ................................................................. 114
Appendix E: Correlations of Items/Parcels Used in Mediated LGM ............................. 115
Appendix F: SIMPLIS syntax: Conﬁgural invariance test ............................................. 118
Appendix G: SIMPLIS syntax: Metric invariance test ................................................... 120
Appendix H: SIMPLIS syntax: Uniqueness invariance test ........................................... 122
Appendix I: SIMPLIS command ﬁle for fully-mediated LGM ...................................... 124
References ....................................................................................................................... 1 29

vii

LIST OF TABLES

Table l: 5-Item HWK goal commitment scale ................................................................. 17
Table 2: Sample size, ethnicity, SAT/ACT z-scores, and wave 1 target variables by
school ........................................................................................................................ 54
Table 3: Mean comparisons for ﬁnal sample respondents and non-respondents ............. 55
Table 4: Item-level descriptive statistics ........................................................................... 59
Table 5: Scale means, SDs, alphas, and correlations ........................................................ 60
Table 6: Invariance test summary ..................................................................................... 64

Table 7: Effects of wave 1 perceived met expectations and perceived P-O ﬁt on wave 2
goal commitment over and above institutional afﬁliation ........................................ 85

Table 8: Goal commitment and distal outcome means, standard deviations, and

intercorrelations ........................................................................................................ 88
Table 9: Study information ............................................................................................. 110
Table 10: Correlations of items/parcels used in mediated LGM .................................... 115

viii

LIST OF FIGURES

Figure 1: Hollenbeck & Klein's (1987) Goal Commitment Model .................................. 13
Figure 2: Locke, Latham, & Erez (1988) Model of Goal Commitment ........................... 13
Figure 3: Spreading Activation of Positive Evaluation to Organizaitonal Goals ............. 44
Figure 4: Conceptual Model of Hypotheses ..................................................................... 49
Figure 5: Conﬁgural model of indicators - Construct relationships at all three times ...... 65

Figure 6: Fully Constrained Invariant Model Including Completely Standardized

Common Metric Estimates ........................................................................................ 70
Figure 7: Basic univariate LGM ....................................................................................... 72
Figure 8: Traditional mediation model ............................................................................. 75
Figure 9: Mediated LGM .................................................................................................. 77

Figure 10: Completely standardized path estimates for intercept and slope factors - Fully
mediated LGM .......................................................................................................... 78

Figure 11: Partially mediated model for intercepts including completely standardized path
estimates .................................................................................................................... 81

Figure 12: Partially mediated model for slopes including completely standardized path
estimates .................................................................................................................... 82

Figure 13: Final retained model including completely standardized path estimates ........ 87

Figure 14: Heuristic model - Combined reciprocal inﬂuences and cognitive sensemaking
forms of interaction ................................................................................................... 93

Figure 15: Proposed integration of self-environment perceptions within the Hollenbeck
and Klein (1987) goal commitment model ............................................................... 97

ix

Introduction

Two major outcomes of interest to organizations are task performance and
voluntary turnover. Performance has received perhaps the most attention in the
organizational literature, including a large quantity of work on antecedents of
performance (e. g., Barrick & Mount, 1991; Schmidt, Ones, & Hunter, 1992), the
performance criterion itself (e.g., Austin & Villanova, 1992; Campbell, 1994; Motowidlo,
2003), and the estimation of the dollar value of performance (e. g., Orr, Sackett, &
Mercer, 1989; Roth, Bobko, & Mabon, 2002). Due to the high costs of losing an
employee or student, voluntary turnover is also a concern for business organizations and
colleges alike. Voluntary turnover is expensive both in terms of productivity and
investment lost when an individuals leaves and in terms of the costs of recruiting,
selecting, and training a replacement. Both performance and voluntary turnover are
crucial determinants of organizational success and ultimate survival. Therefore,
understanding the psychological factors contributing to these outcomes is essential.

In order to understand behavior related to these two criteria, many researchers
have employed theories of motivation, including goal setting theory (Locke, 1968; Locke
& Latham, 1990). In fact, it has been suggested that any discussion of motivated
behavior must include goals in some capacity (Austin & Vancouver, 1996). One of the
most persistent ﬁndings in the goal setting literature is that difﬁcult goals lead to higher
performance than do easy goals (e.g., Locke, Shaw, Saari, & Latham, 1981). However,
this ﬁnding is based on the assumption that commitment to goal attainment exists (Locke,

Latham, & Erez, 1988). Without commitment, goal level has no effect on performance

(Klein, Wesson, Hollenbeck, & Alge, 1999). Therefore, understanding goal commitment
is important for understanding and improving job performance.

Some researchers have also examined the role of goal commitment in the
voluntary turnover process. Traditionally, much of this research has focused on college
student persistence — ﬁnding inverse relationships between commitment to the goal of
obtaining a college degree and dropout (e. g., Getzlaf, Sedlaceck, Kearney, & Blackwell,
1984; Pascarella & Terenzini, 1983; Tinto, 1975). More recently, researchers have
examined the inﬂuence of goal commitment on the attitudes of organizational newcomers
(e.g., Maier & Brunstein, 2001). Because attitudes such as satisfaction and
organizational commitment are thought to be central predictors of voluntary turnover
(e.g., March & Simon, 1958; Mobley, 1977; Tinto, 1975), understanding the relationship
of goal commitment with these attitudes is important to improving our understanding of
the turnover process.

Given that goal commitment has been linked with such essential organizational
outcomes as performance and voluntary turnover, it is crucial that researchers be able to
predict and inﬂuence it. The key to doing so is identifying and understanding its
antecedents. Some empirical support has been provided for models that propose both
person and situation predictors of goal commitment (Hollenbeck & Klein, 1987; Klein, et
al., 1999; Locke, et al., 1988), suggesting that variables related to person—environment
interactions may be important inﬂuences on goal commitment. However, past research
has been limited in that it tends to focus on additive or statistical interactions between
individual and situational antecedents but fails to account for more complex types of

relationships between these two categories of factors.

The aim of the present study is to expand existing models of goal commitment to
more explicitly consider the results of embedding a person within a particular context. I
propose that variables that reﬂect an individual’s perception of him or herself with a
particular situation, speciﬁcally perceptions of person-organization (P-O) ﬁt and
perceptions of met expectations, can be employed to predict goal commitment. I begin
with a brief discussion of ﬁve major forms of person-environment interactions.
Following that, I review the literature on goal commitment, perceived met expectations,
and perceived P-O ﬁt.

Person-Environment Interactions. One of the oldest debates in the history of
psychology concerns whether a person’s behavior is a function of his or her innate traits
(person factors) or a function of his or her environmental context (situation factors).
Most researchers acknowledge that both perspectives are to some extent correct (e. g.,
Caldwell & O’Reilly, 1990; Chatman, 1989; Donavan, Fang, Bendapudi, & Singh, 2004;
Judge & Kristof-Brown, 2004; Lewin, 1951; Mischel, 1977). Different people may
behave differently in the same situation, and the same person may demonstrate little
cross-situational consistency in behavior. One can derive from this compromise position
the proposition that each person/situation combination is unique, and each unique
person/ situation combination may elicit attitudes and behaviors different from any other
person/situation combination. I refer to each individual person/situation combination as a
“contextualized person.”

The distinctions between contextualized persons may become more or less
pronounced over time. When a person is embedded in a situation, the person and the

environment exert reciprocal inﬂuences on one another that, over time, result in emergent

and ﬂexible psychological states. In other words, characteristics of the situation
inﬂuence the person’s cognitions, attitudes, and behaviors, and at the same time, the
person becomes an active element in the situation (e. g., Terborg, 1981). Cumulatively,
these reciprocal relationships inﬂuence the psychological outcomes of the contextualized
person, including perceptions and behaviors. Because this relationship is dynamic and
reciprocal, it is expected that any predictors and outcomes of interest might exhibit
change over time — either within persons, between persons, or both.

The notion that psychological outcomes, including behavior, are a function of the
person/situation combination has been broadly accepted. For example, the ﬁelds of
interactional and environmental psychology are founded on the notion that an
individual’s behavior is a function of both the person and the situation (Caldwell &
O’Reilly, 1990; Chatman, 1989; Donavan, Fang, Bendapudi, & Singh, 2004; Lewin,
1951; Mischel, 1977; Stewart & Barrick, 2004). Researchers within the career
counseling and stress and coping domains have also acknowledged the importance of
studying the relationships between the person and environment, including the ﬁt between
an individual’s characteristics and environmental demands (e.g., Moos & Swindle, 1990;
Swindle & Moos, 1992; White, 1952). Complexity Theory endorses the notion that a
dynamic environment exerts continually-changing demands upon the individual, and the
individual responds to those demands in ways consistent with his or her internal
characteristics (Bar-Yam, 2002). A focus on the joint effects of person and situation is
also not new within the ﬁeld of organizational behavior, as person-environment
interactions have been acknowledged by several researchers in the past (e. g., Chatman,

1989; Judge & Kristof-Brown, 2004; Roberts, Hulin, & Rousseau, 1978).

Forms of Person-Environment Interactions

Five major forms of person-environment interactions have been identiﬁed:
statistical interactions, mediation effects, independent direct effects, reciprocal effects,
and cognitive sensemaking effects (Terborg, 1981). The ﬁrst two forms, statistical
interactions and mediation effects, refer to the multiplication of a person variable and a
situation variable in order to summarize their non-additive effects. According to these
types of interactions, the effects of the person variable on the outcome depend on the
level of the situation variable, or vice versa. The third type of person-environment
interaction is independent direct effects, wherein the person and situation variables both
affect same outcome directly, but are unrelated to each other. In this case, both variables
have additive, linear effects in the prediction of the outcome variable. Due to the heavy
focus on data analytic techniques rather than theoretical mechanisms for person-
environment relationships, Terborg (1981) refers to these three forms of person-
environment interaction as “statistical and mechanistic” (pp. 572).

As opposed to the mechanical forms of interaction discussed above, the ﬁnal two
forms are referred to as “dynamic” due to their theoretical focus on the continuous and
complex interactions between a person and his/her environment (Terborg, 1981, pp. 572).
The fourth form of interaction, reciprocal interactions, incorporates the complex ways in
which individuals and their environments affect each other over time. Environments
affect the attitudes and behaviors of individuals, and those individuals in turn affect the
nature of their environments. From this perspective, individuals, situations, and
behaviors are mutually interdependent over time. Research undertaken from the

perspective of reciprocal person-environment interactions, therefore, should be

longitudinal in nature, as consideration of the ways in which variables, and their
relationships, change over time is essential (Terborg, 1981).

The ﬁnal form of person-environment interaction, the cognitive sensemaking
approach, refers to the notion that different individuals will perceive and interpret the
same situation in different ways (Terborg, 1981). In this sense, between-person
disagreement on the nature of a situation (e.g., organizational climate), should not
necessarily be viewed as error, but instead should be incorporated as a construct of
interest. Research from a cognitive sensemaking approach would tend to focus on
individual differences in perceptions of situations, rather than more “objective” measures
of situational differences.

In previous goal commitment research, when person-environment interactions
have been considered, they have been approached primarily from the mechanical
standpoint. Two major models of goal commitment have incorporated the notion of
independent direct effects, and some research has incorporated the notions of mediation
effects and statistical interactions (e.g., Hollenbeck & Klein, 1987; Klein & Wright,
1994; Locke, Latham, & Erez, 1988). The present study addresses whether dynamic
conceptualizations of person-environment interaction might also inform our
understanding of goal commitment. The present study thus focuses exclusively on
variables conceptualized from the standpoint of reciprocal and cognitive sensemaking
interactions.

In order to approach goal commitment from the dynamic interaction standpoint, I
focus on individual differences in perceptions of the environment (cognitive sensemaking

interactions) rather than on objective features of the environment. In formulating theory

and hypotheses, I focus explicitly on the ways in which the person and his/her
environment may affect each other over time, and I measure the perceptual predictors and
goal commitment longitudinally. Therefore, theoretical and empirical attention is
devoted to the ways in which these variables may change and relate over time (reciprocal
interactions).

Purpose of the Study

Two variables novel in the goal commitment literature, perceived met
expectations and perceived person-organization (P-O) ﬁt, will be examined as potential
predictors of goal commitment. These two variables reﬂect one’s perceptions of the self-
environment combination, and thus, they reﬂect a cognitive sensemaking approach to
person-environment interaction. Perceived met expectations refers to one’s perception of
the extent to which one’s expectations prior to organizational entry (person factor) are
compatible with what is actually experienced in the organizational environment (situation
factor) at a given point in time. Perceived P—Oﬁt refers to one’s perception of the extent
to which some personal characteristic (person factor) is similar to, or compatible with, the
characteristics of the organization in which one is embedded (situation factor).

These two constructs are similar in that both reﬂect the perceptions of a
contextualized person regarding some aspect of his or her unique person/ situation
combination. Thus, perceived met expectations and perceived P-O ﬁt are both
conceptualized as perceptions that are relevant to both the self and a speciﬁc environment
(i.e., organization). It has been well-established that it is often one’s perceptions and
interpretations of a situation, which are usually only partially based on objective reality,

that affect attitudes and behavior (e. g., Cable, Judge, Boudreau, & Bretz, 1995; Fiske &

Taylor, 1991; Kristof-Brown, Zimmerman, & Johnson, 2005; Nisbett & Ross, 1980).
Therefore, knowledge of the perceptions of the contextualized person may be beneﬁcial
for accurate prediction of the person’s goal commitment within a particular situation,
such as one’s organizational environment.

As will become evident, these conceptual similarities between perceived met
expectations and perceived P-O ﬁt have led to parallel debates in the met expectations
and P-O ﬁt literatures regarding appropriate measurement strategies for each (e. g.,
Edwards, 1991; Irving & Meyer, 1994; 1995). These measurement debates, and their
relationships to construct deﬁnitions, will be addressed in more detail during the
literature review.

Study Overview and Roadmap

The purpose of the present study is to examine the extent to which perceived met
expectations and perceived P-O ﬁt predict goal commitment for organizational
newcomers. Thus, this investigation contributes to our existing knowledge of goal
commitment by considering perceptions related to aspects of the speciﬁc self/situation
combination relevant to the goal. I use a longitudinal sample of 566 incoming freshman
students at 10 US. colleges and universities. (Several parallels are later drawn between
college students and organizational employees in order to argue in favor of the
appropriateness of this student sample). Respondents provided data on their perceptions
of met expectations, perceptions of P-O ﬁt, and commitment to the goal of obtaining their
college degree at three time points over their ﬁrst three academic semesters.
Relationships among these constructs were analyzed using a mediated latent growth

model (LGM). LGM is a data analytic technique that expands on traditional regression-

based analyses, allowing the researcher to examine the signiﬁcance of the relationships
between variables’ change trajectories as well as the signiﬁcance of Wave 1 concurrent
relationships. This technique is valuable in longitudinal data analysis; however, it is
important to note that it is correlational in nature. Therefore, while LGM provides an
assessment of the extent to which changes in variables are signiﬁcantly related, the reader
should keep in mind that it does not allow causal inferences to be made.

I begin with a brief literature review of the three major constructs involved in the
present study: goal commitment, perceived met expectations, and perceived P-O ﬁt. The
reviews of perceived met expectations and perceived P-O ﬁt will include theoretical
issues such as the distinction between the “objective” and “perceived”
conceptualizations, and methodological issues related to that distinction. Following that
literature review, I present several hypotheses. One of the basic tenants of the reciprocal
form of person-environment interaction is that the person and situation inﬂuence one
another reciprocally and dynamically over time. Research undertaken from this
perspective must be not only longitudinal, but it must specify mechanisms for change
(Terborg, 1981). Therefore, hypotheses are presented in reference to both concurrent
relationships among variables and relationships in the ways in which these variables
change over time.

Following the presentation of the hypotheses, I describe the study sample and
procedure, followed by a presentation of the results, which will include descriptive
statistics, preliminary analyses, and LGM analyses testing the hypotheses presented.
Finally, I discuss implications of the study, potential limitations, and suggestions for

future research.

Literature Review
Goal Commitment

In the past two decades, goal setting theory has become one of the preeminent
theories of human motivation (Austin & Vancouver, 1996; Locke, et al., 1988). People
in organizations may have a variety of different types of goals, including goals such as
getting a promotion, learning new skills, or being well—liked. In institutions of higher
learning, many students have the goals such as being on the Dean’s list, being elected to
student government, or successfully earning their degree.

Regardless of the nature of the goal content, one of the major ﬁndings from goal
setting research has been that having a speciﬁc, difﬁcult goal leads to higher performance
than having an easy goal. However, this effect is dependent on one’s goal commitment.
If one is not committed to achieving the goal, the goal has no effect on behavior (Locke,
et al., 1988). For this reason, goal commitment has become a major focus of research on
goal setting (Klein & Wright, 1994).

Goal commitment was originally defined by Locke (1968) as one’s resistance to
changing one’s goal over time; however, the accepted deﬁnition has become broader as
research has progressed (Donovan & Radosevich, 1998). In their deﬁnition, Campion
and Lord (1982) focused on the extension of effort over time as well as an unwillingness
to change the goal. Locke and Latham (1990) deﬁned goal commitment as one’s
determination to reach a goal, and later, Hollenbeck and colleagues (Hollenbeck & Klein,
1987; Hollenbeck, Williams, & Klein, 1989) deﬁned goal commitment as one’s

determination to pursue a goal and persistence in goal pursuit over time.

10

More recently, DeShon and Landis (1997) deﬁned goal commitment as, “the
degree to which the individual considers the goal to be important, is determined to reach
it by expending effort over time, and is unwilling to abandon or lower the goal when
confronted with setbacks and negative feedback” (pp. 106). Note that this deﬁnition
reﬂects several common themes in previous goal setting literature (e. g. Campion & Lord,
1982; Keman & Lord, 1988; Latham & Locke, 1991; Locke, Latham, & Erez, 1988;
Naylor & Ilgen, 1984; Riedel, Nebeker, & Cooper, 1988; Tubbs, 1993; Wright, O’Leary-
Kelly, Cortina, Klein, & Hollenbeck, 1994) and incorporates attitudinal and behavioral
components.

Some researchers have argued that goal commitment should be distinguished
from goal acceptance, which is deﬁned as initial agreement that the goal is an adequate
performance standard (Hollenbeck & Klein, 1987; Tubbs & Ekeberg, 1991). This
distinction is not always supported empirically (Locke & Latham, 1990; Rem,
Danehower, Swiercz, & Icenogle, 1999). Some scholars have asserted that acceptance is
a special case of commitment to an assigned goal (Locke et al., 1988; Locke, et al.,
1981), and several researchers consider goal commitment and goal acceptance to be
synonymous (e.g., Hollenbeck & Klein, 1987; Locke, et al., 1988; Tubbs & Dahl, 1991).
Until more conclusive evidence is presented indicating that goal acceptance and goal
commitment are, in fact, separate constructs, it is reasonable to continue the existing
tradition of considering these constructs to be equivalent.

The key to understanding, predicting, and inﬂuencing goal commitment is the
identiﬁcation of its antecedents. Two major models of the predictors of goal commitment

have been proposed. The ﬁrst, and most inﬂuential, model of the goal commitment

11

process was proposed by Hollenbeck and Klein (1987) and is displayed in Figure 1. In
this model, goal commitment is conceptualized as stemming from two proximal
antecedents based in the expectancy theory framework: the attractiveness of goal
attainment and the expectancy of goal attainment. Each of these proximal antecedents is
proposed to be affected by the distal personal and situational antecedents, as displayed.

Many of these propositions have received empirical support (Hollenbeck, Klein,
O’Leary, & Wright, 1989). A recent meta-analysis (Klein, et al., 1999) found that
attractiveness, expectancy, and their interaction (termed motivational force) were
signiﬁcantly related to goal commitment. In addition, signiﬁcant relationships were
found between goal commitment and several of the proposed distal antecedents,
including ability, volition/participation, affect, goal speciﬁcity, task experience, feedback
provision, and feedback type. However, these relationships should be interpreted
cautiously, as the number of studies examining the distal variables was relatively small
(Klein, et al., 1999).

A second model of the goal commitment process was proposed by Locke and
colleagues (1988). Like the Hollenbeck and Klein (1987) model, this framework
incorporates internal (person) and external (situation) predictors of goal commitment.
This model is displayed in Figure 2. Several of the predictors proposed in this model are
also included in the Hollenbeck model, such as peer group inﬂuence, rewards and
incentives, self-efﬁcacy, and expectancy. Although the Locke, et a1. (1988) model has
inspired somewhat less empirical work than has the Hollenbeck and Klein (1987) model,
support for the utility of these factors was provided by the Klein, et a1. (1999) meta-

analysis.

12

Figure 1
Hollenbeck & Klein’s (1 98 7) Goal Commitment Model

 

 

 

Situational Personal Situational Personal
Factors Factors Factors Factors

Volition Endurance Social inﬂuence Ability

Explicitness Type A Task complexity Past success

Reward Personality Performance Self-esteem

structures Org. Commit constraints Locus of

Competition Job Supervisor control

Involvement supportiveness

 

 

 

 

 

 

 

 

 

 

 

V V

[ Attractiveness of Goal Attainment Expectancy of goal attainment I

Task performance

 

 

 

   

Goal level

Figure 2
Locke, Latham, & Erez (1988) Model of Goal Commitment

 

 

 

 

 

 

 

 

 

 

 

External Goal

Factors E“ Commitment
Authority a
Peer group 25’
inﬂuence 5:

External .2 r
rewards& El; Performance
incentives o I

U 1
Interactive
M Internal Factors
Pamelpation ——Expectancy Goal
Competition S elf- e fﬁ c acy Content
Internal rewards

 

 

 

 

 

Both the Hollenbeck and Klein (1987) and Locke, et al. (1988) models

acknowledge that person-environment interactions may affect goal commitment. Both

incorporate the direct effects of person and situation factors, reﬂecting the “independent
effects” form of interaction (Terborg, I981). The Locke, et al. (1988) model also
suggests elements of other types of interaction. For example, under the category of
“interactive factors,” the model includes the extent to which the individual is permitted to
participate in the setting of the goal and the extent to which there is competition for goal
attainment. These factors are considered interactive in the sense that the relationship of
participatory goal setting with goal commitment will depend on whether the individual’s
cultural values and/or desire to participate in goal setting, and the effect of competition
will depend on whether the individual desires competition (Klein & Wright, 1994; Locke,
et al., 198 8). These factors therefore reﬂect the “statistical interaction” form of person-
environment interaction (Terborg, 1981).

The Locke, et al. model also brieﬂy acknowledges the importance of cognitive
processing. This is consistent with Terborg’s emphasis on the importance of the
individual’s cognitive interpretation of the situation (Terborg, 1981). Although Locke, et
al.’s model incorporates the “cognitive processing” box, the authors say little about the
meaning or content of that box, and little subsequent attention has been paid to this more
subjective element of the goal commitment process. The present study aims to address
this gap in the literature by examining the extent to which perceptions of the
contextualized person can be employed to predict goal commitment. More speciﬁcally, I
focus on perceptions of met expectations and perceptions of P-O ﬁt.

I proceed with the caveat that the more traditional predictors of goal commitment,
including individual measures of person and situation factors, are not measured in the

present study. Therefore, the extent to which perceived met expectations and perceived

l4

P-O ﬁt might predict additional variance in goal commitment over and above these more
traditional predictors cannot be determined here. The purpose of the present study is
simply to assess whether perceptions of the person/ situation combination, speciﬁcally
perceived met expectations and perceived P-O ﬁt, might be useful predictors of goal
commitment in and of themselves.

Goal Commitment Measurement. A 9-item self-report measure was developed
by Hollenbeck and colleagues (Hollenbeck, et al., 1989; henceforth referred to as HWK).
Subsequently, Tubbs and colleagues (Tubbs & Dahl, 1991; Tubbs, 1993) argued that self
report measures of goal commitment in general, and the HWK scale in particular, require
caution. They argued instead that goal commitment should be assessed as the
discrepancy between assigned goals and self-set goals.

Research on this issue has shown that self-report and discrepancy measures of
goal commitment are typically only weakly related (Donovan & Radosevich, 1998;
Tubbs, 1993; Wright et al., 1994), suggesting that there may be differences in the
constructs being measured. It has been suggested that self-report measures of goal
commitment might actually assess three concepts: motivational force, choice of a
personal goal, and maintenance of that choice (Tubbs, 1993; Tubbs & Ekeberg, 1991).
As described above, motivational force is conceptualized as an antecedent of goal
commitment rather than a component of goal commitment itself (Hollenbeck & Klein,
1987; Klein, et al., 1999). Therefore, this claim posed a serious threat to the validity of
direct measures of goal commitment, and Tubbs and colleagues (1991) suggested that the
discrepancy between commitment to a personal goal and an assigned goal be used to

measure goal commitment.

15

The Hollenbeck camp disputed Tubbs and colleagues’ suggestion that the
discrepancy measure of goal commitment should be used (Wright, et al., 1994). They
argued persuasively that discrepancy-based measurement of goal commitment is
inappropriate for four theoretical reasons. First, Wright, et al. (1994) argued that
discrepancy-based measurement is inconsistent with the deﬁnition of goal commitment as
extension of effort over time and unwillingness to abandon a goal. Whereas items in a
self-report measure can directly address this deﬁnition, researchers using the discrepancy
measure must infer these characteristics. Second, they argued that discrepancy—based
measurement is problematic in that discrepancy measures are highly related to their
components. It has been found that the discrepancy between personal and assigned goals
is correlated so highly with personal goal levels that they are almost indistinguishable (rs
= .79 and .91; Tubbs & Dahl, 1991). Third, Wright, et al. (1994) pointed out that the
discrepancy measure as deﬁned by Tubbs and colleagues does not consider the sign of
the deviation, effectively equating positive and negative discrepancies. This, they point
out, is contradictory to several motivation theories that argue that efforts to reduce
negative discrepancies differ from efforts to create positive discrepancies (e. g., control
theory and self-efﬁcacy theory). Finally, they propose that the arguments in favor of the
use of a discrepancy measure inappropriately over-emphasize effect size ﬁndings relative
to theory as the standard for construct validity. The four points presented above strongly
argue in favor of the use of self-reports measures of goal commitment as opposed to a
discrepancy-based measure.

Since the arguments presented by Wright, et al., (1994), the HWK scale has

become the most frequently-used method of measuring goal commitment (DeShon &

l6

Landis, 1997; Klein, et al., 1999; Klein, Wesson, Hollenbeck, Wright, & DeShon, 2001).
This scale was developed by compiling several single-item measures of goal commitment
that were in use at the time of the scale’s conception and writing additional items
(Hollenbeck, et al., 1989). The items selected and written for this scale, therefore, were
based on the deﬁnitions of goal commitment in use during the 19803. Despite the
theoretical foundation and popularity of this scale, some measurement controversy has
persisted. Speciﬁcally, while some researchers presented evidence suggesting that the 9-
item HWK scale was unidimensional (Hollenbeck, et al., 1989; Wright, et al., 1994),
other evidence suggested that the scale might in fact represent two factors (DeShon &
Landis, 1997; Tubbs & Dahl, 1991). Authors from both camps collaborated in order to
identify a 5-item version of the HWK scale with strong evidence for unidimensionality
(Klein, et al., 2001). These ﬁve items are displayed in Table 1, and this version of the
HWK scale is the most recommended method for measuring goal commitment.

Table l
5-Item H WK Goal Commitment Scale

 

It’s hard to take this goal seriously (R)

Quite frankly, I don’t care if I achieve this goal or not (R)
I am strongly committed to this goal

It wouldn’t take much to make me abandon this goal (R)
I think this is a good goal to shoot for

.V‘PE’JNT‘

 

Summary. Goal commitment has been found to predict two of the most important
organizational outcomes: performance and voluntary turnover. Therefore, predicting and
understanding goal commitment is essential. Although several person and situation
factors have been linked to goal commitment via direct effects, the perceptions arising
from the dynamic and reciprocal relationship between a person and his/her context have

been neglected in the goal commitment literature. I next discuss two such perceptions

17

that are examined in the present study: perceptions of met expectations and perceptions
of person-organization ﬁt.
Perceived Met Expectations
Typically, the voluntary turnover rate tends to be higher for new
organizational members than for those with longer tenure, regardless of whether
the organization in question is business-related or academic in nature (e. g., ACT,
2004; Gerdes & Mallinckrodt, 1994; Helland, Stallings, & Braxton, 2001-2002;
Mallinckrodt & Sedlacek, 1987; Shaw, 1968). Also regardless of the type of
organization, research has suggested that individuals’ pre-entry expectations
about various aspects of organizational membership are often unrealistically high
(Haynes, Ruthig, Perry, Stupinsky, & Hall, 2006; Wanous, 1977). The
combination of these ﬁndings has led to the met expectations hypothesis, which
states:
The concept of met expectations may be viewed as the discrepancy
between what a person encounters on the job in the way of positive and
negative experiences and what he [sic] expected to encounter. Thus, since
different employees can have quite different expectations with respect to
payoffs or rewards in a given organizational or work situation, it would
not be anticipated that a given variable (e. g., high pay, unfriendly work
colleagues, etc.) would have a uniform impact on withdrawal decisions.
We would predict, however, that when an individual ’s expectations —

whatever they are — are not substantively met, his [sic] propensity to
withdraw would increase. (Porter & Steers, 1973, pp. 152)

According to this met expectations hypothesis, the negative effects of an unfavorable
organizational situation, such as an unjust supervisor or professor, will be attenuated if
the individual has realistically negative expectations for fairness. Conversely, the
positive attitudinal effects of bonus pay or a scholarship will be attenuated if the

individual had expected a larger bonus or scholarship than received.

18

Since this original statement, researchers have acknowledged that the direction of
the discrepancy from expectations is meaningful. More speciﬁcally, a positive
discrepancy (i.e., experiences exceed what was expected) is considered to be different
than a negative discrepancy (i.e., experiences fall short of expectations). Met
expectations, then, can fall on a linear scale ranging from “unmet expectations” to “met
expectations” to “exceeded expectations” (e.g., Major, Kozlowski, Chao, & Gardner,
1995). Therefore, although the term used here is “met expectations,” my
conceptualization of the construct includes the notion that expectations can be exceeded,
and exceeded expectations are considered to be more positive than expectations that are
simply met.

Research on met expectations has generally found signiﬁcant relationships
between the extent to which pre-entry expectations have been met or unmet and several
attitudes and behaviors, including turnover. A meta-analysis by Wanous, Poland,
Premack, and Davis (1992) directly examined the met expectations hypothesis and found
that across a sample of 31 studies and 17,241 participants, met expectations were
positively related to a number of attitudes and behaviors, including job satisfaction (r =
.39), organizational commitment (r = .39), intent to remain with the organization (r =
.29), and turnover (r = -.19).

This consistent ﬁnding linking met expectations to attitudes and turnover has
inspired practical applications, such as realistic job previews (RJPs). RJPs are
interventions designed to provide job applicants with more realistic expectations for what
actual work experiences will entail. Based on research suggesting that the expectations

of incoming organizational members are usually inﬂated (Wanous, 1977), RJPs typically

19

provide applicants with realistically negative information about the job in order to deﬂate
their unrealistically high expectations prior to acceptance or organizational entry. Several
reviews of the literature on RJPs have been conducted (e.g., Premack & Wanous, 1985;
Wanous, 1977; Wanous & Colella, 1989). In general, these reviews have found that RJPs
are successful in lowering applicant expectations and that RJPs are linked with decreased
intent to turnover and improved attitudes in organizational newcomers.

In the present study, I am concerned with perceptions of met expectations. I
deﬁne perceived met expectations as the extent to which an individual believes that the
expectations he or she had prior to organizational entry have been met. Note then, that a
person’s perception of met expectations may differ from any objective measure of the
degree to which pre-entry expectations have been met. This perspective is somewhat
different from the way in which met expectations has been characterized in the past.
Traditionally, met expectations has been viewed as a discrepancy-related construct that
remains relatively ﬁxed upon organizational entry (e. g., Dabos & Rousseau, 2004; Horn,
Griffeth, Palich, & Bracker, 1998; Ilgen & Seely, 1974; Major, et al., 1995). When we
consider perceived met expectations, a more complex picture is revealed. This picture
involves a construct representing a person’s perceptions regarding the context in which
he or she is embedded, regardless of whether those perceptions are true to organizational
reality. This focus on perceived met expectations is consistent with the cognitive
sensemaking approach to person-environment interactions, which suggests that
individuals will interpret the same situation in different ways (Terborg, 1981).

In addition, one’s perceptions of met expectations may change over time. As

suggested for researchers approaching person-environment interactions from the

20

reciprocal interactions standpoint (Roberts, et al., 1978; Terborg, 1981), I specify several
theoretical mechanisms by which perceptions of met expectations may change over time.

Changes in perceptions of met expectations might very well reﬂect real changes
occurring in the organization. For example, an employee’s job requirements or group of
coworkers might change, altering the experiences portion of the comparison. Parallels in
an academic setting might be a new semester of coursework or new roommates.
Regardless of whether the context is business or academic in nature, to the extent that
people perceive that the changes have occurred, their perceptions of met expectations are
likely to change. Furthermore, because people are active agents in their environments,
they may be able to actively change the nature of their environment to more closely meet
their expectations, which is also likely to affect perceptions of met expectations.

Alternatively, changes in perceptions of met expectations may occur
independently of any real changes in the organizational context. For example,
organizational socialization processes could affect the person’s perception of the extent
to which the situation is discrepant from expectations. These changes in perceptions
could occur without a change in any objective aspect of the person or the situation. Over
time, the inﬂuence of the situation may even change the person’s memory of what his or
her initial expectations were prior to organizational entry, thereby affecting perception of
the degree to which those expectations have been met.

In summary, as a perception, perceived met expectations represents the
individual’s cognitive interpretation of the complex, reciprocal, and dynamic interplay of
person and situation. Changes in these perceptions over time may or may not reﬂect

actual changes in the degree to which one’s pre-entry expectations have truly been met.

21

The focus on perceived met expectations has implications for an ongoing discussion of
measurement issues related to the met expectations construct (e. g., Irving & Meyer, 1994,
1995)

Perceived Met Expectations Measurement. Drawing on a critique of the
methodology used in the literature on person-job ﬁt (Edwards, 1991), Irving and Meyer
(1994) presented a critique of several methods used to assess met expectations, including
both discrepancy-based and direct measures. Irving and Meyer contend that direct
measures of met expectations, in which respondents directly report the degree to which
their expectations have been met by actual work conditions, are problematic in that
researchers are unable to assess the separate and combined effects of pre-entry
expectations and post-entry experiences. In other words, they argue that when using
direct measures, researchers are unable to say with conﬁdence that it is met expectations,
rather than only expectations or only organizational experiences, that affects the outcome
of interest. In addition, when direct measures are used, the effects of expectations and
experiences are weighted by each individual who responds to the survey items according
to his or her individualized preferences. Therefore, a standardized weighting of the
effects of each component is not possible using the direct measurement method.

However, in the present study, I do not hypothesize any relationships involving
“objective” met expectations. Instead, I focus on perceptions of met expectations, which
is a subjective construct. Given this perspective, it makes theoretical and practical sense
to use direct measures of perceived met expectations. From the reciprocal and cognitive
sensemaking interaction perspectives, no prescriptive scheme exists for the weighting of

the person and situation effects. Instead, individual differences should exist in both the

22

importance of various expectations relative to one another and in the effects of the degree
of perceived violation of any particular expectation. Prior research suggests that people
weight some dimensions of their experiences more heavily than others when rating their
perceptions of met expectations (Irving & Meyer, 1995). These differences are consistent
with the cognitive sensemaking interactional perspective and therefore, direct measures
are appropriate when measuring perceived met expectations. From this standpoint,
individual differences in weighting are theoretically expected and should not be viewed
as measurement error but should instead be considered part of the construct variance.

The focus on the perception of met expectations also eliminates a second potential
cause for worry when direct measures are used: memory biases. Some scholars have
raised concerns regarding the potential for memory distortions to affect direct ratings of
met expectations (Irving & Meyer, 1995). However, when one views met expectations as
a perception, this concern become less troubling. Because the interest here is in perceived
met expectations, accurate recollection of expectations prior to organizational entry is not
necessary. Instead, the ways in which a person’s organizational experiences affect his/her
memory as time passes (which might occur naturally or through the socialization process)
become theoretically interesting. The psychological state that is characterized by the
perceived met expectations variable is a perception of reality, and while this perception is
almost certainly inﬂuenced by objective reality, it is the perception and not the reality
that is of theoretical interest in the present study.

Finally, the reciprocal interactive perspective on perceived met expectations
allows for the possibility that perceptions of met expectations could be, and should be,

measured repeatedly over time. A study that is theoretically based on the reciprocal

23

inﬂuences form of interaction should conceptualize relationships in a longitudinal sense
(Terborg, 1981). As discussed previously, changes in perceptions of met expectations
over time could theoretically result from changes in actual organizational experiences,
changes in perceptions independent of actual changes in experiences, or both. Such an
interest in changes in perceptions of met expectations essentially demands direct
measurement of met expectations, as is demonstrated by the following scenario. Imagine
a situation in which a researcher collected pre-entry expectations data and a series of
three waves of post-entry work experiences measures. Assume for the purpose of this
example that equal weighting of expectations and experiences is valid, and the researcher
calculates discrepancy scores between pre-entry expectations and actual experience at
each wave of data collection. Because the pre-entry expectations remain constant over
time, the discrepancy score essentially reﬂects only changes in experiences. In this case,
the discrepancy method of measuring met expectations would reﬂect actual work
experiences subtracted from a constant and would be essentially useless in a longitudinal
data collection. Instead, the reciprocal inﬂuences approach demands that perceptions of
met expectations be measured repeatedly so that changes in the construct can be assessed.
Summary. Perceived met expectations represent the degree to which one believes
that his or her pre-entry expectations have been met. These perceptions are subjective and
may or may not reﬂect “objective” met expectations. Although concerns have been raised
regarding the construct validity of direct measures of met expectations (Irving & Meyer,
1994; 1995), when researchers are interested in perceptions of met expectations, direct
measures are theoretically and practically appropriate. The focus on perceptions of met

expectations allows for individual differences in the weighting of various expectations

24

relative to each other as well as for individual differences in the weighting of the effects
of violations of these expectations.

The second potential predictor of interest in the present study is perceived person-
organization (P-O) ﬁt. The concerns raised regarding direct measurement of P-O ﬁt
parallel those presented against direct measurement of met expectations. Similarly, I will
argue that because the focus in the present study is on perceptions of P-O ﬁt, the
construct is best assessed via direct measurement.

Perceived Person-Organization Fit

Consistent with the interactive perspective, the concept of ﬁt stems from the idea
that many outcomes are a function of the combination of characteristics of a person and
characteristics of his or her environment (Chatman, 1989). The person-organization (P-
0) ﬁt construct reﬂects the extent to which a person is compatible with the characteristics
of his or her organization. Thus, it considers the combination of a person’s
characteristics and the characteristics of his/her organizational environment. P-O ﬁt has
been distinguished from other forms of person-situation ﬁt, such as person-group ﬁt,
person-job ﬁt, and person-vocation ﬁt (Kristof, 1996) and has been conceptualized in
several ways, including similarity, need-satisfaction, and demand-ability match (Kristof-
Brown, et al., 2005).

In the past decade, a general consensus has emerged that there are two general
types of P-O ﬁt. The ﬁrst type, supplementary congruence, refers to the extent to which
the person and the organization are similar on some dimension (Kristof, 1996). For
example, supplementary congruence would occur if the person and the organization as a

whole both valued diversity, both preferred competition to collaboration, or both

25

endorsed work-life balance. The second type of P-O ﬁt, complementary congruence,
refers to the extent to which the person and the organization contribute to one another’s
needs (Kristof, 1996). Complementary congruence does not imply similarity, but instead
denotes a symbiotic relationship between person and organization. For example, imagine
a person who is highly creative but weak on ability to follow through on his or her ideas.
An organization might provide this person with a structured environment in order to
facilitate follow through. Complementary congruence would be achieved in that the
organization beneﬁts from the individual’s creativity and the individual beneﬁts from the
organization’s structure and resources.

Although both supplementary and complementary congruence are generally
accepted as valid forms of P-O ﬁt and are both compatible with the person-environment
interaction perspective, most empirical research has focused on supplementary ﬁt
(Kristof-Brown, et al., 2005). Hypotheses concerning the supplementary form of P-O ﬁt
(which focuses on similarity), have largely grown from Schneider’s (1987) Attraction-
Selection-Attn'tion (ASA) framework. The ASA model proposes that certain types of
people are attracted to organizations where they perceive the organization and its people
as being similar to themselves in some way; that organizations select applicants who they
feel are similar to current members; and ﬁnally, that selected members who perceive that
they are dissimilar to the majority of organizational colleagues will attrit, or leave the
organization. The result of this process is a relatively homogeneous group of people
within the organization, as well as the formation of an aggregate organizational

“personality.”

26

Person-organization similarity could occur on several dimensions, including goal
congruence (e. g., Vancouver & Schmitt, 1991) or personality-climate congruence (e.g.,
Christiansen, Villanova, & Mikulay, 1997; Ryan & Schmit, 1996). By far, the most
commonly-used dimension on which P-O ﬁt is assessed is values congruence (Kristof,
1996; Verquer, Beehr, & Wagner, 2003). This emphasis on values can be largely
attributed to Chatman’s (1989) inﬂuential theory of P-O ﬁt, as well as to the validation of
a values-based P-O ﬁt measure, the Organizational Culture Proﬁle (O’Reilly, Chatman,
& Caldwell, 1991; Kristof-Brown, et al., 2005). With some exceptions (e.g., Cable &
Judge, 1996; 1997), many researchers have recently deﬁned P-O ﬁt as synonymous with
value congruence (e.g., Cable & DeRue, 2002; Cable & Parsons, 2001; Chatman, 1989;
Kristof, 1996). Empirical support for this conceptualization is provided by Cable and
Judge (1996), who found that values similarity was signiﬁcantly related to P-O ﬁt, but
demographic similarity was not.

Consistent with the ASA framework (Schneider, 1987), P-O ﬁt has been linked
with several positive organizational outcomes, including organizational commitment,
satisfaction, lower intent to turnover, willingness to recommend the organization to
others, and citizenship behavior (e.g., Judge & Bretz, 1993; Cable & DeRue, 2002; Cable
& Judge, 1996; Chatman, 1989; Goodman & Svyantek, 1999; Meglino, Ravlin, &
Adkins, 1989; O’Reilly & Chatman, 1986; O’Reilly et al., 1991; Verquer, et al., 2003;
Westerman & Cyr, 2004). Attitudes and intent to turnover are the most frequently
studied outcomes of P-O ﬁt (Verquer, et al., 2003), and meta-analyses have generally
found signiﬁcant relationships between P-O ﬁt and attitudinal outcomes. Verquer and

colleagues (2003) found signiﬁcant overall mean correlations between value congruence

27

and satisfaction (r = .31), organizational commitment (r = .30), and intent to turnover (r =
-.43). A more recent meta-analysis conducted by Kristof—Brown and colleagues (2005)
found signiﬁcant latent relationships between P-O ﬁt (deﬁned as values congruence) and
satisfaction (r = .50), organizational commitment (r = .65), intent to turnover (r = -.35),
task performance (r = .13), contextual performance (r = .27), and strain (r = -.27). Thus,
P-O ﬁt has been consistently linked with several positive outcomes for organizational
members.

P-O ﬁt has been shown to have positive outcomes for applicants as well as for
current organizational members. For example, P-O ﬁt has been linked to applicant job
choice decisions (Cable & Judge, 1996) and interviewer job offer recommendations
(Adkins, Russell, & Werbel, 1994; Cable & Judge, 1997; Rynes & Gerhart, 1990). The
Kristof-Brown, et al., (2005) meta-analysis supported these conclusions, ﬁnding
signiﬁcant relationships between P-O ﬁt and organizational attraction (r = .46), job
acceptance (r = .24), intent to hire (r = .61 ), and job offers (r = .32). These ﬁndings are
consistent with the proposals of the ASA model (Schneider, 1987) and with research
indicating that applicants prefer organizations which are perceived as being similar to
themselves (Cable & Judge, 1994; Chatman, 1989; Judge & Bretz, 1992). In addition,
applicants who place more weight on P-O ﬁt during the organizational selection process
have higher P-O ﬁt perceptions after organizational entry, and post-entry P-O ﬁt
perceptions are in turn linked to higher satisfaction and commitment (Cable & Judge,
1996). Overall, research supports the proposition that P-O ﬁt is inﬂuential in the

organizational selection process as well as after organizational entry.

28

Given the strength of the evidence supporting relationships between P-O ﬁt and a
wide array of positive outcomes, we might next ask what the antecedents of P-O ﬁt might
be. Most research has focused on selection and socialization processes as the two main
factors that affect P-O ﬁt. In each case, attention has been focused on both individual
differences and characteristics of the organization.

Much research has assumed that applicants affect their future P-O ﬁt by selecting
organizations where they believe they will ﬁt well and by self-selecting out of selection
processes for organizations where they do not believe they will ﬁt (e. g., Bowen, Ledford,
& Nathan, 1991; Schneider, 1987). Some empirical research has supported this
proposition, generally using a policy-capturing approach in which participants make
judgments about the attractiveness of hypothetical organizations (Bretz, Ash, & Creher,
1989; Burke & Deszca, 1982; Cable & Judge, 1994; Judge & Bretz, 1992; Turban &
Keon, 1993). In addition, Cable and Judge (1996) found that applicants who reported
placing greater emphasis on P-O ﬁt concerns during selection report experiencing higher
levels of P-O ﬁt after organizational entry. Cumulatively, this research supports the
notion that P-O ﬁt can be affected by individuals’ efforts to select organizations in which
they believe they will ﬁt.

P-O ﬁt can be affected by organizations as well as by individuals. Cable and
Judge (1997) found that interviewers were able to make accurate judgments about the
extent to which applicant and organizational values were congruent and that these
judgments signiﬁcantly affected the interviewers’ recommendations. The emerging

consensus, therefore, is that P-O ﬁt is a concern of both applicants and organizations, and

29

that decisions made during the selection process affect subsequent P-O ﬁt. Clearly, P-O
ﬁt is a function of both individual and organizational inﬂuences in the selection process.

P-O ﬁt can also be affected after organizational entry during the socialization
process. Because one of the major goals of the socialization process is to encourage the
continuation of important organizational values (Jones, 1986; Van Maanen & Schein,
1979), and because P-O ﬁt is often conceptualized as value congruence, one could say
that one of the major purposes of the socialization process is to ensure P-O ﬁt. It has
been suggested that when individuals enter an organization, they often experience “reality
shock” characterized by high uncertainty, anxiety, and differences between their
expectations about the organization and organizational reality (Jones, 1986; Louis, 1980;
Van Maanen & Schein, 1979). Research has shown that socialization techniques can
signiﬁcantly improve P-O ﬁt by changing the individual’s values (Cable & Parsons,
2001). In these ways, the organizational environment inﬂuences P-O ﬁt.

Although the majority of socialization research has centered on the extent to
which the organization actively inﬂuences P-O ﬁt, it has also been proposed that P-O ﬁt
can be improved through individual efforts to change the organization. Research has
shown that people can and do actively change their organizational situations (e. g., Kohn
& Schooler, 1978; Meyerson & Scully, 1995; Miner, 1987). Although such bottom-up
effects may be weaker or slower to emerge than the top-down effects that the individual
has on the organization, these individual efforts may result in shifting norms and may
eventually have large effects on the organizational climate. It has been recommended
that researchers consider the possibility of such individual inﬂuences on P-O ﬁt

(Chatman, 1989).

30

In the present study, consistent with the cognitive sensemaking perspective on
person-environment interaction, I am interested in perceptions of P-O ﬁt. The perceptions
perspective allows for P-O ﬁt to exist as long as it is perceived to exist, regardless of the
actual degree of ﬁt (Kristof, 1996). Like perceived met expectations, perceived P-O ﬁt
may or may not mirror the “objective” nature of the situation. In other words, an
individual may perceive value congruence even when the discrepancy between the
individual’s values and separately-assessed measures of the organization’s values is
large.

I also consider the ways in which changes in perceived P-O ﬁt might occur.
Changes in perceptions of P-O ﬁt may or may not reﬂect actual value change on the part
of either the individual or the organization. Some preliminary research has focused on the
extent to which the organization can affect individual perceptions of P-O ﬁt. For
example, research has shown that socialization techniques can signiﬁcantly improve P-O
ﬁt by changing the individual’s values (Cable & Parsons, 2001). Actual changes in the
individual’s values should often be related to changes in perceptions of person-
organization value congruence. However, it is also possible that the socialization process
might result in increased perceptions of the extent to which the individual’s values match
those of the organization without creating any actual change in individual values.

It has also been suggested that perceptions of P-O ﬁt can be improved through
individual efforts to change the organization. Research has shown that people can and do
actively change their organizational situations (e. g., Kohn & Schooler, 1978; Meyerson
& Scully, 1995; Miner, 1987), and it seems reasonable to believe that these actual

changes may be correlated with changes in perceptions of the situation. In summary,

31

research suggests that one’s perceptions of P-O ﬁt can change over time and can be
affected by both the organization’s inﬂuences on the individual and the individual’s
inﬂuence on the organization.

Perceived P-O Fit Measurement Issues. Several methods for measurement and
computation of P-O ﬁt have been proposed. Direct, or subjective, P-O ﬁt measures
directly ask respondents to report how well they think their characteristics match the
organization’s characteristics. Indirect measures ask respondents to rate both themselves
and the organization on dimensions such as values. For indirect measures, the ratings of
the organization’s characteristics can be provided either by the individual or a separate
rater, or an aggregate of numerous responses may be used. Using indirect measures, the
degree of ﬁt is then computed using, for example, a difference score, interaction,
correlation coefﬁcient, or polynomial regression.

P-O ﬁt has been the subject of measurement discussion that is largely parallel to
the issues discussed relevant to met expectations (Edwards, 1991; Irving & Meyer, 1994;
1995; Kristof-Brown, et al., 2005; Meglino & Ravlin, 1998; Verquer, et al., 2003). The
concerns raised regarding direct measurement of P-O ﬁt (e.g., Edwards, 1991) largely
parallel those raised by Irving and Meyer (1994, 1995) regarding direct measurement of
met expectations. Edwards (1991) argued that direct measures of P-O ﬁt are problematic
in that they do not allow researchers to distinguish between the effects of personal values
and organizational values and that unless the speciﬁc values or other dimensions to be
considered are speciﬁed, researchers cannot ensure that commensurate dimensions are
being used when respondents estimate ﬁt. In addition, direct measures have been

criticized to the extent that signiﬁcant relationships between scores on these measures

32

and responses on job attitude measures may be created by respondents’ desire to be
cognitively consistent (Salanick & Pfeffer, 1977).

Other P-O ﬁt researchers have responded to these claims by arguing convincingly
that direct and indirect measures may, in fact, be tapping different constructs (Kristof-
Brown, et al., 2005). Research suggests that direct and indirect measures of P-O ﬁt are
only weakly related (Cable & Judge, 1997; Kristof-Brown & Stevens, 2001). It has been
proposed that the weak correspondence between these two types of measures might be
due either to individuals’ failure to interpret characteristics of the environment accurately
or failure to accurately self-assess (French, Rogers, & Cobb, 1974). These two errors
would cause one’s perception of the ﬁt between oneself and the environment to differ
from the objective discrepancy between personal and organizational values.

Several studies have supported the notion that discrepancy scores and direct
measures of P-O ﬁt may not be measuring the same construct. First, as mentioned above,
these measures tend to be only weakly correlated (Cable & Judge, 1997; Kristof-Brown,
et al., 2005). Second, research has found that these two types of measures are
differentially related to outcomes. For example, Cable and Judge (1997) found that
perceived values congruence was signiﬁcantly related to interviewers’ hiring
recommendations and organizational hiring decisions, whereas actual values congruence
was not. A study undertaken to explicitly examine the relationship of discrepancy and
direct measures of P-O ﬁt found that discrepancy scores do not relate to perceived P-O ﬁt
in the expected manner, and it was suggested that ﬁt discrepancy and perceived ﬁt be
considered distinct constructs (Edwards, Cable, Williamson, Lambert, & Shipp, 2006).

The consensus, then, is that perceived P-O ﬁt is not the same thing as objective P-O ﬁt,

33

and when researchers are interested in assessing perceptions of P-O ﬁt, direct measures
are appropriate (Cable, et al., 1995; French et al., 1974; Kristof, 1996; Kristof-Brown, et
aL,2005)

In the present study, I am interested in perceived P-O ﬁt. Because the cognitive
sensemaking approach to person-environment interaction speciﬁes that individuals
should be expected to perceive situations in unique ways, it allows for P-O ﬁt to exist as
long as it is perceived to exist, regardless of the actual degree of ﬁt (Kristof, 1996;
Terborg, 1981). In addition, when direct measures of perceived P-O ﬁt are used,
respondents are able to weight various factors as appropriate for their perceptions,
allowing for individual differences in the salience of various dimensions (Kristof-Brown,
et al., 2005). In support of the idea that perceptions are stronger drivers of attitudes and
behaviors than are objective realities, research has shown that direct measures of P-O ﬁt
perceptions tend to be more strongly predictive of several outcomes than are objective ﬁt
measures (Kristof-Brown, et al., 2005). Theoretically, direct measurement is more
appropriate than discrepancy-based measurement when researchers are interested in
perceived P-O ﬁt.

Summary. Several of the theoretical and measurement concerns raised in the
literature regarding met expectations have also been raised regarding P-O ﬁt. These
concerns are based on issues regarding the construct deﬁnition of these variables — more
speciﬁcally, the distinction between “actual” met expectations or “actual” P-O ﬁt and
perceptions of these constructs. In the case of both perceived met expectations and

perceived P-O ﬁt, it is clear that the perceptions of a contextualized person, who both

34

inﬂuences and is inﬂuenced by his or her environment, are more adequately assessed via
direct measurement than by discrepancy scores.
Generalizability to Student Samples

In the previous sections, I have reviewed the literature on goal commitment,
perceptions of met expectations, and perceptions of P-O ﬁt. This review focused
primarily on research from the organizational literature. In the present study, I use a
sample of undergraduate college students. Because the generalizability of ﬁndings from
a student sample to organizational employees (and vice versa) may be a concern for some
readers, I now present several arguments in favor of the appropriateness of the student
sample used in this study.

First, it is important to note that the hypotheses tested here are high-level in nature
and are proposed to reﬂect broad psychological processes that function in equivalent
ways in multiple situations. Because the perceptions addressed here refer to one’s
perception of one’s relationship with a particular environment, the nature of the speciﬁc
environment (e. g., organizational or academic) is taken into account. The individual
“objective” features of the environment that affect one’s construction of perceptions of
met expectations or P-O ﬁt may vary between academic and organizational environments
(just as they are expected to vary across persons). However, I propose that once those
perceptions have been constructed, the high-level psychological process by which those
perceptions are related to one another and to goal commitment is likely to apply
regardless of the speciﬁc nature of the individual’s environment.

Furthermore, several similarities between student and employee populations have

been identiﬁed that will be relevant to the hypotheses presented here. These four

35

similarities support the use of the student sample in the present study. First, both students
and employees accept offers to enter their organizations based on expectations about
what organizational life will be like. This has been found to be true for both high school
seniors accepting college admission offers (Braxton, Vesper, & Hossler, 1995; Chapman,
1981) and for job applicants (Cable & Judge, 1994; Cable & Judge, 1996; Judge & Cable,
1997). Second, on average, both incoming college students and new organizational
employees tend to have unrealistically inﬂated expectations that are often unmet upon
entry (ACT, 2004; Buckley, 1971; King & Walsh, 1972; Lauterbach & Vielhaber, 1966)
and lead to high initial turnover rates (Gerdes & Mallinckrodt, 1994; Helland, et al.,
2001-2002; Mallinckrodt & Sedlacek, 1987; Shaw, 1968).

Third, college students and employees hold personal goals and vary in their
commitment to those goals. Much research on the types of goals individuals hold has
focused on achievement goals, and this research suggests that both college students and
organizational employees hold varying levels of achievement goals (e.g., Button,
Mathieu, & Zajac, 1996). Research on more speciﬁc types of goals, including
individually-selected personal work goals for organizational employees and the goal of
college degree obtainment for college students, supports the inference that people have
goals to which they are committed in varying degrees, and that those goals are related to
important outcomes such as attitudes, performance, well-being, and retention (e.g., Allen
& Nora, 1995; Kaufman & Creamer, 1991; Maier & Brunstein, 2001; Moore &
Davidson, 2006; Roberson, 1990).

Finally, research suggests that both students and employees care about P-O ﬁt.

Fit or perceptions of ﬁt have been related to several important student outcomes,

36

including GPA, attitudes, and turnover intentions (Cabrera, Nora, & Castaneda, 1993;
Pascarella, Smart, & Ethington, 1986; Schmitt, Oswald, F riede, Imus, & Merritt, 2007);
as well as for similar employee outcomes including performance, attitudes, and turnover
intentions (Kristof-Brown, et al., 2005; Sims & Kroeck, 1994).

The arguments presented above suggest that college students and employees may
experience similar psychological processes involving perceptions of met expectations,
perceptions of ﬁt, and goal commitment following organizational entry. In the present
study, I hypothesize that perceived met expectations and perceived P-O ﬁt will be
signiﬁcant predictors of goal commitment. Furthermore, because perceptions are
proposed to change over time as the person and situation exert reciprocal inﬂuences on
one another (Terborg, 1981), the relationship between changes in these factors over time
and changes in goal commitment over time will be examined.

General Study Rationale

Research has not yet directly incorporated self-environment perceptions in the
prediction of goal commitment. However, there is some evidence that suggests that
consideration of the interaction between person and situation may contribute to the
prediction of goal commitment. For example, Klein and Wright (1994) found that the
multiplicative interaction of self-esteem and normative information about peers’
performance and the interaction of self-esteem and rewards captured 14% of the variance
in goal attractiveness over and above the direct effects of these variables. Because goal
attractiveness has been shown to be a proximal antecedent of goal commitment, a
possible effect of the interaction on goal commitment itself can be inferred. In addition,

Hollenbeck et al. (1989) found that the interaction of need for achievement and goal

37

origin (self-set or assigned) signiﬁcantly added to the prediction of goal commitment.
These studies suggest that person and situation factors may combine to affect goal
commitment and that further investigations of such relationships may be fruitful.

Additional evidence that interactive constructs may be signiﬁcant predictors of
goal-related constructs was provided by Kristof-Brown and Stevens (2001). These
authors studied goal congruence (mastery and performance goals) in project teams. It
was hypothesized that when an individual team member perceives that other team
members share his or her goals, he or she might believe that the team is more likely to
support the attainment of valued outcomes. In other words, they hypothesized that
perceived goal congruence would affect team member expectancy for goal attainment.
They found that perceived goal congruence among team members was signiﬁcantly
related to increased team member satisfaction and increased work-related contributions to
the team project. Kristof-Brown and Stevens (2001) did not explicitly measure
expectancy, nor did they assess the extent to which team member goals affected and were
affected by other team members throughout the 7-week task. However, their study points
to the potential contribution of perceptions of met expectations and perceptions of P-O ﬁt
in predicting goal-related attitudes and behaviors.

In summary, the present study contributes to the literature by investigating the
extent to which perceptual variables based theoretically in the dynamic forms of person-
environment interaction are signiﬁcant predictors of goal commitment. Although
previous studies have considered the possibility that the mechanical forms of person-

environment interaction might relate to goal commitment, research has yet to consider

38

variables based in the reciprocal inﬂuences and cognitive sensemaking forms of
interaction.
Hypotheses

The ﬁrst two hypotheses concern the relationship between perceived met
expectations and perceived P-O ﬁt. According to the ASA model (Schneider, 1987),
individuals are attracted to organizations that they perceive as being similar to themselves
on the things they value. This proposition has been supported empirically by studies
ﬁnding that interviewers make offers, and applicants make acceptance decisions, based in
part on P-O ﬁt concerns (e.g., Arthur, Bell, Villado, & Doverspike, 2006; Cable & Judge,
1996, 1997). These ﬁndings could be interpreted as suggesting that interviewers and
applicants make decisions based on expectations about P-O ﬁt. In other words, an
applicant selects an organization in which he or she believes that he or she will achieve
the greatest degree of ﬁt on the values that are most important to that individual
applicant, whatever those values might be. Once the individual has entered the
organization, the extent to which he or she believes that these pre-entry expectations are
being met is likely to be related to the degree to which P-O ﬁt is perceived.

For example, an individual who values autonomy is likely to accept membership
in an organization which he or she believes will provide him or her with some sufﬁcient
degree of autonomy. Following organizational entry, the extent to which his or her
expectations regarding autonomy are perceived as being met is likely to be related to the
extent to which congruence on the value of autonomy is perceived. For example, if a

student is enrolled in classes that assign points for daily attendance (thereby limiting his

39

or her autonomy to make decisions about class attendance), s/he is likely to perceive that
while he or she values autonomy, the organization does not.

Hypothesis 1 .' Perceived met expectations will be positively related to

perceived P-Ofit.

Because perceptions of met expectations and P-O ﬁt are likely to be constructed
as a function of both the person and the environment, any changes in perceptions of the
self or the environment are likely to result in changes in the constructs over time.
Changes in perceptions of met expectations may occur as a result of changes in beliefs
about what one’s pre-entry expectations were or changes in one’s perceptions of the
environment. It is likely that the former would be caused by memory distortion or
cognitive dissonance, whereas the latter is likely to be related to actual changes in work
experiences.

Regardless of whether or not changes in perceptions are related to actual changes
in the work situation, the change trajectories of perceived met expectations and perceived
P-O ﬁt are expected to work in tandem. For example, memory bias may affect one’s
beliefs about pre-entry expectations such that the expectations are remembered as higher
than they actually were. As long as the organizational situation is perceived as remaining
constant, perceptions of met expectations should then decrease. Because the individual
believes that he/ she entered the organization expecting to ﬁt into a more positive
environment than the one actually experienced, reports of perceived P-O ﬁt should also
decrease.

Relatedly, one’s memory of his or her pro-entry expectations may remain

constant, but one’s perceptions of the environment might change. Following the example

40

regarding autonomy presented previously, if one values autonomy yet perceives a
decrease in the amount of autonomy that one has, one might experience decreased
perceptions of the extent to which the organization values autonomy (i.e., decreased
perceptions of value congruence). Therefore, we would expect to see both a decrease in
perceptions of met expectations and a decrease in perceptions of P-O ﬁt.

Hypothesis 2: Changes in perceptions of met expectations will be

positively related to changes in perceived P-Ofit.

Hypotheses 3 and 4 concern the relationship between perceived P-O ﬁt and goal
commitment. The Hollenbeck and Klein (1987) model proposes that some attitudinal
constructs, including job involvement and organizational commitment, are antecedents of
goal commitment. Some research has begun to provide empirical support for these
proposed relationships (e.g., Maier & Brunstein, 2006); however, no research has, to my
knowledge, directly linked perceived P-O ﬁt with goal commitment. One might question
why this link would be expected. Whereas a relationship between perceived P-O ﬁt and
institutional commitment would be somewhat commonsense, a relationship between
perceptions of P-O ﬁt and goal commitment may seem counterintuitive, particularly in
the academic domain. Why would poor ﬁt at one school cause a student to be less likely
to complete a degree? Why would it not result in a transfer to a different school?

Perceived P-O ﬁt is hypothesized to be signiﬁcantly associated with goal
commitment for three reasons. First, recall that perceived P—O ﬁt represents the degree of
perceived similarity between individual values and organizational values. To the extent
that there is value congruence, the goals the individual sets for him or herself are likely to

be in line with organizational values. Self-set goals that are aligned with organizational

41

values are more likely to further organizational objectives and be rewarded by the
organization. The increased potential for rewards related to goal attainment should
increase goal attractiveness, which has in turn been linked to goal commitment (Klein, et
aL,1999)

Second, and perhaps more importantly, when there is perceived value congruence,
the goals that the organization assigns to its members are less likely to violate individual
values. For example, an organization that values competition would be more likely to
assign an individual the goal of outperforming other employees or classmates. An
individual being assigned this goal is likely to have more commitment to this goal to the
extent that he or she also values competition. Therefore, perceived P-O ﬁt is likely to be
signiﬁcantly related to goal commitment, particularly for assigned goals.

The third reason to hypothesize a relationship between perceptions of P-O ﬁt and
goal commitment stems from recent work in the social cognition literature and is
displayed in Figure 3. The rationale presented here implies a causal relationship and is
presented with an acknowledgment that the causal mechanism discussed is not tested in
the present study. It is merely discussed as rationale for the hypothesis relating
perceptions of P-O ﬁt and goal commitment. The foundation of the argument to be
presented is the basic notion that perceived P-O ﬁt represents perceptions of similarity
(i.e., congruence) between the self and the organization. The combination of two
ﬁndings from research on mental networks supports the hypothesis that perceived P-O ﬁt
will affect goal commitment, as discussed below.

The ﬁrst relevant ﬁnding is that research has shown that most people tend to hold

positive evaluations of themselves (Bosson, Swarm, & Pennebaker, 2000; Greenwald &

42

Farnham, 2000; Koole, Dijksterhuis, & Van Knippenberg, 2001). Therefore, attitude
objects that are more closely associated with the self should be evaluated more positively
than attitude objects that are viewed as more distal from the self. Because perceptions of
P-O ﬁt are conceptualized in terms of congruence between the self and the organization,
greater perceptions of P-O ﬁt should be represented in the form of stronger mental links
between the self and the organization. In contrast, any perceived mis-ﬁt between the self
and the organization should lead to weaker mental links between the self and the
organization. Because activation spreads more reliably through stronger mental links than
through weaker ones, the positive association with the self should spread through the
strong link to the organization when perceptions of P-O ﬁt are strong. Therefore,
perceived P-O ﬁt is proposed to affect the extent to which the positive evaluation
associated with the self transfers, via spreading activation, to a positive evaluation of the
organization.

The second relevant ﬁnding is that when the organization concept becomes
mentally activated, any concepts strongly related to the organization will also become
activated. This is likely to include any goals related to the organization. These goals
may be either self-set (e. g., career goals) or assigned. However, it is likely that
organization-assigned goals may be particularly strongly associated with the organization
in memory. Some portion of any positive evaluation that has become associated with the
organization through high perceptions of P-O ﬁt is now likely to spread to the goal itself.
This positive evaluation should increase the degree to which the goal is perceived as
attractive. Because goal attractiveness has been established as a proximal determinant of

goal commitment (Hollenbeck & Klein, 1987; Klein, et al., 1999), increased goal

43

attractiveness should lead to increased goal commitment. Therefore, increased P-O ﬁt
perceptions should be related to increased goal commitment via its effects on goal
attractiveness.

Figure 3
Spreading Activation of Positive Evaluation to Organizational Goals

 

 

P-O Fit

 

 

 

   
 

 

 

 

Org. Goals

 

 

Positive
Evaluation

 

The three arguments presented above provide the logic for the hypothesis that...

Hypothesis 3: Perceived P-Oﬁt will be positively related to goal

commitment.

The theoretical discussion presented regarding Hypothesis 3 also implies that
changes in perceptions of P-O ﬁt will relate to changes in goal commitment over time.
The concept of changes in goal commitment is not problematic, as goal commitment is
not deﬁned as a constant, unchanging construct (Hollenbeck, et al., 1989). Instead, goal
commitment is theoretically seen as a state-like construct that can change as efforts to
achieve the goal are implemented and evaluated. According to the Hollenbeck and Klein
(1987) model, goal commitment may change either as a function of changes in the

attractiveness or the expectancy of goal attainment.

44

The rationale behind the relationship between changes in perceptions of P-O ﬁt
and changes in goal commitment parallels that for Hypothesis 3. First, regardless of the
reason for changes in perceived P-O ﬁt, decreases in these perceptions will be associated
with an increased likelihood that the individual’s goals will not reﬂect organizational
values or that the goals assigned to the individual by the organization will be
incompatible with the individual’s values. In both cases, these mismatches are expected
to relate to decreased attractiveness of goal attainment. In this way, changes in
perceptions of P-O ﬁt should relate to commitment to newly set and assigned goals over
time.

Furthermore, changes in perceptions of P-O ﬁt are expected to be associated with
changes in commitment to the same goal over time. It has recently been proposed that
strengthening the link between the self and an attitude object will cause evaluations of
that object to become more positive (Walther, Nagengast, & Traselli, 2005). As
perceived P-O ﬁt increases, the strength of the mental link between the self and the
organization increases, leading to more positive evaluations of the organization. As
discussed above, the stronger that link becomes, the more positive the evaluation of the
organization and its associated goals is expected to become. In this manner, as perceived
P-O ﬁt increases, the attractiveness of organizational goals is expected to increase. This,
in turn, leads to increased goal commitment.

Hypothesis 4: Changes in perceived P-Oﬁt will be positively related to

changes in goal commitment.

The hypotheses, as stated above, imply that perceptions of P-O ﬁt will

mediate the relationship of perceived met expectations and goal commitment. In

45

other words, it is expected that some portion of the effects of perceived met
expectations on perceived P-O ﬁt will be transferred onto goal commitment, such
that perceived met expectations will have a signiﬁcant indirect effect on goal
commitment. Likewise, I expect that some portion of the effects of changes in
perceptions of met expectations will be transferred onto changes in goal
commitment via their effects on changes in perceived P-O ﬁt. Therefore, I
present Hypotheses 5 and 6:

Hypothesis 5 .' Levels of perceived P-O ﬁt will mediate the relationship

between levels of perceived met expectations and levels of goal

commitment.

Hypothesis 6: Changes in perceived P-Ofit will mediate the relationship

between changes in met expectations and changes in goal commitment.

In summary, it is hypothesized that two speciﬁc perceptions, perceived met
expectations and perceived P-O ﬁt, are predictors of goal commitment, as shown in
Figure 4. These hypotheses contribute to the previous models of goal commitment in an
important way. Although past goal commitment theory and research have considered
mechanical forms of person-environment interaction, including independent directs and
statistical interactions, research has largely failed to incorporate the dynamic forms of
interaction. Personal factors such as ability, task experience, and affect have been shown
to affect goal commitment, as have situational factors including goal speciﬁcity,
participation, and feedback (Klein, et al., 1999). However, the literature is lacking an
examination of the cognitive processing elements of the Locke, et a1. (1988) model,

including perceptual constructs. Because people’ perceptions of situations are the driving

46

force behind their reactions to them (e.g., Fiske & Taylor, 1991; Nisbett & Ross, 1980),
consideration of the perceptions arising from the combination of a unique person with a
particular situation has on the individual’s perceptions is essential.

The present study represents a preliminary step in addressing this gap in
the literature. The predictor variables incorporated here represent the perceptions
of the contextualized person, and as such, they are dynamic constructs that
represent something more than a simple, mechanical combination of person and
situation factors (Edwards, et al., 2006; Terborg, 1981).

Some past research has suggested that perceptions of the self-environment
relationship may be more strongly predictive of outcomes than are such simple
mathematical combinations of person and situation factors (Kristof-Brown, et al.,
2005). However, in the present study, I did not have the data necessary to
compare the relative effect sizes of person, situation, and perceptual variables.
Therefore, the present study simply seeks to provide evidence regarding the extent
to which perceived met expectations and perceived P-O ﬁt in and of themselves
can be used to predict goal commitment. Although the results of the present study
cannot speak to the question of “value added,” they can provide evidence
regarding the extent to which perceptions representing the reciprocal and
cognitive sensemaking approaches might be useful individual predictors of goal
commitment.

In addition to this theoretical contribution to the goal commitment
literature, the present study provides a demonstration of a relatively new

methodological technique — mediated latent growth modeling using three latent

47

intercept and three latent slope terms. This technique allows the researcher to
statistically test the extent to which changes in a variable have indirect effects on
the change trajectory of another variable. The ability to test this indirect effect is
central to the evaluation of Hypotheses 5 and 6. The mechanics of this technique
will be described in the following section, following a description of the study

sample, procedure, and measures used.

48

   

   

   

   

E08 E05

    

 

      
 
 

     

      

   

   

 
 

“5:55:80 em HE OE NE .9Q 62
Row uniooaom woZooaom
E0835 E0285 E0885
EoEHEEoU mm HE Oi ~E .Exm BE

 
 

  
  

Uoioouom quoocom

  
  

ammmﬁomamcxo ENE: 333.3on
v 0.2%?“

49

Method

The hypotheses presented above were tested using an existing dataset obtained
with the support of The College Board. The data collection procedure, sample
characteristics, and descriptive statistics are described below. Following that discussion,
I describe the results of the hypothesis tests.
Data Collection Procedure

The dataset used in the present study was collected in three waves following a
larger, initial data collection effort. Participants were incoming students at ten US.
colleges and universities, including Michigan State University, Winston-Salem State
University, California State University-Fullerton, Indiana University, Ohio State
University, Spelman College, University of Iowa, University of Chicago, University of
Michigan, and Virginia Polytechnic Institute and State University. The procedures for
the initial data collection and the subsequent three waves of data collection are described
below.
Initial Data Collection and Sample

The initial data collection effort was coordinated by researchers from Michigan
State University and was conducted on-site by representatives from each of the
participating colleges and universities. A variety of techniques were used to recruit
participants for this data collection effort according to the preferences of each school.
Appendix A contains study information by school including recruitment methods,
location of sessions, number of sessions, and the number of proctors used during the data
collection sessions. Participants were paid $40 to complete questionnaires for a two-hour

time period, with the exception of students from the University of Chicago, who were

50

paid $50 at the recommendation of organizational representatives. This data collection
took place in late summer/early fall 2004, as students participated in orientations and
began classes in their ﬁrst semester. The total sample size across all ten schools was
2771. The average age of participants was just over 18 years; over 97% of the sample
was either 18 or 19 years of age. Sixty-four percent of the sample was female, 96% were
US. citizens, and 94% indicated that English was their native language. Ethnically, this
sample was 55% Caucasian, 25% African-American (due to intentional oversampling of
this subgroup), 6 % Hispanic, 7% Asian, and 7% other ethnicities.
Wave 1, 2, and 3 Data Collections

As described above, 2771 incoming college freshmen participated in the initial
data collection in summer/fall 2004. Of these participants, 2631 (95%) granted the
researchers permission to re-contact them in order to solicit their participation in future
paid research studies. These 2631 participants were contacted to participate in the Wave
1, 2, and 3 data collections, which formed the dataset to be used in this study. All
recruiting and data collection for these three waves were conducted via e-mail. For each
administration, three to four e-mail communications were sent to the e—mail address
provided by participants on their “permission to recontact” forms. The ﬁrst e-mail was a
pre-notice advising participants that they would be receiving a formal invitation to
participate in research in the near future. The second e-mail was the research invitation
giving details about the survey and compensation, and the third and fourth e-mails were
reminders about the survey.

The Wave 1 data collection took place in winter 2004, at the end of the

participants’ ﬁrst semester. This timing allowed respondents to have sufﬁcient experience

51

with their organizations to effectively rate themselves on the variables of interest at the
ﬁrst Wave. The Wave 2 and 3 data collections took place near the end of the respondents’
second and third semesters, respectively. Administering Waves 2 and 3 in the students
second and third semesters provided them chances to respond to the different experiences
inherent in having different coursework and professors across semesters. Therefore, I
expect that one data collection per semester should be sufﬁcient time for changes in the
constructs of interest to emerge.

In order to facilitate data collection at the 10 participating institutions, an Internet-
based survey was created. Participants could complete this survey from any computer
with access to the Internet. All aspects of the experimental process, including consent,
measures, and debrieﬁng, were completed via this web survey. As compensation for
completing each wave, participants received an electronic gift certiﬁcate for $20,
redeemable at Amazon.com®. In addition, for each wave, all participants were entered
into a drawing for a $100 cash prize. Of the 2631 initial participants who granted us
permission to re-contact them, 1234 (47%) students provided useable data in Wave 1,
1044 (40%) provided useable data in Wave 2, and 904 (34%) provided useable data for
Wave 3. Complete data for purposes of analyses to be conducted in this study were
obtained from 566 students.

The average age of this sample of 566 respondents was just over 18 years; over
97% of our sample was either 18 or 19 years of age. Sixty-ﬁve percent of the sample was
female, 93% were US. citizens, and 93% indicated that English was their native
language. Ethnically, this sample was 68% Caucasian, 7% African-American, 4%

Hispanic, 13% Asian, and 9% other ethnicities.

52

Table 2 breaks down the sample descriptive statistics, including racioethnicity,
gender, average z-transformed SAT score, and average scores on Wave 1 target variables,
by school. To provide information regarding the sample representativeness within
school, the overall school’s gender and racioethnic percentages and average 2-
transformed SAT score for the Fall 2004 incoming class are also displayed. SAT scores
were z-transformed using national norms for the 2004 year. Note that for the University
of Chicago, racioethnicity statistics for the Fall 2004 incoming class were not available;
therefore, 2006 statistics were used (the admissions ofﬁce indicated that no major
changes in racioethnic proﬁle had occurred between 2004 and 2006). Ns for some
institutions (e.g., Spelman College and Winston-Salem State University) are low,
reﬂecting difﬁculties retaining these participants throughout the waves of data collection.
Some of these difﬁculties may have related to students providing us with generic e-mail
addresses at the beginning of the study (i.e., yahoo or hotrnail addresses) but later perhaps
switching to school e-mail addresses (i.e., .edu addresses), making sustained contact
difﬁcult. Participants from the schools with higher Ns tended to provide us with .edu e-
mail addresses initially.

As the table indicates, participants in the sample generally had higher SAT scores
than did non-participants, and females were over-represented in the sample.
Racioethnically, the samples from some schools overrepresented Whites, while the
sampling at other schools overrepresented Blacks. Hispanic participants were generally
underrepresented, and Asian participants were generally representatively sampled with

the exception of the University of Chicago, in which they were oversampled.

53

.C E 28$. cease. 8am
..o3£.§8 E: VQQN nosxaotmhﬁa bmowsim + .VQQN :QS 338. 3 ESPN rush M23563 3:3 mmmeatmoxom oEQEmoEmh

 

 

E238 Eek? C S 23:52 .wmwsomamx 138.. has 333 3:345 8 ~33 g N 8 33» 3: AS: amMeEmoLmo‘ $32
84. N2 a2 a3 an. :s an E a as o as a as ex 2 have <>
m3. Ea Ba am. 3 of as R a: 2 as _ as a. as ea 8 5min:
8.4 3a 8a a5 a? as 2. as : as a. E N as R a: saw 25
and. Sa :a a3 2. as mm as o as a As 2 as 8 mm 232:
one Ea mam a5 a. as 2 as _ S _ as _ as aw ea «32
8.4 m3. mom 63 8. as o as o as o as 2: as o E 55.0%
52mm
8a 2:. mom a: 2.- as o E o S o :6 mm a: mm m age;
64 aa aa aw : a: as R a s a. as a. as a as 2 M: .692ch .3
as am :a a3 am. as am as a. S o as a as a 2: am swag:
34 can 8a a: a. as cm as am as a. as o as S. cm 8:25
.58 E ed .mxm a: 283 2a: 53. seamen seam £55 a 2 693
So egseom Basses 539% a a a a

 

Beach «3 335.3; 8335 N Sex: has anaconda FUSE Em. .bBESm swam mawﬁom.

N 2an

54

One-sample t-tests were conducted to test for signiﬁcant mean differences
between respondents who were included in the ﬁnal sample (i.e., respondents who did not
have missing data on the items of interest at any wave) and those who participated at
Wave 1 but were not included in the ﬁnal sample due to missing data. The sample sizes,
means, and standard deviations for non-respondents and respondents can be found below
in Table 2. The Cohen’s d statistics for each mean difference are also displayed. The
results of the t-tests indicated that participants included in the ﬁnal sample scored
signiﬁcantly higher on met expectations (p < .05) than those who were not. The Cohen’s
d for this mean difference is d = -.13, which is considered a small effect size. No
signiﬁcant differences were found between respondents and nonrespondents on

perceptions of P-O ﬁt or goal commitment.

 

 

 

Table 3
Mean comparisons for ﬁnal sample respondents and non-respondents
Non-respondents Final Sample
Scale N Mean SD N Mean SD d
Met Expectations 621 3.09 .62 566 3.17 .60 -.13*
P-O Fit 632 3.77 .80 566 3.84 .78 —.09
Goal Commitment 652 4.55 .47 566 4.58 .43 -.07

 

* The mean difference was significant at p < .05

Measures

Perceived Met Expectations. The extent to which participants perceived that their
expectations were met regarding academic aspects of their colleges or universities was
measured using a four-item scale developed by the researchers for the purpose of this
study. The focus on academic/class-related aspects of the respondents’ experience may
be analogous to examining perceptions of met expectations regarding task-related aspects
of the organizational experience. Although perceptions of met expectations regarding

social aspects of one’s experience may also show signiﬁcant relationships with outcomes,

55

 

we focus on the academic/task aspects because organizations have a greater ability to
leverage these factors relative to social aspects in order to create effective changes. All
four scale items are displayed in Appendix B. Following the suggestion of Wanous, et al.
(1992) and the precedent set in previous research (e.g., Major, et al., 1995), these
perceptions were measured on a ﬁve-point Likert-type rating scale with anchors
including 1 (expectations unmet), 3 (expectations met), and 5 (expectations exceeded).

Perceived Organizational Fit. The extent to which participants perceived ﬁt with
their schools was measured using a four-item scale. Three of the items were adapted
from an existing scale (Lauver & Kristof-Brown, 2001), and the fourth item was
generated by the researchers for the purpose of this study. All items were answered on a
5-point, Likert-type scale ranging from strongly disagree to strongly agree. The four
items are listed in Appendix C.

Goal Commitment. Goal commitment was assessed using an adaptation of the 9-
item HWK scale (Hollenbeck, et al., 1989). These items were adapted to speciﬁcally
reference the goal of obtaining a college degree. In the Wave 1 data collection, all 9
items were administered and were answered on a 5-point Likert-type scale ranging from
“strongly disagree” to “strongly agree.” After Wave 1, the three items with the lowest
item-total correlations were dropped from the scale as part of an effort to reduce survey
length. All nine items are listed in Appendix D, with retained items indicated. The
resulting six-item scale was administered in Waves 2 and 3. Four of the ﬁve items
suggested by Klein, et al. (2001) were retained, along with two additional items with
higher item-total correlations. Concerns may be raised regarding the unidimensionality

of the retained six-item scale; however, an exploratory factor analysis (EFA) conducted

56

on the six retained items indicated that only one signiﬁcant factor was present
(eigenvalue = 2.86), which accounted for 48% of the variance in item responses. Note
that for the analyses in the present study, only the six retained items were used for the
Wave 1 goal commitment measure as well as for the Wave 2 and 3 goal commitment

measure.

57

Results

Descriptive Statistics

The item-level means, standard deviations, and item-total correlations for each
Wave of data collection are presented in Table 3. In addition, the scale means, standard
deviations, reliabilities, and intercorrelations for each Wave are presented in Table 4. As
displayed, all scales reached the traditional criterion for acceptable levels of or=.70 with
the exception of the Wave 1 perceived met expectations scale. The lower internal
consistency levels for this scale in general may reﬂect the nature of the scale items —
more speciﬁcally, the fact that each item addresses a different aspect of a student’s
academic experience. The low alpha at Wave 1 was largely due to a single item
reﬂecting expectations about course availability. It is possible that this item failed to
correlate highly with the others at Wave 1 due to students’ lack of experience with course
scheduling. As shown in Table 3, the item showed higher item-total correlations at
Waves 2 and 3. Because dropping this item would have caused the alpha reliability
estimates to fall below or=.70 at Waves 2 and 3, the item was retained at all three time

points.

58

 

Table 4

Item-Level Descriptive Statistics

 

 

Standard Item-
Wave Item Mean Deviation Total r
l Perceived Met Expect. 1 3.08 .85 .44
1 Perceived Met Expect. 2 3.39 .99 .42
1 Perceived Met Expect. 3 2.67 1.17 .27
1 Perceived Met Expect. 4 3.30 .95 .53
2 Perceived Met Expect. 1 3.15 .81 .48
2 Perceived Met Expect. 2 3.42 .90 .48
2 Perceived Met Expect. 3 2.92 1.13 .43
2 Perceived Met Expect. 4 3.35 .95 .60
3 Perceived Met Expect. l 3.15 .77 .52
3 Perceived Met Expect. 2 3.39 .92 .50
3 Perceived Met Expect. 3 3.01 1.04 .41
3 Perceived Met Expect. 4 3.34 .92 .57
1 Perceived P-O Fit 1 3.85 .89 .84
1 Perceived P-O Fit 2 3.90 .85 .84
l Perceived P-O Fit 3 3.69 .93 .65
1 Perceived P-O Fit 4 3.93 .89 .76
2 Perceived P-O Fit 1 3.78 .86 .84
2 Perceived P-O Fit 2 3.80 .84 .89
2 Perceived P-O Fit 3 3.62 .92 .72
2 Perceived P-O Fit 4 3.84 .89 .79
3 Perceived P-O Fit 1 3.79 .89 .85
3 Perceived P-O Fit 2 3.81 .88 .88
3 Perceived P-O Fit 3 3.66 .98 .73
3 Perceived P-O Fit 4 3.84 .93 .80
1 Goal Commitment 1 4.36 .79 .53
1 Goal Commitment 2 4.72 .57 .51
1 Goal Commitment 3 4.30 .92 .52
1 Goal Commitment 4 4.86 .45 .54
1 Goal Commitment 5 4.71 .66 .52
1 Goal Commitment 6 4.49 .65 .50
2 Goal Commitment 1 4.34 .78 .49
2 Goal Commitment 2 4.60 .65 .56
2 Goal Commitment 3 4.20 .91 .54
2 Goal Commitment 4 4.78 .50 .60
2 Goal Commitment 5 4.64 .57 .64
2 Goal Commitment 6 4.39 .67 .49
3 Goal Commitment 1 4.30 .83 .63
3 Goal Commitment 2 4.58 .66 .60
3 Goal Commitment 3 4.15 1.00 .58
3 Goal Commitment 4 4.70 .58 .64
3 Goal Commitment 5 4.55 .72 .64
3 Goal Commitment 6 4.33 .73 .56

 

59

3. V E B “neoﬁtmwm was $833.28 3.4 “stowage 2% E hmaﬁgn and 9.3533me 38%

 

 

mm. hm. we. Nm. E. C. N. NN. 3. cm. 314 #:0838800 :80 m 963 a
an. mm. ~N. ON. w_. 3. N. C. we. 9; EoEdEEoU 300 N 9.63 w
3.. NN. 2. MN. 2. N. 3. me. wmé EoEumEEoU 300 ~ 263 n
Na. no. em. am. S. a _. Nw. Rum HE Oi woZoBoE m 263 o
3. ac. Nm. ow. ON. 3.. and HE Oi “octagon N 0E? m
ca. wN. mm. N. as. vwd HE OE woﬁoocom _ 263 v
Eu vm. cm. 50. NN.m meowﬁoExm BE @oZoBom m 963 m
an. Nm. on. _N.m meoﬂSoExm BE boﬁoouom N 263 N
N». co. :6. mcouﬁoExm BE BEoBom _ 963 H
o m N. o m w m N M Om :32 So:

 

.2 we EEK? 69% 6:33. £39m.
m 033.

60

Data Analysis Steps

Data analyses were conducted in three major steps: 1.) establishing measurement
invariance across time points, 2.) building univariate latent growth models (LGMs), and
3.) constructing the mediated LGM. The item and parcel-level correlations for the
variables used in these analyses can be found in Appendix B. Each of these three steps is
discussed in further detail below.

Step 1: Establishing measurement invariance. Invariance means that “Individuals
who are identical on the construct being measured, but who are from different
populations, have the same probability of achieving any given score on the test” (Millsap
& Kwok, 2004, pp. 93). In other words, tests of invariance seek to answer the question,
“To what extent are manifest variables’ (i.e., Xs’) measurement properties transportable
or generalizable across populations?” (V andenberg & Lance, 2000, pp. 8). It has been
recommended that invariance tests be conducted prior to making comparisons between
demographic, cultural, or experimental groups (e. g., Chan & Schmitt, 1997); as a method
of identifying alpha, beta, and gamma forms of change over time (Golembiewski,
Billingsley, & Yeager, 1975; Schmitt, 1982); and as a precursor to conducting latent
growth modeling (e.g., Chan, 1998). Invariance testing prior to LGM is designed to
determine the extent to which the measures used are invariant across time periods. The
procedure for invariance testing is to examine a set of increasingly restrictive models.
This procedure, and the results of these tests, are discussed further below and are
summarized in Table 5.

Note that in order to simplify analyses, item parcels were created from the goal

commitment scale. Three parcels were created at each time point, and each parcel was

61

composed of the mean of two items. The scale was unidimensional in factor analyses;
therefore, parcel 1 was composed of the item with the highest factor loading and the item
with the lowest factor loading, parcel 2 was composed of the items with the second
highest and second lowest factor loadings, and parcel 3 was composed of the items with
the third highest and third lowest factor loadings (Hall, Snell, & F oust, 1999). This
procedure tends to produce uniquenesses that are maximally equivalent for each item
parcel.

Conﬁgural invariance. In the conﬁgural invariance test, a baseline model is
established based on relevant theory. The baseline model for the present study is
displayed in Figure 5. This model was applied across all three waves of data collection
and represents the model for which invariance was tested. The ﬁt statistics displayed in
Table 5, then, represent the ﬁt of this model with various constraints applied across time
points.

The conﬁgural invariance test assesses whether the pattern of ﬁxed and free factor
loadings holds across groups. In this model, three latent factors were speciﬁed, with the
perceived met expectations items set to load onto the perceived met expectations factor,
the perceived P-O ﬁt items set to load onto the perceived P-O ﬁt factor, etc. The ﬁt of
this model was assessed for each wave of data collection. Poor ﬁt at any given time point
would indicate that conﬁgural invariance does not hold. Although the )8 test is the most
common statistic used to assess model ﬁt, it is highly sensitive to sample size such that
when large samples are used, this index is almost always statistically signiﬁcant (Chan &
Schmitt, 1997). Therefore, I examined several different indices of model ﬁt including

Steiger’s (1990) root mean square error of approximation (RMSEA), Bentler’s (1990)

62

comparative ﬁt index (CPI), and Bentler and Bonnett’s (1980) non-normed ﬁt index
(NN F 1).

Because the pattern of factor loadings is an empirical representation of the
cognitive frame of reference participants use to respond to items, different patterns of
factor loadings at each wave of data collection would indicate that frames of reference
were not equivalent across groups (Vandenberg & Lance, 2000). If the conﬁgural
invariance test were to fail, it would indicate that different constructs were being
measured at each time point.

The syntax used to conduct the conﬁgural invariance test is displayed in
Appendix F. The global goodness-of-ﬁt statistics indicated that the baseline model ﬁt the

data closely (x2 = 244.63, p < .05, df= 123, RMSEA = .04, CFI = .99, NNFI = .98). The

x2 was statistically signiﬁcant; however, as stated above, this result was not surprising
given the relatively large sample size used. Overall, the goodness-of-ﬁt statistics
indicated close model ﬁt across time points, supporting the inference of conﬁgural

invariance.

63

 

 

£363 coast?»

8% 52 am. am. we. oodnm mmEom 3:3 on 8 meant?» .868 E83 Embmcou 85:5

mo> 5_ wo. wo. mo. NmNmm mo>m>> mmonow Race on 8 38833:: cﬁbmcov mommocosvmcb

mo> 02 3. am. we. wNNcN 8%? 36ch 323 on 8 meRB 88mm £83.80 oEoE

8% NM mo. am. we. moéVN 353 $88 :33 on 8 830:8...“ 88mm £88850 ESwEoU

aeoeoaaam a £22 £0 Ewe/E ax comatose use a oeaz
mountain

 

REESE 2Q PEESEN
o 833.

64

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2m in on mm mm on mm «m mw mm 6
30 3e 3e 3e 3e 3e 30 30 3e 3c 30
m :55 m .85 _ 38E v .8: m so: N Eu: _ as: v so: m :5: N :5: ﬂ 80:
m.mx< m.~X< m; X< N.vx< N.mX< N.NX< N. _ X< _.vx< _.mX< _NX< _._ X<

m.~e

MEG

 

88g ~39er 2V 3 akimnctﬁmm 852209 I ESSNESXQ 330: Nuxzmmbsb

m 2:5

65

Metric Invariance Test. This step of the measurement invariance process tests
whether the factor loading estimates (Ax matrices) are equivalent across time points. To
understand the implications of this test, recall that in the equation representing the

measurement model in SEM,
X = 2x: + 5 Equation]

the Ax parameter represents the regression weight of the latent variable on the observed
score. Each Ax parameter, then, represents the expected change in the observed item
score per unit change in the latent construct. This implies that if the Ax parameters are
signiﬁcantly different between time points, then beta change has occurred, or in other
words, respondents were using a different scale to respond to the items at different time
points (Schmitt, 1982; Vandenberg & Self, 1993).

In order to test for metric invariance, I contrasted the three-factor model I retained
in the test of conﬁgural invariance with a more restrictive model in which the Ax
parameters were constrained to be equal across time points. A chi-square difference test
is commonly used to determine whether the restricted model ﬁts the data signiﬁcantly
worse than the unrestricted model. However, because the chi-square test is strongly
affected by sample size, it has been recommended that a decrease in the CPI statistic of
.02 or more should be considered grounds for rejecting invariance (Cheung & Rensvold,
2002).

The syntax used to test metric invariance is displayed in Appendix G. In the

model created by the syntax, the factor loadings (Ax parameters) were constrained to be

equal across time points. This model also ﬁt the data well (x2 = 262.28, p < .05, df =

66

139, RMSEA = .04, CFI = .98, NNFI = .99). A x2 difference test was conducted in order
to determine whether this restriction resulted in a signiﬁcant decrease in model ﬁt. This
x2 difference test was nonsigniﬁcant (Ax2 = 17.65, Adf= 16, n. 5.), and the decrease in
CFI was less than .02, supporting the inference of metric invariance across waves of data
collection.

Invariance of Uniquenesses. Although conﬁgural and metric invariance are often
considered adequate for providing evidence of measurement invariance prior to
conducting LGM (Alwin & Jackson, 1981; Reise, Widaman, & Pugh, 1993; S6rbom,
1974), the invariance of the item uniquenesses can provide additional support. This test
evaluates the extent to which the measurement errors, or uniquenesses, are invariant
across waves (equivalence of the (95 matrix). Invariance of uniquenesses is tested by
comparing the previously retained model with a model in which the 95 matrix is
constrained to be equal across groups. In order to conduct this test, the model retained
from the metric invariance test was compared with a model in which the uniquenesses
were speciﬁed to be equal across waves. This test can provide evidence that the scale
reliabilities are equivalent across times and also that the effects of any unmodeled
predictors are consistent across times (DeShon, 2004). Although Chan (1998) states that
this standard is “extremely demanding, and most researchers recognize that it is
unrealistic to expect such extreme invariance to hold in actual data except in highly
contrived situations” (pp. 434), I conducted this test in order to obtain additional evidence
in support of invariance.

The syntax used to conduct this test is found in Appendix H. In this test, the error

variances for each item were constrained to be equal across waves. This very restrictive

67

model was contrasted with the model retained in the prior test (in which the Ax
parameters are constrained to be equal across waves). Like the other models tested, this
model exhibited good ﬁt (x2 = 352.32, df= 161, RMSEA = .05, CFI = .98, NNFI = .98).
The chi-square difference test was signiﬁcant (Ax2 = 90.04, Adf= 22, p < .05); however,
the CPI statistic did not decrease relative to the previous model, supporting the invariance
of the uniquenesses across waves of data collection (Cheung & Rensvold, 2002).

Factor Variance Invariance Test In some situations, a ﬁnal invariance test is
conducted which assesses whether the variances of the latent factors are invariant across
groups (the variances of the latent factors are contained in the diagonal of the (1) matrix).
In order to test for factor variance invariance, the previous retained model would be
compared with a model constraining the latent factor variances to be equal across groups.
When this type of invariance holds, it indicates that equivalent ranges of the construct
continua were used when responding to items at each time point (Vandenberg & Lance,
2000). Furthermore, it would support the inference of equivalent reliabilities based on
the establishment of uniqueness invariance (Vandenberg & Lance, 2000). If factor
variance invariance does not hold, it may indicate that beta change (i.e., differential scale
calibration) has occurred (Schmitt, 1982; Vandenberg & Self, 1993).

However, the factor variance test may not be a meaningful indicator of invariance
in the present study. When using LGM, a violation of invariance on this test might
simply indicate that substantive variance in slopes is present (Chan, 1998). For example,
if respondents have relatively homogeneous levels of perceived P-O ﬁt at Wave 1, but
then diverge over time, signiﬁcant differences in the variance of the latent perceived P-O

ﬁt factor would be expected over time. Because I believe that there is signiﬁcant variance

68

in change trajectories over time, I would expect that the variance of the latent factors
would vary across waves.

To conduct this test, the model retained after the uniqueness invariance test was
compared with a model in which the latent factor variances ((DS) were also constrained to
be equal. As expected, the model ﬁt the data well (x2 = 397.09, df = 167, RMSEA = .05,
CFI = .98, NNFI = .98), and the CPI did not decrease relative to the previous model, but
the chi-square test was signiﬁcant (Ax2 = 26.77, Adf= 6, p < .05). For perceived P-O ﬁt
and goal commitment, the factor variances increased over waves, which suggested that
individuals’ scores are likely to be diverging over time.

The common metric completely standardized parameter estimates for the fully
constrained model (including constrained factor variances) is displayed in Figure 6.

Step 1 Summary. Tests of measurement invariance are necessary prior to
conducting LGM analyses in order to make meaningful interpretations of results. When
conducting LGM, tests of conﬁgural and metric invariance are most essential (Chan,
1998). Additional evidence for measurement invariance is provided if the invariance of
the item uniquenesses can be established. In the present study, support for all three types
of invariance was provided by decreases in the CPI statistic of less than .02 (Cheung &

Rensvold, 2002). I next describe the tests of univariate LGMs.

69

 

 

 

 

 

 

 

 

 

 

8. mm. S.
33 A33 30
m 383 N $83 ~ 32$
3w. me. K.
2: n e
.5800
~80

mm. 9V. :. 2. NV. 3. 8. S.
33 3c 30 30 33 Se 33 3c
4 :5: m :5: m :5: _ as. 4 E»: m as: N :5: _ so:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/

»

mm. :m «d.

      

 

 

cc; H e
“E Oi
Bioobm

 

\

Na.

/

ch.

   
    

mV.

2: n 6
.95 82
Bioeom

mm.

   

\1

838th 6.23: 29.2569 Bmﬁhxcntoﬁ $633569 M53365 3303 333:5 BmSuhwnoU 33k

0 08mm

 

70

Step 2: Univariate LGMs. When conducting LGM analyses, it is important to
establish well-ﬁtting univariate models prior to conducting cross-domain analyses
(Willett & Sayer, 1994). These univariate analyses serve two major purposes: 1.) they
determine the nature of the change occurring over time, and 2.) they assess the extent to
which there is variance in change over time. If there is not signiﬁcant variance in the
way in which responses are changing over time, hypotheses regarding relationships
between the slopes of different constructs cannot be tested. This is analogous to having

insufﬁcient scale variance to test correlations between constructs.

The generic model for univariate LGMs is displayed in Figure 7. An LGM is a
hierarchical factor model in which each of the individual items or parcels is speciﬁed to
load onto the relevant Time factor, and each Time factor in turn loads onto second-order
latent intercept and slope variables. Note that for factor scaling purposes, one item at
each time point must be ﬁxed to 1.0 (Kline, 1998). The uniquenesses of the same item
across time points are allowed to covary in order to reﬂect shared variance due to item
wording. These correlated uniquenesses are a common feature of longitudinal models in
SEM (Kline, 1998). Because the intercept represents the ﬁxed value of the construct at
Time 1 and does not change over time, all paths loading onto the intercept factor were
ﬁxed to 1. The slope factor loadings are ﬁxed to represent the hypothesized nature of
change over time. For example, a linear trajectory could be represented by factor
loadings of 0.0, 1.0, and 2.0; while a quadratic trajectory could be represented by factor
loadings of 0.0, 1.0, 4.0, and 16.0. Ideally, the third factor loading would be left free to

be estimated by the program; however, this often causes problems with model

71

convergence. If the model is unable to converge with the third factor loading freed, it is
recommended that this factor loading be initially ﬁxed to represent a linear change
trajectory (Willett & Sayer, 1994). If this linear trajectory demonstrates poor model ﬁt,
the researcher may ﬁx the third factor loading to represent another trajectory as

appropriate.

Figure 7
Basic Univariate LGM

 

Perceived Met Expectations. As described above, the slope factor loadings for
perceived met expectations were initially ﬁxed to represent a positive, linearly increasing
trajectory. This linear model ﬁt the data closely according to several goodness-of—ﬁt
indices 08:98.73, df=49, RMSEA = .04, CFI = .97, NNFI = .96). Therefore, the positive
linear model was retained. The mean slope estimate was K = .04 (n. 3.), indicating that
the average change trajectory for met expectations was not signiﬁcantly different from

zero. However, the variance in slopes was signiﬁcant (p < .05). The signiﬁcance of the

72

variance indicated that although the average form of change was not signiﬁcantly
different from zero, there were individual differences in the ways that some individuals

were changing over time.

Perceived P-O Fit. A separate univariate LGM was conducted in order to identify
the nature of changes in perceived P-O ﬁt over time. A linear trajectory also ﬁt these
data acceptably (x2 = 198.82, df=49, RMSEA = .07, CFI = .98, NNFI = .97). The mean
slope was K = -.04 (p < .05), indicating that the average change over time was
signiﬁcantly different from zero and slightly negative. Like perceived met expectations,
the variance of perceived P-O ﬁt slopes was signiﬁcant (p < .05), indicating that

sufﬁcient variance in slopes existed for further testing.

Goal Commitment. Finally, the univariate LGM for goal commitment ﬁt the data
closely (x2 = 44.17, df=22, RMSEA = .04, CFI = .99, NNFI = .98). The mean slope
estimate was K = -.05 (p < .05), and the variance of the slopes was statistically signiﬁcant

(p < .05).

Step 2 Summary. In summary, the univariate LGMs indicated that a linear
change model ﬁt the data for each of the three constructs. In addition, all three constructs
showed signiﬁcant variance in the slope estimates, indicating that the hypotheses

regarding slope relationships could be tested.

Step 3: Mediated LGM The ﬁnal step in the data analysis was to construct the
mediated LGM used to test the hypotheses presented previously. This mediated model

combines the LGM method with the traditional SEM mediation framework. 1 will begin

73

by describing the traditional method for testing mediation in SEM, and then I will
illustrate how my hypotheses were tested within this framework.

Traditional Mediation. Although multiple methods for testing the signiﬁcance of
indirect/mediation effects have been proposed (e.g., Baron & Kenny, 1986), the product
of coefﬁcients technique is both commonly-used and highly recommended (Kenny, 2006;
MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). This technique calculates the
indirect effect of X on Z by multiplying the coefﬁcient for the path from X to Y by the
coefﬁcient from the path from Y to Z. The logic behind this test is that X has some direct
effect on Y, but only a portion of that effect is transferred onto Z. This product term is
then divided by its standard error, which is then compared to a Z table in order to
determine its statistical signiﬁcance. Various methods for computing the standard error
of the product term have been proposed. The most commonly-used of these is the Sobel
(1982) formula (MacKinnon, et al., 2002), which is displayed in Equation 2. This
formula is used by LISREL (J oreskog & Sorbom, 1993) in computing the signiﬁcance of

indirect effects (MacKinnon, et al., 2002).

 

2 2 2 2 ,
O'aﬂﬁm =Ja 0' ﬂ+ﬂ 0' a Equat10n2

An illustration of the mediation model using hypothetical data is displayed in
Figure 8. As demonstrated, the magnitude of the direct effect of A on B is .36 (found in
the Gamma (D matrix). In the Beta (ﬂ) matrix, we see that the direct effect of B on C is
.27. Therefore, the standardized indirect effect of A on C is .36*.27 = .10. LISREL
computes the signiﬁcance of this estimate using the Sobel (1982) method and produces a
Z-estimate of 4.04, which is signiﬁcant at p < .05. Therefore, the indirect effect of A on

C is signiﬁcant, and the mediation hypothesis would be supported.

74

In order to determine whether the mediation effect should be considered full (the
effect of A on C operates entirely via A’s effects on B) or partial (A also has a direct
effect on C), a second model is speciﬁed. In the second model, in addition to the
mediated path, we also specify a direct link between variables A and C. This second
model, therefore, is a partially-mediated model. The goodness-of-ﬁt statistics for the
fully and partially mediated models are compared. If the ﬁts of the two models are not
signiﬁcantly different according to a chi-square difference test, we retain the more
parsimonious fully-mediated model. If the partially-mediated model shows signiﬁcantly
better ﬁt, the partial mediation model is retained.

Figure 8
Traditional Mediation Model

 

 

 

 

 

 

 

 

 

 

 

 

 

Mediated LGM. In the present study, I combined the LGM technique with the
mediation framework presented above in order to examine the hypotheses presented
previously. The three univariate LGMs constructed in Step 2 were combined in order to
examine the relationships among intercepts and slopes. The model tested is presented in
Figure 9. The technique used here is similar to the technique described by Cheong,
MacKinnon, and Khoo (2003); however, it is somewhat different in that the Cheong, et
al. procedure used a single manifest predictor in the place of latent variable A. To my
knowledge, this study represents the ﬁrst demonstration of a mediated LGM using three

latent intercept factors and three latent slope factors.

75

Because the technique to be used is new, it was difﬁcult to calculate a sample size
required to achieve adequate statistical power. However, a large sample size may be
needed. It has been suggested that when using LGM, sample sizes over 500 are often
required to detect signiﬁcant covariances between latent slope and intercept terms when
three to four waves of data collection are used (Hertzog, Lindenberger, Ghisletta, & Von
Oertzen, 2006). The sample used in the present study consisted of 566 participants.

Overall Model Fit

The overall ﬁt of the mediated LGM was examined using several goodness-of-ﬁt
statistics (x2 = 873.85,p < .05, df= 475, RMSEA = .04, CFI = .98, NNFI = .98). As a
whole, the goodness-of-ﬁt statistics indicated that the mediated LGM ﬁt the data closely
(Browne & Cudeck, 1992). The full LISREL syntax command ﬁle for the model tested
can be found in Appendix I. The completely standardized path estimates for the fully-
mediated LGM are displayed in Figure 10.

Hypotheses 1 and 2

Hypotheses 1 and 2 concerned the relationships between the perceived met
expectations and perceived P—O ﬁt intercepts (Hypothesis 1) and slopes (Hypothesis 2).
The standardized relationship between the intercepts was signiﬁcant (,6 = .59, p < .05),
supporting Hypothesis 1. In addition, the standardized relationship between the
perceived met expectations and perceived P-O ﬁt slopes was also signiﬁcant (,6 = .99, p <
.05), supporting Hypothesis 2. Therefore, perceived met expectations and perceived P-O
ﬁt were signiﬁcantly related at Wave 1 of data collection, and the change trajectories of

these variables from Wave 1 to Wave 3 were also signiﬁcantly related.

76

Figure 9

eptual Model — Mediated LGM

Conc

 

77

«39:5:— 30235 309.8:—
EoO 78G . «E Oi . num— ~92

 

EON 326.8»: 53% I 282$ 28%. 33 38x25 S\§EE.EN Sam nmnﬁxownu‘m. ARENQSQD

2 05mm

78

Hypotheses 3 and 4

The relationships between perceptions of P-O ﬁt and goal commitment intercepts
and slopes were also tested (Hypotheses 3 and 4, respectively). The standardized path
estimate between the intercepts of perceived P-O ﬁt and goal commitment was signiﬁcant
(,6 = .38, p < .05), indicating that perceptions of P-O ﬁt and goal commitment were
signiﬁcantly related at Wave 1 of data collection. Also signiﬁcant was the standardized
path estimate between the perceived P-O ﬁt and goal commitment slopes (,6 = .66, p <
.05), suggesting that perceptions of P-O ﬁt and goal commitment changed similarly over
time.

Hypotheses 5 and 6

Hypotheses 5 and 6 proposed that perceptions of P-O ﬁt will mediate the
relationship between perceptions of met expectations and goal commitment, both for
intercepts (Hypothesis 5) and slopes (Hypothesis 6). As previously discussed, these
hypotheses were tested using the product of coefﬁcients method (Mackinnon, et al.,
2001; Sobel, 1982). The standardized indirect effect of the perceived met expectations
intercept on the goal commitment intercept was statistically signiﬁcant (indirect effect =
.22, p < .05). Thus, at Wave 1, perceptions of P-O ﬁt mediated the relationship between
perceptions of met expectations and goal commitment, supporting Hypothesis 5.

In addition, the standardized indirect effect of the perceived met expectations
slope on the goal commitment slope was also signiﬁcant (indirect effect = .66, p < .05),

supporting Hypothesis 6. This indicates that perceptions of met expectations and goal

79

commitment changed in similar ways, and that this effect was mediated by changes in
perceptions of P-O ﬁt.

Variance Accounted For

In addition to considering the model ﬁt statistics and the statistical signiﬁcance of
the path estimates, it is also important to examine the degree to which the proposed
model accounts for variance in the outcome of interest — in this case, goal commitment.
If the model fails to account for some meaningful amount of variance in the outcome, the
model cannot be considered practically signiﬁcant. In the present study, the model
accounted for 14% of the variance in the goal commitment intercept (goal commitment at
Wave 1) and 41% of the variance in the goal commitment slope (changes in goal
commitment from Wave 1 to Wave 3).

Partial Mediation Tests

As an alternative to the fully mediated model presented above, models in which
perceptions of P-O ﬁt were speciﬁed as a partial mediator was tested. A signiﬁcantly
better ﬁt of the partial mediation model would suggest that perceptions of met
expectations have direct as well as indirect effects on goal commitment. If the ﬁts of the
two models do not differ signiﬁcantly, the more parsimonious fully mediated model
should be retained.

The ﬁrst alternative model tested was one in which only the intercept terms were
speciﬁed as partial mediators. The completely standardized path estimates for the
intercept and slope factors in this partial intercept mediation model are displayed in
Figure 11. With the addition of the direct effect from the perceived met expectations

intercept to the goal commitment intercept, the path from the perceptions of P-O ﬁt

80

intercept to the goal commitment intercept became nonsigniﬁcant. In addition, the
indirect effect of the perceived met expectations intercept on the goal commitment
intercept became nonsigniﬁcant (ab = .08, n.s.). Therefore, if this model were to be
retained, it would imply that one’s Time 1 perceptions of met expectations would be
directly related to one’s Time 1 goal commitment, rather than indirectly via perceptions
of P-O ﬁt. A chi-square difference test was conducted in order to determine whether the
partial mediation model ﬁt the data signiﬁcantly better than did the fully mediated model.
This partially mediated model ﬁt the data well (12 = 872.22, df = 474, RMSEA = .04,
CFI = .98, NNFI = .98, GFI = .91). However, the chi-square difference test was non-
signiﬁcant (Ax2 = 1.63, Adf = 1, n.s.), indicating that there was no signiﬁcant difference
in the goodness-of-ﬁt for the fully and partially mediated models. Therefore, the more
parsimonious fully-mediated model was retained for intercepts.

Figure 11

Partially Mediated Model for Intercepts Including Completely Standardized Path

Estimates

 

 

.34

 

 

  
   

  
     
     
   

     
    
   
   
   

      
     

Goal
Comm.
Intercept

Perceived
P-O Fit
Intercept

Perceived
Met Exp.
Intercept

.64 .15

-.35*

-. 1 6 ,
Perceived

P-O Fit
Slope

Perceived
Met Exp.
Slope

.71

81

Next, partial mediation was tested for the slope factors. The model tested,
including completely standardized path estimates, is displayed in Figure 12. In this
model, when the direct path was added between the slope of perceived met expectations
and the slope of goal commitment, the path from the perceived P-O ﬁt slope to the goal
commitment slope became non-signiﬁcant, as did the indirect effect estimate for
perceived met expectations slope on goal commitment slope (ab = -.28, n. 5.). If this
partially mediated model were to be retained, then, we would conclude that changes in
perceptions of met expectations were directly linked with both changes in P-O ﬁt and
change in goal commitment, and there would be no evidence for a mediation effect for
slopes. The results of the chi-square difference test were signiﬁcant (Ax2 = 6.58, Adf = 1,
p < .05). Therefore, the partially mediated model was retained for the slope factors, and
because the indirect effect in this model was nonsigniﬁcant, the model provides no
evidence in favor of a mediation effect for the slope factors.

Figure 12

Partially Mediated Model for Slopes Including Completely Standardized Path Estimates

   
     
   
     
 

     
 

     
  
     
   

  

Goal
Comm.
Intercept

Perceived
P-O Fit
Intercept

Perceived _
Met Exp.
Intercept

.65 .39

    

-,24* .09

-.23 .05

Perceived Perceived Goal
MCI EXP. .79 P-O Flt _34 Comm.
Slope Slope Slope

 

 

1.12

 

 

Results Summary

82

The ﬁnal, retained model, including completely standardized parameter estimates,
is displayed in Figure 13. In this model, the levels of perceived met expectations and P-O
ﬁt at Wave 1 were signiﬁcantly related, as were the levels of perceived P-O ﬁt and goal
commitment. Furthermore, the indirect effect of Wave 1 level of perceived met
expectations on the Wave 1 level of goal commitment was signiﬁcant, suggesting that
perceived P-O ﬁt fully mediated the relationship between these two variables. Overall,
the model accounted for 14% of the variance in Wave 1 level of goal commitment.

Changes in perceptions of met expectations from Wave 1 to Wave 3 were
signiﬁcantly related to changes in perceptions of P-O ﬁt, such that as one increased, the
other increased, and vice versa. In the ﬁnal model, changes in perceptions of met
expectations were also positively and signiﬁcantly related to changes in goal
commitment. However, changes in perceived P-O ﬁt were unrelated to changes in goal
commitment. Furthermore, in the ﬁnal model, the indirect effect of changes in perceived
met expectations on changes in goal commitment was nonsigniﬁcant; therefore, there was
no evidence for a mediation effect. Overall, the model accounted for 77% of the variance
in changes in goal commitment.
Relationships with Distal Outcomes and Supplemental Analyses

The present study examined the relationships of perceptions of met expectations
and perceptions of P-O ﬁt with goal commitment. Because past research has
demonstrated signiﬁcant links between goal commitment and task performance as well as
with turnover, the results found here have implications for both organizational
performance and employee retention. Although formal tests of the relationships between

goal commitment and such outcomes are beyond the scope of this paper, I have provided

83

the correlations between goal commitment and college GPA (adjusted for college
difﬁculty/selectivity) and intent to turnover in Table 6 for the participants in this study.
As displayed in the table, goal commitment was signiﬁcantly related to both GPA and
intent to turnover. The relationships were particularly strong at concurrent waves of data
collection.

The present study focused on person-environment interactions and the
relationships of perceptions resulting from such interactions on goal commitment. The
speciﬁc person and situation elements that lead to the formation of such perceptions are
beyond the scope of this study. However, in order to conclude that perceptions of met
expectations and P-O ﬁt add something to the prediction of goal commitment over and
above the situation itself, supplemental analyses were conducted. Because between-
school differences (such as differences in freshman orientation procedures) might be
associated with the variables in the present study, these analyses are reported in order to
provide preliminary evidence that the perceptions of interest predict goal commitment
over and above institutional afﬁliation. It is important to note that these analyses are only
preliminary and are not intended to imply a strict test of the relative effects of pure
situations and self-environment perceptions, particularly because there is likely to be a
wide variety of situations encountered even within a particular school.

A hierarchical linear regression was conducted in order to determine whether
perceptions of met expectations and perceptions of P-O ﬁt predicted goal commitment
over and above institutional membership. Dummy codes reﬂecting institutional
membership were entered in Block 1, with Michigan State serving as the required

reference group. Wave 1 perceived met expectations and perceived P-O ﬁt were entered

84

in Block 2. The dependent variable for this analysis was Wave 2 goal commitment; thus,
the regression tested the extent to which perceptions of met expectations and P-O ﬁt
predicted future goal commitment over and above the effects of institutional membership.

The results of this regression are reported in Table 7.

Table 7

Eﬂects of Wave 1 Perceived Met Expectations and Perceived P—O Fit on Wave 2
Goal Commitment Over and Above Institutional Affiliation

 

 

 

 

Block 1 R2 ARZ Variable B

.039 .039* Constant 4.42*
Fullerton -.08
Michigan State (reference group)
U. Chicago .10
Winston-Salem .09
Spelman .02
Iowa .01
Indiana -.01
Ohio State .12*
U. Michigan .11*
VA Tech. .06

Block 2 RT AR2 Variable B

.077 .038* Constant 3.83*
Fullerton -.07
Michigan State (reference group)
U. Chicago .08
Winston-Salem .08*
Spelman .00
Iowa .04
Indiana .00
Ohio State .11*
U. Michigan .13 *
VA Tech. .07
W1 Perceived Met .11*
Expectations
W1 Perceived P-O Fit .14*

 

* p < .05 Dependent variable: Wave 2 goal commitment.

As Table 7 shows, the dummy coded institutional variables were signiﬁcantly
associated with goal commitment, accounting for 3.9% of the variance (p < .01). Adding

perceived met expectations and perceived P-O ﬁt almost doubled the amount of variance

85

accounted for (AR2 = .038, p < .01). A second analysis was conducted which was
identical to the ﬁrst except that Wave 3 goal commitment was the dependent variable and
Wave 2 perceptions of met expectations and goal commitment were used. The results of
this analysis were very similar to the ﬁrst; the addition of the perceptions in Block 2 more
than doubled the amount of variance in goal commitment captured by the model. The
results of these regressions therefore provide some indication that perceptions of met
expectations and P-O ﬁt add to the prediction of goal commitment over and above the

predictive power of the situation.

86

wa. M £22 .3. n Eu .3. u Eng .8. v a at. “as .53” u «a
be $355» 3&9 33:65 of 22393.? on: 3.83% mﬁ .202

          
     

\ngmi- Jﬂﬂtuﬁﬂmlwlli/
\\ \x, #4., v: {VA/Java... 7

.\ .\. \.\\.\ \\vv/\ X ..// .x. z// t
x \.. \\.\\\ xxx" .\ / \ J. \ / \V/x... , ,1,

max... 3.

 

      

. . a. .w a... w. 4.
E E f. .L. _.. .2.--.
. _ . r + a x a m a a
m». 3. ms. 2.. N». we. mm. 2.. 2.. am. R. 3. ma. mm. in no. 3.
_ g

 

         

   

«3.028:—
aam 32

33.52.—
EcU 13¢

E023:—
=h Oé

  

835th Sam nonwhxenaesw $233.50 Max:335 Eco: huESmk EEK

2 8am

87

 

 

3. v a ..
.thGUh Lm>©=§= 8 2x22» V23 utmshssob NBQM 3% LQ.\.~Q=©MB.~.Q max» 5 33thme MKS mmmuthEme .802
a. 1%. *2. at.- ..2.- an..- .me amm- aa. :2 8m 5393 3 pass m? w
S. .3. .2.- .2.- La.- .a? 1%.- No. a? 8m 5565 9 835 g h
as. 8.- 8.- t8.- ..Nm; Lam.- S. a? mom 5885 8 625 :5 a
2: an. .2. .8. :. a. 4?- 9% $6 m? m
2: km. *8. *3. we. 9% 2% $5 NB 4
ma. .3. .2.. am. .31. 8m 88 38 m? m
an. an ”a. 9:. 8m 88 Row 9». N
R. 3. Re 8m Boo Boo :5 a
w a a m a m N a 8 502 z

 

EEEEESSEN 35» 6:33ng Exehtﬁm .wtomxé 38m.
x 033.

88

Discussion

In the past, when goal commitment research has considered person-environment
interaction, it focused almost exclusively on mechanical forms of interaction, such as
non-additivity/statistical interaction, mediation effects, and independent direct effects
(e.g., Hollenbeck & Klein, 1987; Klein, et al., 1999; Locke, et al., 1988). The present
study considered the potential impact of the dynamic forms of interaction on goal
commitment, more speciﬁcally reciprocal inﬂuences and cognitive sensemaking. Based
on this interactive approach, I proposed that perceived met expectations and perceived P-
0 ﬁt, which represent the dynamic, emergent perceptions of an individual regarding some
aspect of him or herself within the organizational context, might be useful in
understanding the goal commitment process. This effort thus moves beyond the
relatively simple, mechanical conceptualizations of person-situation effects employed in
much past research and delves further into the mind of the “contextualized person.” In
doing so, it allows for the fact that when a person is embedded in a context, he or she is
affected by the organization and also becomes a part of the organization. It also allows
for the fact that individuals may perceive the same situation in different ways (Terborg,
1981).

In this study, perceived met expectations and perceived P-O ﬁt were found to be
signiﬁcantly associated with goal commitment. Perceptions of met expectations and
perceptions of P-O ﬁt were signiﬁcantly associated with each other at Wave 1, and
changes in these variables over time were also signiﬁcantly related. Perceived P-O ﬁt

was signiﬁcantly related to goal commitment levels at Wave 1, but changes in

89

perceptions of P-O ﬁt were not signiﬁcantly associated with changes in goal commitment
over time.

It was interesting to note that for concurrent relationships, perceptions of P-O ﬁt
fully mediated the relationship of perceived met expectations and goal commitment. In
other words, perceived met expectations had no signiﬁcant relationship with goal
commitment other than indirectly via its relationship with perceptions of P-O ﬁt.
However, no mediation effect was evident for slopes/change trajectories. The
implications of this seem to be that, consistent with past research, students make college
selection decisions based on their expectations about P-O ﬁt on valued dimensions
(Braxton, Vesper, & Hossler, 1995; Chapman, 1981). Once they have entered the
college, the relationship of their perceptions of met expectations with concurrent levels of
goal commitment is entirely dependent on the relationship with concurrent perceptions of
P-O ﬁt. However, the direction of changes in P-O ﬁt may be unrelated to the direction of
changes in goal commitment. Instead, changes in perceptions of met expectations were
strongly related to changes in goal commitment. These results imply that it is important
for researchers to examine both concurrent and longitudinal relationships among
variables, as the relationships may differ.

Perceived met expectations and perceived P-O ﬁt were used in the present study
to represent the dynamic forms of person-environment interaction. However, additional
constructs may also represent the dynamic forms of person-environment interaction when
conceptualized as perceptions arising from the dynamic interaction of person and
situation factors. For example, perceptions of racial or gender discrimination are thought

to result from a combination of personal factors (e.g., one’s demographic status and

90

group identiﬁcation) and situational factors (e. g., the attitudes and behaviors of others in
the organization, organizational policies and procedures; e. g., Operario & F iske, 2001).
Over time, these person and situation factors may interact to produce changes in
perceptions of whether discrimination is taking place. Individuals may experience
incidences of prejudice which then affect their subsequent attitudes, affect, and behavior.
These changes in behavior might then affect the extent to which others in the
environment have stereotypical or prejudiced attitudes, thereby affecting incidences of
discrimination, and so on. Thus, additional work on reciprocal and cognitive
sensemaking interactions might be conducted across a broad array of substantive topics.

It is worthy of note that the perceptual variables incorporated in the present study
are not thought to include any implicit effects of the person and situation on behavior;
that is, the respondent must be able to consciously access the perception of interest.
However, while the perception must by deﬁnition be explicit (i.e., consciously
accessible), the basic person and situation factors that give rise to the perception need not
be. For example, a person may perceive high levels of P-O ﬁt without being speciﬁcally
aware of which aspects of the self and the organization have caused this perception to
occur. Regardless of whether the perceptions of met expectations and P-O ﬁt emerge
from implicit or explicit person and situation factors, these perceptions are expected to
drive attitudes and behaviors.

From a broad theoretical standpoint, then, the combination of the reciprocal and
cognitive sensemaking forms of interaction might be conceptualized as displayed in
Figure 14. The perceptions, which represent the cognitive sensemaking approach, result

from the dynamic and reciprocal interaction of the person and situation over time, which

91

represents the reciprocal inﬂuences approach. These perceptions may partially mediate
the relationship between the person and situation factors and behavior. The mediation
effect of these perceptions is expected to be only partial in order to account for possible
implicit effects of both the person and situation factors on behavior.

The effect is proposed to be partial for the person factor because research has
shown that implicit (i.e., unconscious) attitudes and beliefs can affect behavior without
resulting in relevant conscious perceptions (Bargh, 2002; Bargh & Chartrand, 1999).
Similarly, partial mediation for the situation factor is consistent with past research
demonstrating that aspects of one’s situation can have implicit effects on one’s behavior
(Bargh, Chen, & Burrows, 1996; Latane & Darley, 1970; Schwarz & Clore, 1983).
Finally, consistent with the reciprocal inﬂuences approach, a feedback loop is displayed
which represents the fact that a contextualized person has effects on his or her situation

over time.

92

Figure 14

Heuristic Model — Combined Reciprocal Inﬂuences and Cognitive Sensemaking Forms of
Interaction

 

 

=35" - .\ ' \
/ f y’ \ '\.
Person
/ Factor x '

 

 

 

 

 

 

/
[I H Perceptions -____p Behavior K.

\\ Situdtion J \“
\ Factor / .

j
\ ge\/ ./'

 

 

 

 

 

 

 

 

 

 

 

 

 

al.
\

It is important to note that in this study, the goal to which respondents were
committed was held constant. All respondents reported their commitment to the goal of
graduating from college. However, this process could generalize across a wide range of
goals. For example, junior faculty typically set a goal of obtaining tenure. When
entering an academic institution, new faculty have expectations about the types of work
they will be required to perform. For instance, the new faculty member at a research
university might expect to spend most of his or her time conducting research. To the
extent that these expectations are perceived as being met, the faculty person should
believe that both he or she and the university value research productivity. In other words,
perceptions that his or her expectations are being met should be related to perceptions of
P-O ﬁt, or value congruence. If, however, perceptions of P-O ﬁt are low (e. g., the faculty
member values research but perceives that the university strongly values committee
service), the faculty member may become less committed to obtaining tenure at that

university and begin searching for other options. The above example illustrates that the

93

process proposed in the present study is not expected to be unique to the goal of
graduating from college, but that it is expected to generalize broadly across goals.

The present study examined incoming college students, but a parallel process may
occur for employees in organizations. If these results generalize to work situations, the
ﬁndings have at least two implications. First, organizations should pay attention to the
ways in which employees’ perceptions of met expectations are changing over time.
Decreases in perceptions of met expectations may be, in the long term, directly linked
with decreases in both perceptions of P-O ﬁt and decreases in goal commitment.
Perceptions of P-O ﬁt have been linked with important attitudinal and behavioral
outcomes, including tumover (Kristof-Brown, et al., 2005; Sims & Kroeck, 1994), while
goal commitment has been associated most consistently with performance (e. g., Locke, et
al., 1981). Thus, in order to avoid turnover and performance decrements, it may not be
sufficient to simply implement a one-time realistic job preview in order to deﬂate initial
expectations. Organizations may need to consistently monitor employees’ perceptions of
met expectations in order to detect signiﬁcant decreases. If such decreases are detected,
organizations might most easily intervene by changing aspects of the employee’s
organizational experience (e. g., job description, pay, etc.) to more closely match
expectations. If implemented effectively, such interventions are expected to increase
perceptions of met expectations, which may in turn be associated with increased P-O ﬁt
perceptions and goal commitment.

Second, when organizations are only interested in short-term levels of goal
commitment, a focus on perceptions of P-O ﬁt might be most beneﬁcial. Certain

organizational goals may be very short-term in nature. Because full mediation was found

94

for concurrent levels of goal commitment, organizations might be best served by efforts
to increase perceptions of P-O ﬁt, regardless of whether those efforts also increase
perceptions of met expectations. When more long-term, stable increases in goal
commitment are desired, however, organizations should emphasize both perceptions of P-
0 ﬁt and perceptions of met expectations for maximum results.

The present study has at least four implications for goal commitment models such
as those proposed by Locke, et al. (1988) and Hollenbeck and Klein (1987). First, the
ﬁnding that perceptions such as perceived met expectations and perceived P-O ﬁt are
linked with goal commitment implies that the cognitive processing element of the Locke,
et a1. (1988) model deserves additional theoretical and empirical attention. The notion
that one’s perceptions of the environment (whether accurate or inaccurate) drive one’s
behavior deserves increased attention, and further incorporation of individual differences
in environmental perceptions may be a useful addition to goal commitment models.

Second, the results of the present study imply that goal commitment models might
beneﬁt from further attention to the dynamic forms of person-environment interaction
(Terborg, 1981). While these models have incorporated the notion of mechanical person-
environment interactions, the present study suggests that consideration of dynamic forms
may lead to the identiﬁcation of new variables that might be used to predict goal
commitment, such as perceived met expectations and perceived P-O ﬁt.

Third, goal commitment researchers should devote increased consideration to the
role of time and changes over time. The present study suggests that the relationships

among variables at the same time point may differ from the relationships of their change

95

trajectories. Therefore, goal commitment models may need to distinguish between
concurrent and longitudinal relationships among variables.

Finally, perceived met expectations and perceived P-O ﬁt should be added to the
models as antecedents of goal commitment. These factors can be incorporated within the
existing models; however, the models may need to be expanded or modiﬁed to more
explicitly incorporate these types of variables. Where in the models these factors best ﬁt
is a question that requires future research, as discussed further below (see “suggestions
for future research”). For the time being, I present a preliminary model that might serve
as a starting point for future research endeavors.

Figure 15 presents a proposed integration of perceived met expectations and
perceived P-O ﬁt within the framework of the Hollenbeck and Klein (1987) model.
Figure 15 also incorporates the cognitive processing box ﬁom the Locke, et al. (1988)
model. Several elements in this proposed model are worthy of note. First, the
relationship of person and situation factors with the self-environment perceptions is
mediated by cognitive processing. This relationship is consistent with the cognitive
sensemaking approach (Terborg, 1981), which suggests that due to individual
differences, different persons will perceive and process situational elements differently.
Second, the proximal predictors of goal commitment (goal expectancy and goal
attractiveness) are proposed to mediate the relationship of self-environment perceptions
with goal commitment. It is also consistent with the heuristic model presented in Figure
14.

Third, I broaden the label in the performance box to include goal-related behavior

in general (independent of the value-laden performance judgment, which may be

96

inﬂuenced by factors other than the individual’s behavior). Finally, consistent with the
reciprocal interaction perspective (Terborg, 1981), I propose that the individual’s
behavior feeds backward, affecting elements of the person and the situation. Over time,
therefore, the variables in the model are proposed to reciprocally inﬂuence one another.
Figure 15

Proposed Integration of Self-Environment Perceptions Within the Hollenbeck and Klein

(1 98 7) Goal Commitment Model

 

Person Factors Situation Person Factors Situation i
Factors Factors 1

 

 

 

 

 

 

 

 

 

 

    
   

 

 

 

   
     

Cognitive Processing

/\

Perceived Met Perceived P-O
Expectations F it

M

Goal Expectancy Goal Attractiveness

\/

Goal Commitment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Goal
Level

 

 

 

 

1

Behavior /
Performance

 

 

 

 

 

 

 

 

 

Potential Limitations
There are at least four potential limitations to the present study -- the potential for

common method effects, the inability to make causal inferences, the construct

97

representativeness of the perceived met expectations measure, and the use of a student
sample. Each of these potential limitations is discussed in turn.

First, as in any study using self-report measures, the potential inﬂuence of
common method bias cannot be ruled out. The effects of common method may have
been rendered less likely, however, by three characteristics of the survey used. First, the
scales measuring perceptions of met expectations, perceptions of P-O ﬁt, and goal
commitment were each separated by approximately 50 items in the survey. This
separation of the scales may have created a temporal or psychological separation between
the measurement of the three constructs that could reduce the potential effects of the
common method (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Second, the survey
responses were conﬁdential, and the respondents were guaranteed that no identifying
information would be revealed to anyone associated with their organizations. The
guarantee of conﬁdentiality can reduce method effects related to social desirability
concerns (Podsakoff, et al., 2003). Finally, because the study was conducted
longitudinally, method effects can be expected to be limited due to the lag of several
months between waves of data collection (Podsakoff, et al., 2003).

The second potential limitation of the present study is that when using cross-
domain LGM analyses, it is not possible to make causal inferences. While the model
presented in the present study proposes that perceptions of met expectations may predict
perceptions of P-O ﬁt, the causal direction of this relationship cannot be established here.
Similarly, the direction of the relationship between perceptions of P-O ﬁt and goal
commitment cannot be established with certainty. Additional longitudinal research will

be necessary to ﬁrmly establish the causal nature of the relationships among perceived

98

met expectations, perceived P-O ﬁt, and goal commitment. Such future research might
employ cross-lagged panel designs to more ﬁrmly establish causal direction. A
combined autoregressive / LGM technique is also available when the data collection
includes at least four waves (cf. Bollen & Curran, 2004). Such analyses would allow the
research to make inferences regarding the causal direction of the proposed relationships.

Third, the use of archival data dictates that one must use the measures available.
Due to practical constraints on data collection, the scales used had a relatively small
number of items. In particular, the perceived met expectations and perceived P-O ﬁt
scales had only four items each. Because the perceived P-O ﬁt items assessed global
elements of P-O ﬁt, they may have been likely to cover a broad range of the latent
perceived P-O ﬁt construct.

However, in the perceived met expectations measure, three of the four items
assessed more speciﬁc factors regarding which one’s expectations could be met or unmet
(e. g., course organization, competence of professors). The fourth item assessed
perceptions of met expectations for the overall academic experience. The ﬁrst three
items may not have covered the entire range of one’s perceptions of met expectations,
even for the academic focus; other factors such as amount of homework or test difﬁculty
might also account for part of the construct domain. Furthermore, the restriction of the
items to the academic part of the organizational experience may further exclude latent
variance in the perceived met expectations construct (e.g., perceptions of met
expectations related to social relationships). The restriction of the focus to academic
matters reﬂected a concern for the practical implications of the study. Organizations can

more easily intervene to affect perceptions of met expectations for academic or task

99

related aspects of the organizational experience than they can for social aspects. Still, the
perceived met expectations measure seems likely to lack full construct
representativeness.

However, the extent to which a lack of full construct representation might affect
the results of the present study is unknown. If the portion of latent variance in perceived
met expectations that is left uncaptured by the measure is not associated with variance in
perceptions of P-O ﬁt and goal commitment, then the relationships found here would
have been inﬂated. Conversely, if the uncaptured variance is signiﬁcantly associated
with perceptions of P-O ﬁt and goal commitment, then the relationships found here
would have been deﬂated. I believe that the latter possibility is the more likely one. I
propose that perceived met expectations variance associated with social matters
(uncaptured variance in the present study) is likely to be related to one’s perceptions of P-
0 ﬁt and goal commitment; however, future research is needed to address this question.

The fourth potential limitation of the present study is the nature of the sample
used. The use of a student sample to make inferences relevant to organizations raises
generalizability questions. However, in the present study, the process examined is
expected to generalize across academic and workplace situations. The process supported
here is broadly psychological in nature, and the effects of perceptions of met expectations
and perceptions of P-O ﬁt are not expected to differ between student and employee
samples. Furthermore, several established similarities between college student and
organizational samples were previously presented which support the potential for

generalization of the present study’s results to organizational employees.

100

Perhaps a greater limitation on the generalizability of the results presented here is
the issue of nonrandom attrition. Students with initially lower levels of perceived met
expectations at Wave 1 were signiﬁcantly more likely to attrit from the study than were
students with higher initial perceptions of met expectations. Although this effect size was
small (d = -.13), the results of the present study must be interpreted in light of this effect.

Furthermore, supplemental analyses suggest that students with lower academic
ability (measured by SAT/ACT score) were also more likely to attrit from the study
between Wave 1 and Wave 3 (d = .80). Also, as Table 2 shows, participants in the
sample generally had higher than average SAT scores relative to their schools’ incoming
classes. Therefore, the results found in the present study may be more representative of
higher-ability individuals than of lower-ability individuals. Theoretically, higher-ability
individuals may perceive that they have more available alternatives to staying in an
organization in which they perceive low ﬁt, where as lower-ability individuals, who
perceive fewer available alternatives, may remain committed to an organization in which
they perceive poor ﬁt (Allen & Griffeth, 1999). Therefore, ability may moderate the
relationships among perceptions of met expectations, perceptions of P-O ﬁt, and goal
commitment. However, additional research is needed to examine this possibility and to
explore other potential moderators of the relationship presented here.

Suggestions for Future Research

The present study provided a ﬁrst step in the examination of whether perceptions
of the self-environment relationship, such as perceived met expectations and perceived P-
0 ﬁt, can add to our understanding of goal commitment. This study laid the groundwork

for further investigation, and several avenues for future research are open. I now present

101

suggestions for future research that can further our understanding of the relationships
found here.

First, while the present study established that perceived met expectations and
perceived P-O ﬁt show signiﬁcant relationships with goal commitment, it would be
useful to establish in future research how much variance in goal commitment can be
attributed to these perceptions relative to variance attributed to more traditional predictors
of goal commitment, such as incentive structure or participation in goal setting. Future
research might examine this question using both levels of goal commitment and changes
in goal commitment as outcomes of interest.

Second, future researchers may wish to examine whether the effects of perceived
met expectations and perceived P-O ﬁt differently affect each of the proximal antecedents
of goal commitment. It may be reasonable to hypothesize that self-organization
perceptions may have stronger effects on goal attractiveness than on goal expectancy.
The degree to which one perceives ﬁt with the organization could operate indirectly on
expectancy by affecting the extent to which the individual has access to needed materials
and support. However, the effects of perceived P-O ﬁt may be stronger for goal
attractiveness, as individuals and organizations with congruent values may be more likely
to agree on the types of goals that should be set and the types of reward systems that
should be put in place for goal achievement.

Third, researchers may wish to examine the degree to which the predictors
examined here are related to commitment to both self-set and organizationally-set goals.
The present study found signiﬁcant effects for perceived met expectations and perceived

P-O ﬁt on a (presumably) self-set goal, the goal of graduating from college. However,

102

one might hypothesize that the effects of self-organization perceptions might operate
more strongly on goals that are prescribed to the individual by the organization. When
setting their own goals, individuals may attempt to set goals that are somewhat aligned
with their personal values regardless of their perceptions of P-O ﬁt. It seems that when
goals are assigned to the individual by the organization, however, there is more likely to
be variance on goal attractiveness, and perceptions of P-O ﬁt (value congruence) are
more likely to predict this variance.

Finally, future researchers should consider how person and situation factors lead
to cognitive sensemaking perceptions. Edwards and colleagues (2006) have
demonstrated that individuals’ perceptions of a situation do not reﬂect separately-
obtained measures of the person and situation factors in a straightforward manner. The
attentional and information processing processes leading to the development of these
emergent perceptions currently remain shrouded in mystery, and the identiﬁcation of
relevant antecedents and moderators of this emergent process is an interesting and likely
complex question. Furthermore, a greater understanding of the psychological factors that
inﬂuence the development of self-organization perceptions can yield practical
interventions designed to affect the types of perceptions that emerge from an
organizational situation. The ability to inﬂuence these perceptions may have important
implications for organizationally-relevant outcomes such as goal commitment,
performance, and turnover.

Practical Applications
The research presented in the present study has practical implications for

organizations wishing to increase commitment to valued goals. The results of this study

103

suggest that goal commitment may be increased via interventions designed to improve
perceptions of met expectations and/or perceptions of P-O ﬁt. These interventions may
be targeted at the selection or socialization processes, or they might be implemented at
any time a new goal is assigned. Past research suggests interventions and techniques that
might be used to inﬂuence perceptions of met expectations and perceptions of P-O ﬁt, as
will be discussed below. However, a caveat is offered. As discussed previously, our
perceptions of situations may not necessarily mirror the reality of those situations. For
most psychologically healthy individuals, though, perceptions and reality are likely to be
related. The practical applications presented here may only be effective to the extent that
individuals’ perceptions are in tune with the reality of the organizational situation and
changes in that situation.

Realistic Job Previews. For several years, realistic job previews (RJPs) have been
used to give applicants a more accurate view of the positive and negative aspects of
working in the organization (Ilgen & Seely, 1974). Based on work demonstrating that
applicant expectations for the organization tend to be inﬂated, RJPs provide potential
organizational members with a greater degree of negative information about the
organization than would typically be provided during the selection process (Wanous,
1977). The theory behind RJPs is that by reducing initial expectations, individuals are
less likely to experience perceptions of unmet expectations after organizational entry,
thus reducing post-entry turnover. Meta-analyses of the effectiveness of RJPs support
their use for the reduction of turnover (Wanous, et al., 1992). The present study suggests
that in addition to the reduction of turnover, RJPs might also be beneﬁcial for the

improvement of commitment to organizational goals. Applicants who accept

104

membership in the organization following the RJP may have higher perceptions of met
expectations and may thus be more likely to show greater perceptions of P-O ﬁt and
stronger commitment to any organizational goal assigned following organizational entry.
However, as noted previously, organizations should also continue to monitor changes in
perceptions of met expectations following organizational entry.

Socialization. Once individuals have entered the new organization, the
socialization process begins. The results of the present study have implications for
suggested content and form of the socialization process. Research has shown that in
addition to inﬂuencing newcomers’ knowledge, attitudes, and behaviors, socialization
can affect perceived levels of P-O ﬁt (Cable & Parsons, 2001). Although several
socialization tactics exist (Jones, 1986; Van Maanen & Schein, 1979), two particular
tactics have been linked to increases in newcomers’ perceptions of P-O ﬁt, as discussed
further below: the use of content that is sequential and ﬁxed, and the use of social aspects
that are serial and investiture in nature (Cable & Parsons, 2001).

Socialization processes provide new organizational members with information
about the sequence of events they will encounter and the likely timeline for encountering
those events (sequential and ﬁxed content). For example, business organizations might
provide information about the typical career paths and requirements for promotion, while
academic organizations might provide students with counseling on the degree
requirements and sequence of required courses. These sequential and ﬁxed socialization
tactics have been linked with increased perceptions of P-O ﬁt (Cable & Parsons, 2001).
The implication is that in order to increase perceptions of P-O ﬁt, organizations should

provide new members with information regarding the typical or desired progression in

105

the organization, as well as information about how performance will be evaluated and
what resources are available for individuals. To the extent that perceptions of P-O ﬁt are
increased, goal commitment is expected to increase as well.

It has also been found that new members who were provided with experienced
mentors or role models (serial and investiture tactics) had increased levels of perceived P-
0 ﬁt relative to those who were not provided with such a mentor/role model (Cable &
Parsons, 2001). Thus, organizations may wish to invest in mentoring programs for in
which experienced employees or senior students are paired with newcomers in order to
provide information and social support. In order to achieve the greatest beneﬁts for
perceived P-O ﬁt from mentoring, organizations should be mindful of who is selected to
be a mentor. In selecting mentors, organizations should choose experienced persons who
themselves perceive high levels of P-O ﬁt. This is expected to increase newcomer
perceptions of P-O ﬁt, and thus have a positive inﬂuence on goal commitment.

Finally, the results of the present study imply that organizations might improve
goal commitment by increasing perceptions of P-O ﬁt at the time that the goal is
introduced. By linking the goal to values that are shared between the organization and its
members, organizations might increase the attractiveness of the goal. The author recently
observed an example of the effectiveness of this process in a small manufacturing
company. This organization had recently implemented a new program designed to
improve the cleanliness and organization of the factory. As part of this program, each
individual was assigned goals for keeping his or her area clean. Quite a bit of skepticism
and grumbling was overheard from the production personnel, who preferred to use their

time to produce product rather than “wasting” their time on “making the place look

106

pretty.” Management, sensing that the employees did not truly understand the purpose of
the program, then appealed to the shared values of safety and productivity. They argued
that the new program, in addition to making the factory look better, would also make the
plant more safe and productive, as production workers would no longer trip over tools
left in the aisles or have to spend valuable time searching for a tool that had been moved
from its proper location. Following this appeal to shared values, commitment to the new
program and its associated goals increased. The implication here is that organizations
should planfully consider the manner in which new goals will be introduced, and the
introduction of new goals should include appeals to shared values. Because these appeals
are likely to increase perceptions of P-O ﬁt, they are also likely to increase goal
commitment.
Conclusion

The study presented here represents a ﬁrst step in the examination of the extent to
which variables theoretically grounded in the reciprocal inﬂuences and cognitive
sensemaking forms of person-environment interaction (Terborg, 1981) can be used to
increase our understanding of goal commitment. The results suggested that perceptions
of met expectations and perceptions of P-O ﬁt are signiﬁcantly related to goal
commitment, and that perceived P-O ﬁt fully mediates the relationship between
concurrent levels of perceived met expectations and goal commitment but partially
mediates the relationship between changes in perceived met expectations and goal
commitment. Although the person-situation debate has a long history, the present study
takes a unique perspective on the complex nature of a contextualized person and links

this perspective to goal commitment. It is my hope that future research will continue to

107

examine the nature and effects of self-environment perceptions on goal commitment and

on other outcomes of interest.

108

APPENDICES

109

 

«gangsta .28.»on

 

 

 

 

 

coo—ho -
o N 5 8:8 a 98 N N _ A23: is: $95.3. {new nor-N
:2 0582 owes mEoemmﬂu :2 8302 owes Seesaw—o omen; mEoocmv-m—o .5
98 :2 02.363. Em: 8.: 235-3. 2: Eco-Q6 max-N
.33 _ 3 3R Cw voom .eonEoEom
. . . =30: . . m a note-mm m u 33 a. 83 B
383:2 353 333 x5 «94% 8 N a “2% a Eng 3 a C a C a is
€motoE< .3530»
"3sz Ba 6:33 am...“ mowammoE 8:86:35 3:36?
. 58:052. :33. 3:255 m<0 3:0 o: wszenxz 3332 new 3283 39: ES SENSEQQQ ~36QO
:motoE<-§o.£<v 3:on .3352 38329.8 =m
3:253 .3852 mﬁocom 563:0
m 88-0 omenwmﬁ 353.83
_. New 0 _ m 535?: com com a .3383 6%: “shunt-E x2332
.23: E252
26 no; 25.? . . . E EPA—m wczmoa
n20 E: b .HS . b t 3 we o3 >230ch mﬁznmeKQ 223:9me

05 :o saga-.85

 

 

110

98 m_maE-m =m 8 E3. Ewarm maggots ﬂcouam =m 9 Bow Ear-m .EEwoi
owwtm 8885
09330 89—on basic: baa-.333 coup—Em
ea $8.65 :30 amen-sea 2% 53:22 asses -esm 258:8
88m SawEBE
=ote§8§~ 335%.
a 3an

< XHQmem<

 

5583:: 88m Eu_mm-=8m:_3 Bu 03235 bunch—=0 8: Stag—£5 .802

 

20.3690 QUGN uc MUQBVENaB

 

 

 

 

 

 

m o 0% v m S “(883% SQSE 2R
EE.§6=< E::86=< 808:3 :2 2302 owhﬁ :m: 8502 ems: E
. . . . .E:Eo:v=< 38‘ “.3: 20.513, 2: 253% MEAN
voonQw 3m:
25$on . voontw 35:3 €993 vooQEE 3R5. :o N 32933. E: 33 3253 BS:
383 :w .
0 £285 0 o 0 38.563
Z @958 05 >30 Z Z Z 39: BS QEBBQQQ 3.8QO
cocﬁcoto
:8» 3E 3:095 mwamnms MRSFBE
coo E 33¢?an 82 39A “2 =< comm a 3383 GS: “$836“.xe x3532
0:3 338 =<
2233
558:28 deﬁance
. . . . .32: 05 .850 828255
32:; x83 65 :05
8 ﬂaoﬁoogog mo :3 ma xoizm 5 5:835 mo 05 Set .38 nngzﬁmk
40203 05 3865.53 3:0— w 3388 532 w BEE. 83 .886 :< Mz§zbmx\o zea$bwmQ
. . 5859a 3823 203 $5an . . . . .
catacoto 203 $38 so» a :8 :8 EPA am
.3863 among “we wEEOoE =< _ U 82 _ =<
E Eon 35m
lwmnéoo FEE—09m Degas gwinﬁz «32 333:5
cook 35?; .«o b65323 mo \EﬂgED 88m 030

 

6:53 a 23

.DmE 88w 29:3 05 @ﬁozu 2 332.28% 95%
$5 3898 Eocoaomom 5653:: 83m sawing—z 8 3:25; awozoo socﬁocowéme 8.“ anew 39:8 btoEE bEwEEovoa a 2 88on 3G 05. .632

111

APPENDIX B

Perceived Met Expectations Scale Items
Instructions. Think back to the expectations you had of what college would be like.

Now, indicate the extent to which your actual college experiences have met your

expectations.
1 2 3 4 5
Does not Meets Exceeds
meet expectations expectations

expectations

1. The way my courses are organized
2. The competence of the professors
3. The course availability

4. My overall academic experience

112

APPENDIX C
Perceived P-O Fit Scale Items

Please indicate the extent to which you agree with the following statements about your
school.

1 2 3 4 5
Strongly Disagree Neither Agree Strongly
Disagree agree nor Agree

disagree

1. My personality is a good match for this school
2. I am the right type of person for this school
3. My values match or ﬁt the values of this school

4. I feel like I belong at this university

113

APPENDIX D
Goal Commitment Scale Items

Please indicate the extent to which you agree with the following statements.

1 2 3 4 5
Strongly Disagree Neither Agree Strongly
Disagree agree nor Agree

disagree

1. It is hard to take the goal of graduating from college seriously” (R)

2. It is unrealistic to expect that I will graduate from college* (R)

3. I might rethink my goal of graduating from college, if things go differently than I
expect* (R)

4. Quite frankly, I don't care if I graduate from college or not* (R)

5. I am strongly committed to pursuing the goal of graduating from college*

6. It would not take much for me to abandon the goal of graduating from college (R)

7. I think that the goal of graduating from college is a good goal to shoot for

8. I am willing to put forth a great deal of effort, beyond what I would normally do,
to graduate from college“

9. I will not gain much by trying to achieve the goal of graduating from college (R)

* Indicates that the item was retained during scale reﬁnement

114

 

:. 2. 8. 2. 2. :. N_. 8. 8. 2. 2. 2. 2. 2. 8. 8. 2. 25525280 :5 mN
2. 2.. a. 2.. 2.. 8. _N. 8. NN. AN. 2. 2. 2. 2. 8. 8. :. 2. E oi 8285i 25 ..N
2.. 2... 2... 8. 2.. SN. 2. 2. NN. _N. 2. 2. 2. :. No. No. 8. m E oi 8285i 25 «N
8. 2.. 2.. 2.. 2.. 8. mN. NN. 8. a. _N. NN. NN. _N. 8. 8. 2. N E oi 8285i 25 NN
8. 2.. am. 2.. a. 3.. 8N. :N. x. 0N. ON. ON. 8. 8. 8. 8. 2. _ E oi 8283i 25 .N
a. in :.. 8. a. NN. 2. NN. NN. 8. 8. 8. cm. NN. 8. 2. :. 2. E oi 3202i 25 8N
2. 2.. 8. 2.. on. NN. 2. 2. mN. NN. NN. mN. MN. 2. :. 8. 2. N E oi 3283i 25 2
2. in. on. 8. E. mN. 2. 2. 8. 8. «N. N. 8. 2. 8. 8. 8. N E oi 8285i 25 2
2: mm. 2.. 2... 8. _N. 2. 2. mN. 2“. ON. 2. eN. 2. 8. 8. :. _ E oi 3383i 25 2
2: 8. E. E. 2.N. 2. 2. 2. MN. N. _N. «N. 8. 2. 2. 2. 2. E oi 8302i :5 2
2: s. No. MN. 2. 2. _N. N. NN. 2. 2. 2. :. 8. 2. N E oi 3283i :5 2
2: 8. _N. 2. 2. 2. 8N. 2. mN. 2. 8. N_. :. 2. NE oi 8283i :5 E
2: N. 2. 2. MN. MN. .N. NN. 2. NN. _ _. 2. _N. _ E oi 8285.2 :5 2
2: 8. 2.. on on 8N. 8. 8. NN. 2. 2. NN. .. .92 .22 8233i 25 2
2: Nm. NN. ON. 2.. MN. NN. 2. 8. 8. 2. m ixm 222 3285i 25 :
2: 3.. ”N. 2. 2.. NN. MN. 8. NN. N_. N .92 a: 2283i 25 2
2: mN. 2. mN. 3.. 2. 8. 2. w _. _ .92 a: 8283i 25 i
2: 2.. 2.. 2.. E. mN. _N. .N. 2. .92 a: 8285 NB 2:
2: ON. 2N. NN. 2m. 2. 2. m .92 .22 328$ 25 N
2: S. 8. 2. 2.. N. N .92 a: 8285i 25 e
2: 8. 2. SN. NM. _ .92 a: 8302i 25 m
2: cm. 2.. in. 2. .22 a: 8285i :5 ..
2: 2. 2. m .92 a: 8385i :5 n
2: 2.. N .92 a: 8285i :5 N
2: _ .92 22 828.8 :5 _

2 2 2 2 2 2 : 2 i a h e m .. n N _

 

m vans/Emma».

:bq 222.8»: 2.: 23.5 ﬂooxcmhﬁmQRo 222223.230
A: 2an

115

 

2: 2. 2. 2. 2. 2. 8. 2. 255258838 :5 2
2: S. 8. 2.2 8. 2.. NW. 2. E oi 8285i 25 ..N
2: 8. 8. 2.. in 2. N E oi 3285i 25 «N
2: 8. mm. 2.. E. N E oi 8283i 25 NN
2: E. 2.. No. _ E oi 8285i 25 _N
2: 3.. R. .12 oi 8283i 25 8N
2: E. N E oi 8285i 25 2
2: N E oi 8385i N5 2
_ E oi 8285i 25 2
2. E oi 8302i :5 2
N E oi 2383i :5 2
N E oi 3202i :5 2
_ E oi 8283i :5 2
.. .92 .22 3282 25 N—
N .92 22 3383i 25 :
N .92 a: 2285i 25 2
_ .92 22 8385i 25 i
2. .92 22 8283i 25 a
N .92 222 8385i 25 h
N .92 22 3285i 25 e
_ .92 22 838.2 25 m
2. .92 a: 8385i :5 ..
N .22 222 8283i :5 N.
N .92. $2 2385i :5 N
_ .92 .22 3285i :5 _
2. Na 5 an 8N 8 FN 2 2 ..N NN NN _N 8N 2 2

 

2.252 2 223

116

 

 

 

 

 

cc._ hm. cm. xv. cm. 2m. cm. Nm. VN. cm. _N. cm. 0N. E. Z. 2. m “co—525800 :30 m3 mm
cc; No. hm. av. Nm. aN. me. Nm. wN. cN. aN. cN. cN. m _. C. N #:2528280 :30 m3 Nm
cc._ cm. 2.. 9.. 0N. Nm. am. cN. N_. NN. cN. 2. ac. N_. _ 3253580 .200 m3 um
cc; mm. 5.. 9.. cv. mm. m _. 2. MN. NN. mN. o _. _N. m 2255:5800 :30 N3 cm
cc._ mm. cm. cm. mm. _N. 2. cN. a: M2. 2. C. N 25:98:80 300 N3 aN
cc._ NN. mm. 3.. ac. 5c. m _. N2. 3. cc. 2. _ Soﬁa—5:00 300 N3 cN
cc._ Nm. xv. M2. 2. a_. cN. :. c2. 2. m 222025252250 :80 :5 FN
cc._ 3. E. E. 2. M2. 2. cc. 3.. N 2:08225800 .2200 :5 9N
mm Nm 2” cm aN wN 5N cN mN VN MN NN —N cN a— c—
c_. c_. c_. I. :. VN. :. _N. N. E. c_. 2. c_. N_. cc. 2. c_. MycogméoUEoO m3 mm
2. _N. E. E. E. _N. C. 2. 3. f. c_. N_. C. 2. cc. ac. 3. NEuEtEEoOEoO m3 NM
2. c2. ac. wc. wc. cN. :. 3. S. 3. 2. N... 2. 2. c2. 2. cc. :congEoU 300 m3 —n
.N. C. 2. 2. M2. 2. 3. N_. 2. E. m_. cN. .2. ac. cc. 2. Z. May—0258280306 N3 cm
3. C. 2. m2. .2. 2. cc. no. N_. 2. c_. I. 3. ac. cc. hc. ac. Nazca—228500.80 N3 aN
mc. :. cc. ac. cc. N_. no. no. _ 2. E. 2. m _. 3. 3. cc. cc. hc. _ EoﬁEEEoU 300 N3 wN
3. NN. a_. MN. MN. I. N_. no. c2. 3. 2. E. I. 2. vc. cc. 2. misﬁt—5:00:30 :5 bN
:. 2. cc. 2. E. N_. N2. vc. vc. 3. N_. 2. ac. 2. mo. no. N_. NEoﬁzﬁﬁoUEoO :5 eN
5 3 m— z m— N— = c— a c b c m v m N _

 

€22.82 2: 2222.

117

APPENDIX F

SIMPLIS Syntax: Conﬁgural Invariance Test

Configural Invariance Test

group Wavel

observed variables: expl exp2 exp3 exp4 fitl fit2 fit3 fit4 goall goa12
goa13

latent variables: exp fit goal

correlations

1.00

0.40 1.00

0.18 0.14 1.00

0.39 0.43 0.30 1.00

0.21 0.10 0.11 0.27 1.00

0.16 0.11 0.12 0.28 0.86 1.00

0.14 0.08 0.11 0.17 0.62 0.61 1.00

0.18 0.13 0.10 0.33 0.74 0.74 0.58 1.00

0.14 0.05 0.05 0.14 0.21 0.20 0.13 0.21 1.00

0.11 0.07 0.07 0.14 0.13 0.14 0.06 0.14 0.46 1.00
0.11 0.03 - .01 0.11 0.19 0.19 0.15 0.19 0.55 0.51 1.00

standard deviations

.85 .99 1.17 .95 .89 .85 .93 .89 .46 .64 .60
sample size = 566

expl = 1*exp

exp2 exp3 exp4 = exp
fitl = 1*fit
fit2 fit3 fit4 = fit

goall = 1*goal

goa12 goa13 = goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the error variance of expl - goal3 be free
let the covariance of exp and fit be free

let the covariance of fit and goal be free

let the covariance of exp and goal be free

group Wave2

correlations

1.00

0.37 1.00

0.30 0.29 1.00

0.46 0.47 0.41 1.00

0.26 0.19 0.20 0.30 1.00

0.28 0.23 0.25 0.35 0.87 1.00

0.23 0.25 0.23 0.27 0.65 0.71 1.00

0.30 0.28 0.20 0.38 0.75 0.77 0.64 1.00

0.18 0.19 0.14 0.15 0.21 0.20 0.14 0.21 1.00

0.13 0.10 0.12 0.16 0.13 0.15 0.07 0.17 0.51 1.00
0.15 0.19 0.14 0.19 0.12 0.14 0.10 0.18 0.59 0.57 1.00

standard deviations
.81 .90 1.13 .95 .86 .84 .92 .89 .49 .66 .57

118

sample size
expl
exp2 exp3 exp4
fitl
fit2 fit3 fit4
goall =
goa12 goal3
let
let
let
let
let
let
let

.-

the
the
the
the
the
the
the

1*exp

1*fit

1*goal

566

exp

fi

goal

t

variance of exp be free
variance of fit be free
variance of goal be free
error variance of
covariance of exp
covariance of fit
covariance of exp

group Wave3

CDCDCDCDOI—|

expl — goal3 be free
and fit be free

and goal be free

and goal be free

.00
.68
.78
.30
.28
.32

.98

OOOOH

.93

.00
.67
.23
.15
.19

1.00

0.30 1.00

0.24 0.53 1.00

0.26 0.73 0.65 1.00

.56 .72 .65

expl - goal3 be free
and fit be free
and goal be free

correlations

1.00

0.44 1.00

0.27 0.32 1.00

0.50 0.42 0.38 1.00

0.34 0.21 0.26 0.34 1.00

0.35 0.22 0.25 0.38 0.89

0.27 0.14 0.15 0.26 0.68

0.32 0.20 0.21 0.33 0.74

0.22 0.17 0.13 0.23 0.27

0.15 0.16 0.15 0.23 0.23

0.19 0.17 0.13 0.22 0.26
standard deviations

.77 .92 1.04 .92 .89 .88

expl = 1*exp

exp2 exp3 exp4 = exp

fitl = 1*fit

fit2 fit3 fit4 = fit

goall = 1*goal

goal2 goal3 = goal

let the variance of exp be free
let the variance of fit be free
let the variance of goal be free
let the error variance of

let the covariance of exp

let the covariance of fit

let the covariance of exp

options set ad=off

path diagram

lisrel output:

end of problem

it=500

and goal be free

rs ef va tv sc mi

119

APPENDIX G

SIMPLIS Syntax: Metric Invariance Test

Metric Invariance Test

group Wavel

observed variables: expl exp2 exp3 exp4 fitl fit2 fit3 fit4 goall goal2
goa13

latent variables: exp fit goal

correlations

1.00

0.40 1.00

0.18 0.14 1.00

0.39 0.43 0.30 1.00

0.21 0.10 0.11 0.27 1.00

0.16 0.11 0.12 0.28 0.86 1.00

0.14 0.08 0.11 0.17 0.62 0.61 1.00

0.18 0.13 0.10 0.33 0.74 0.74 0.58 1.00

0.14 0.05 0.05 0.14 0.21 0.20 0.13 0.21 1.00

0.11 0.07 0.07 0.14 0.13 0.14 0.06 0.14 0.46 1.00
0.11 0.03 - .01 0.11 0.19 0.19 0.15 0.19 0.55 0.51 1.00

standard deviations

.85 .99 1.17 .95 .89 .85 .93 .89 .46 .64 .60
sample size = 566

expl = 1*exp

exp2 exp3 exp4 = exp
fitl = 1*fit
fit2 fit3 fit4 = fit

goall = 1*goal

goa12 goa13 = goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the error variance of expl - goal3 be free
let the covariance of exp and fit be free

let the covariance of fit and goal be free

let the covariance of exp and goal be free

group Wave2

correlations

1.00

0.37 1.00

0.30 0.29 1.00

0.46 0.47 0.41 1.00

0.26 0.19 0.20 0.30 1.00

0.28 0.23 0.25 0.35 0.87 1.00

0.23 0.25 0.23 0.27 0.65 0.71 1.00

0.30 0.28 0.20 0.38 0.75 0.77 0.64 1.00

0.18 0.19 0.14 0.15 0.21 0.20 0.14 0.21 1.00

0.13 0.10 0.12 0.16 0.13 0.15 0.07 0.17 0.51 1.00
0.15 0.19 0.14 0.19 0.12 0.14 0.10 0.18 0.59 0.57 1.00

standard deviations

.81 .90 1.13 .95 .86 .84 .92 .89 .49 .66 .57
sample size = 566

expl = 1*exp

120

fitl = 1*fit

goall = 1*goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the error variance of expl — goa13 be free
let the covariance of exp and fit be free

let the covariance of fit and goal be free

let the covariance of exp and goal be free

group Wave3
correlations

1.00

0.44 1.00

0.27 0.32 1.00

0.50 0.42 0.38 1.00

0.34 0.21 0.26 0.34 1.00

0.35 0.22 0.25 0.38 0.89 1.00

0.27 0.14 0.15 0.26 0.68 0.68 1.00

0.32 0.20 0.21 0.33 0.74 0.78 0.67 1.00

0.22 0.17 0.13 0.23 0.27 0.30 0.23 0.30 1.00

0.15 0.16 0.15 0.23 0.23 0.28 0.15 0.24 0.53 1.00
0.19 0.17 0.13 0.22 0.26 0.32 0.19 0.26 0.73 0.65 1.00

standard deviations

.77 .92 1.04 .92 .89 .88 .98 .93 .56 .72 .65
expl = 1*exp

fitl = 1*fit

goall = 1*goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the error variance of expl - goal3 be free
let the covariance of exp and fit be free
let the covariance of fit and goal be free
let the covariance of exp and goal be free

options set ad=off it=500

path diagram

lisrel output: rs ef va tv sc mi
end of problem

121

APPENDIX H

SIMPLIS Syntax: Uniqueness Invariance Test

Uniqueness Invariance Test

group Wavel

observed variables: expl exp2 exp3 exp4 fitl fit2 fit3 fit4 goall goa12
goal3

latent variables: exp fit goal

correlations

1.00

0.40 1.00

0.18 0.14 1.00

0.39 0.43 0.30 1.00

0.21 0.10 0.11 0.27 1.00

0.16 0.11 0.12 0.28 0.86 1.00

0.14 0.08 0.11 0.17 0.62 0.61 1.00

0.18 0.13 0.10 0.33 0.74 0.74 0.58 1.00

0.14 0.05 0.05 0.14 0.21 0.20 0.13 0.21 1.00

0.11 0.07 0.07 0.14 0.13 0.14 0.06 0.14 0.46 1.00
0.11 0.03 -0.01 0.11 0.19 0.19 0.15 0.19 0.55 0.51 1.00

standard deviations

.85 .99 1.17 .95 .89 .85 .93 .89 .46 .64 .60
sample size = 566

expl = 1*exp

exp2 exp3 exp4 = exp
fitl = 1*fit
fit2 fit3 fit4 = fit

goall = 1*goal

goa12 goa13 = goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the error variance of expl — goa13 be free
let the covariance of exp and fit be free

let the covariance of fit and goal be free

let the covariance of exp and goal be free

group Wave2

correlations

1.00

0.37 1.00

0.30 0.29 1.00

0.46 0.47 0.41 1.00

0.26 0.19 0.20 0.30 1.00

0.28 0.23 0.25 0.35 0.87 1.00

0.23 0.25 0.23 0.27 0.65 0.71 1.00

0.30 0.28 0.20 0.38 0.75 0.77 0.64 1.00

0.18 0.19 0.14 0.15 0.21 0.20 0.14 0.21 1.00

0.13 0.10 0.12 0.16 0.13 0.15 0.07 0.17 0.51 1.00
0.15 0.19 0.14 0.19 0.12 0.14 0.10 0.18 0.59 0.57 1.00

standard deviations

.81 .90 1.13 .95 .86 .84 .92 .89 .49 .66 .57
sample size = 566

expl = 1*exp

122

fitl = 1*fit

goall = 1*goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the covariance of exp and fit be free
let the covariance of fit and goal be free
let the covariance of exp and goal be free

group Wave3

correlations

1.00

0.44 1.00

0.27 0.32 1.00

0.50 0.42 0.38 1.00

0.34 0.21 0.26 0.34 1.00

0.35 0.22 0.25 0.38 0.89 1.00

0.27 0.14 0.15 0.26 0.68 0.68 1.00

0.32 0.20 0.21 0.33 0.74 0.78 0.67 1.00

0.22 0.17 0.13 0.23 0.27 0.30 0.23 0.30 1.00

0.15 0.16 0.15 0.23 0.23 0.28 0.15 0.24 0.53 1.00
0.19 0.17 0.13 0.22 0.26 0.32 0.19 0.26 0.73 0.65 1.00

standard deviations

.77 .92 1.04 .92 .89 .88 .98 .93 .56 .72 .65
expl = 1*exp

fitl = 1*fit

goall = 1*goal

let the variance of exp be free

let the variance of fit be free

let the variance of goal be free

let the covariance of exp and fit be free
let the covariance of fit and goal be free
let the covariance of exp and goal be free

options set ad=off it=500

path diagram

lisrel output: rs ef va tv sc mi
end of problem

123

SIMPLIS Command File for F ally-Mediated LGM

observed variables

APPENDIX I

w2aexpl w2aexp2 w2aexp3 w2aexp4

w3aexp1 w3aexp2 w3aexp3 w3aexp4

w4aexpl w4aexp2 w4aexp3 w4aexp4

w2gfitl w2gfit2 w2gfit3 w2gfit4

w3gfitl w3gfit2 w3gfit3 w3gfit4

w4gfit1 w4gfit2 w4gfit3 w4gfit4

w2gcom1 w2gcom2 w2gcom3

w3gcoml w3gcom2 w3gcom3

w4gcoml w4gcom2 w4gcom3

latent variables

w2aexp w3aexp w4aexp i_exp s_exp

w2gfit w3gfit w4gfit i_fit s_fit

w2gcom w3gcom w4gcom i_com s_com

correlation matrix

1.00

0.40 1.00

0.18 0.14 1.00

0.39 0.43 0.30 1.00

0.32 0.26 0.13 0.28 1.00

0.25 0.45 0.18 0.38 0.37 1.00

0.16 0.13 0.39 0.27 0.30 0.29 1.00

0.24 0.21 0.25 0.34 0.46 0.47 0.41 1.00

0.18 0.18 0.07 0.18 0.34 0.25 0.17 0.25 1.00

0.12 0.27 0.08 0.23 0.23 0.40 0.13 0.24 0.44 1.00

0.14 0.09 0.28 0.17 0.27 0.23 0.40 0.26 0.27 0.32 1.00

0.22 0.19 0.19 0.27 0.36 0.33 0.29 0.50 0.50 0.42 0.38 1.00

0.21 0.10 0.11 0.27 0.15 0.22 0.21 0.28 0.23 0.17 0.16 0.24 1.00

0.16 0.11 0.12 0.28 0.18 0.23 0.19 0.29 0.19 0.14 0.16 0.21 0.86 1.00
0.14 0.08 0.11 0.17 0.18 0.17 0.22 0.26 0.21 0.10 0.13 0.23 0.62 0.61
0.18 0.13 0.10 0.33 0.24 0.21 0.23 0.33 0.19 0.13 0.18 0.24 0.74 0.74
1.00

0.11 0.04 0.06 0.15 0.26 0.19 0.20 0.30 0.25 0.16 0.15 0.21 0.63 0.56
0.55 1.00

0.08 0.05 0.07 0.16 0.28 0.23 0.25 0.35 0.28 0.18 0.19 0.25 0.61 0.60
0.59 0.87 1.00

0.15 0.09 0.11 0.16 0.23 0.25 0.23 0.27 0.23 0.15 0.16 0.22 0.50 0.48
0.47 0.65 0.71 1.00

0.11 0.13 0.03 0.22 0.30 0.28 0.20 0.38 0.27 0.22 0.18 0.27 0.57 0.56
0.59 0.75 0.77 0.64 1.00

0.15 0.07 0.09 0.20 0.20 0.20 0.20 0.29 0.34 0.21 0.26 0.34 0.51 0.46
0.47 0.58 0.62 0.48 0.54 1.00

0.10 0.08 0.08 0.21 0.23 0.22 0.21 0.31 0.35 0.22 0.25 0.38 0.48 0.47
0.47 0.56 0.61 0.48 0.55 0.89 1.00

0.08 0.02 0.02 0.11 0.16 0.14 0.16 0.21 0.27 0.14 0.15 0.26 0.43 0.36
0.39 0.47 0.53 0.59 0.48 0.68 0.68 1.00

0.11 0.06 0.05 0.18 0.19 0.15 0.15 0.26 0.32 0.20 0.21 0.33 0.46 0.41
0.49 0.51 0.57 0.46 0.58 0.74 0.78 0.67 1.00

0.10 0.05 0.04 0.14 0.16 0.15 0.13 0.14 0.06 0.05 0.12 0.11 0.12 0.14
0.14 0.11 0.13 0.07 0.16 0.13 0.16 0.12 0.12 1.00

0.12 0.05 0.03 0.10 0.09 0.11 0.12 0.14 0.04 0.04 0.12 0.12 0.17 0.16
0.17 0.11 0.14 0.06 0.17 0.18 0.18 0.17 0.16 0.54 1.00

0.13 0.06 0.04 0.15 0.11 0.14 0.13 0.14 0.10 0.07 0.12 0.11 0.23 0.23
0.22 0.14 0.15 0.10 0.17 0.20 0.19 0.15 0.18 0.48 0.52 1.00

0.07 0.06 0.08 0.14 0.14 0.15 0.15 0.18 0.11 0.08 0.08 0.12 0.06 0.09
0.11 0.08 0.10 0.04 0.14 0.12 0.15 0.07 0.09 0.44 0.35 0.27 1.00

0.09 0.07 0.08 0.09 0.14 0.11 0.10 0.15 0.12 0.07 0.08 0.15 0.13 0.15

0.17 0.16 0.17 0.10 0.18 0.18 0.20 0.16 0.21 0.38 0.50 0.36 0.55 1.00

124

H

.00
.58

.42

.50

.60

.41

.39

.40

.50

.37

.08

.08

.19

.06

.13

0.11 0.10 0.06 0.09 0.17 0.20 0.15 0.17 0.18 0.12 0.14 0.17 0.18 0.18 0.16
0.17 0.21 0.21 0.16 0.23 0.22 0.23 0.16 0.15 0.37 0.40 0.45 0.47 0.55 1.00

0.08 0.10 0.10 0.10 0.16 0.12 0.16 0.14 0.16 0.14 0.11 0.20 0.08 0.08 0.09
0.10 0.10 0.12 0.08 0.13 0.20 0.27 0.12 0.20 0.39 0.32 0.26 0.43 0.41 0.36 1.00
0.16 0.09 0.08 0.15 0.17 0.12 0.10 0.18 0.14 0.13 0.17 0.21 0.17 0.16 0.17
0.21 0.11 0.17 0.13 0.20 0.26 0.29 0.20 0.28 0.32 0.43 0.29 0.32 0.49 0.37 0.62
1.00

0.10 0.11 0.06 0.12 0.10 0.15 0.10 0.16 0.24 0.21 0.11 0.24 0.11 0.11 0.10
0.10 0.10 0.15 0.11 0.17 0.26 0.30 0.21 0.30 0.24 0.32 0.36 0.34 0.36 0.48 0.50
0.57 1.00

standard deviations

0.85 0.99 1.17 0.95 0.81 0.90 1.13 0.95 0.77 0.92 1.04 0.92 0.89 0.85 0.93
0.89 0.86 0.84 0.92 0.89 0.89 0.88 0.98 0.93 0.57 0.58 0.54 0.61 0.60 0.54 0.66
0.68 0.65

means

3.08 3.39 2.67 3.30 3.14 3.42 2.92 3.35 3.15 3.39 3.01 3.34 3.85 3.90 3.69
3.93 3.78 3.80 3.62 3.84 3.79 3.81 3.66 3.84 4.54 4.58 4.60 4.47 4.49 4.51 4.44
4.43 4.44

sample size = 556

w2aexpl = 1*w2aexp

w2aexp2 — w2aexp4 = w2aexp
w3aexp1 = 1*w3aexp

w3aexp2 — w3aexp4 = w3aexp
w4aexpl = 1*w4aexp

w4aexp2 — w4aexp4 = w4aexp

w2gfit1 = l*w2gfit
w2gfit2 — w2gfit4 = w2gfit

w3gfitl = l*w3gfit
w3gfit2 — w3gfit4 = w3gfit

w4gfit1 = l*w4gfit
w4gfit2 — w4gfit4 = w4gfit

w2gcoml = l*w2gcom
w2gcom2 w2gcom3 = w2gcom

w3gcoml = 1*w3gcom
w3gcom2 w3gcom3 = w3gcom
w4gcoml = l*w4gcom
w4gcom2 w4gcom3 = w4gcom
w2aexp = l*i_exp

w3aexp = l*i_exp

w4aexp = l*i_exp

w2aexp = 0*s_exp

w3aexp = 1*s_exp
w4aexp = 2*s_exp
i_exp = const

s_exp = const

w2gfit = l*i_fit
w3gfit = l*i_fit
w4gfit = l*i_fit

w2gfit - 0*s_fit

125

w3gfit = l*s_fit
w4gfit = 2*s_fit
i_fit = const
s_fit = const
i_fit = i_exp
s_fit = s_exp
i_fit = s_exp
s_fit = i_exp
w2gcom = l*i_com
w3gcom = l*i_com
w4gcom = l*i_com
w2gcom = 0*s_com
w3gcom = 1*s_com
w4gcom = 2*s_com
i_com = const
s_com = const
i_com = i_fit
s_com = s_fit
i_com = s_fit
s_com = i_fit
let the error covariance

let
let

let
let
let

let
let
let

set

let
let
let

let
let
let

let
let
let

let
let
let

let
let
let

let

the error
the error

the error
the error
the error

the error
the error
the error

the error

the error
the error
the error

the error
the error
the error

the error
the error
the error

the error
the error
the error

the error
the error
the error

the error

covariance
covariance

covariance
covariance
covariance

covariance
covariance
covariance

variance of s_fit to 0

covariance
covariance
covariance

covariance
covariance
covariance

covariance
covariance
covariance

covariance
covariance
covariance

covariance
covariance
covariance

covariance

of w2gcoml
of w2gcoml
of w3gcoml

of w2gcom2
of w2gcom2
of w3gcom2

of w2gcom3
of w2gcom3
of w3gcom3

of w2gfitl
of w2gfit1
of w3gfitl

of w2gfit2
of w2gfit2
of w3gfit2

of w2gfit3
of w2gfit3
of w3gfit3

of w2gfit4
of w2gfit4
of w3gfit4

of w2aexp1
of w2aexpl
of w3aexp1

of w2aexp2

and
and
and

and
and
and

and
and
and

and
and
and

and
and
and

and
and
and

and
and
and

and
and
and

and

w3gcoml
w4gcoml
w4gcoml

w3gcom2
w4gcom2
w4gcom2

w3gcom3
w4gcom3
w4gcom3

w3gfit1
w4gfitl
w4gfitl

w3gfit2
w4gfit2
w4gfit2

w3gfit3
w4gfit3
w4gfit3

w3gfit4
w4gfit4
w4gfit4

w3aexp1
w4aexpl
w4aexpl

w3aexp2

126

be
be
be

be
be
be

be
be
be

be
be
be

be
be
be

be
be
be

be
be
be

be
be
be

be

free
free
free

free
free
free

free
free
free

free
free
free

free
free
free

free
free
free

free
free
free

free
free
free

free

let
let

let
let
let

let
let
let

options:

lisrel output:
end of problem

the
the

the
the
the

the
the
the

error
error

error
error
error

error
error
error

covariance
covariance

covariance
covariance
covariance

covariance
covariance
covariance

of
of

of
of
of

of
of
of

set ad=off it=500
path diagram

w2aexp2
w3aexp2

w2aexp3
w2aexp3
w3aexp3

w2aexp4
w2aexp4
w3aexp4

rs ef va tv mi sc

and
and

and
and
and

and
and
and

w4aexp2
w4aexp2

w3aexp3
w4aexp3
w4aexp3

w3aexp4
w4aexp4
w4aexp4

127

be
be

be
be
be

be
be
be

free
free

free
free
free

free
free
free

REFERENCES

128

REFERENCES

ACT. (2004). The role of academic and non-academic factors in improving college
retention. Iowa City, IA: Author.

Adkins, C. L., Russell, C. J ., & Werbel, J. D. (1994). Judgments of ﬁt in the selection
process: The role of work value congruence. Personnel Psychology, 47, 605—623.

Allen, D. G., & Griffeth, R. W. (1999). Job performance and turnover: A review and
integrative multi-route model. Human Resource Management Review Special
issue: Employee turnover, 9, 525-548.

Allen, D. G., & Nora, A. (1995). An empirical examination of the construct validity of

goal commitment in the persistence process. Research in Higher Education, 36,
509-533.

Alwin, D. F., & Jackson, D. J. (1981). Application of simultaneous factor analysis to
issues of factor invariance. In D. J. Jackson & E. F. Borgatta (Eds), Factor

analysis and measurement in sociological research (pp. 249-279). Beverly Hills,
CA: Sage.

Arthur, W. A., Jr., Bell, S. T., Villado, A. J ., & Doverspike, D. (2006). The use of person-
organization ﬁt in employment decision making: An assessment of its criterion-
related validity. Journal of Applied Psychology, 91, 786-801.

Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure,
process, and content. Psychological Bulletin, 120, 338-375.

Austin, J. T., & Villanova, P. (1992). The criterion problem: 1917-1992. Journal of
Applied Psychology, 77, 836-874.

Bargh, J. A. (2002). Losing conscsiousness: Automatic inﬂuences on consumer
judgment, behavior, and motivation. Journal of Consumer Research, 29, 280-285.

Bargh, J. A., & Chartrand, T. L. (1999). The unbearable automaticity of being. American
Psychologist, 54, 462-479.

Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct
effects of trait construct and stereotype activation on action. Journal of
Personality and Social Psychology, 71, 230-244.

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, 1 173-1 182.

129

Barrick, M. R., & Mount, M. K. (1991). The big ﬁve personality dimensions and job
performance: A meta-analysis. Personnel Psychology, 44, 1-26.

Bar-Yam, Y. (2002). Complexity rising: From human beings to human civilization, A
complexity proﬁle. In Encyclopedia of Life Support Systems. Oxford: UK.
EOLSS Publishers.

Bentler, P. M. (1990). Comparative ﬁt indexes in structural models. Psychological
Bulletin, 107, 238-246.

Bentler, P. M., & Bonett, D. B. (1980). Signiﬁcance tests and goodness of ﬁt in the
analysis of covariance structures. Psychological Bulletin, 88, 588-606.

Bollen, K. A., & Curran, P. J. (2004). Autoregressive latent trajectory (ATL) models: A
synthesis of two traditions. Sociological Methods Research, 32, 336-383.

Bosson, J. K., Swann, W. B., & Pennebaker, J. W. (2000). Stalking the perfect measure
of implicit self-esteem: The blind men and the elephant revisited? Journal of
Personality and Social Psychology, 79, 631-643.

Bowen, D. E., Ledford, G. E., & Nathan, B. R. (1991). Hiring for the organization, not
the job. Academy of Management Executive, 5, 35-51.

Braxton, J. M., Vesper, N., & Hossler, D. (1995). Expectations for college and student
persistence. Research in Higher Education, 3 6, 595-611.

Bretz, R. D., Ash, R. A., & Dreher, G. F. (1989). Do people make the place? An
examination of the attraction-selection-attrition hypothesis. Personnel
Psychology, 42, 561-581.

Browne, M. W., & Cudeck, R. (1992). Alternative ways of assessing model ﬁt.
Sociological Methods & Research, 21, 230-258.

Buckley, H. D. (1971). A comparison of freshman and transfer expectations. Journal of
College Student Personnel, 12, 186-188.

Burke, R. J ., & Deszca, E. (1982). Preferred organizational climates of Type A
individuals. Journal of Vocational Behavior, 21, 50-59.

Button, S. B., Mathieu, J. E., & Zajac, D. M. (1996). Goal orientation in organizational
research: A conceptual and empirical foundation. Organizational Behavior and
Human Decision Processes, 67, 26-48.

Cable, D. M. & DeRue, D. S. (2002). The convergent and discriminant validity of
subjective ﬁt perceptions. Journal of Applied Psychology, 87, 875-884.

130

Cable, D. M., & Judge, T. A. (1997). Interviewers’ perceptions of person-organization ﬁt

and organizational selection decisions. Journal of Applied Psychology, 82, 546-
561.

Cable, D. M., & Judge, T. A. (1996). Person-organization ﬁt, job choice decisions, and

organizational entry. Organizational Behavior and Human Decision Processes,
6 7, 294-31 1.

Cable, D. M., & Judge, T. A. (1994). Pay preferences and job search decisions: A person-
organization ﬁt perspective. Personnel Psychology, 4 7, 317-348.

Cable, T. A., & Judge, D. M., Boudreau, J. W., & Bretz, R. D. (1995). An empirical

investigation of the predictors of executive career success. Personnel Psychology,
48, 485-519.

Cable, D. M., & Parsons, C. K. (2001). Socialization tactics and person-organization ﬁt.
Personnel Psychology, 54, 1-23.

Cabrera, A. F., Nora, A., & Castaneda, M. B. (1993). College persistence: Structural

equations modeling test of an integrated model of student retention. The Journal
of Higher Education, 64, 123-139.

Caldwell, D. F ., & O’Reilly, C. A. (1990). Measuring person-job ﬁt using a proﬁle
comparison process. Journal of Applied Psychology, 75, 648-657.

Campbell, J. P. (1994). Alternative models of job performance and their implications for
selection and classiﬁcation. In: M. G. Rumsey, C. B. Walker, & J. H. Harris

(Eds) Personnel selection and classification (pp. 33-51). Hillsdale, NJ: Lawrence
Erlbaum Associates, Inc.

Campion, M. A., & Lord, R. G. (1982). A control systems conceptualization of the goal-

setting and changing process. Organizational Behavior and Human Performance,
30, 265-287.

Chan, D. (1998). The conceptualization and analysis of change over time: An integrative
approach incorporating longitudinal mean and covariance structures analysis
(LMACS) and multiple indicator latent growth modeling (MLGM).
Organizational Research Methods, 1, 421-483.

Chan, D. & Schmitt. N. (1997). Video-based versus paper-and-pencil method of
assessment in situational judgment tests: Subgroup differences in test

performance and face validity perceptions. Journal of Applied Psychology, 82,
143-159.

Chapman, D. W. (1981). A model of student college choice. The Journal of Higher
Education, 52, 490-505.

131

Chatman, J. A. (1989). Improving interactional organizational research: A model of
person-organization ﬁt. The Academy of Management Review, 14, 333-349.

Cheong, J ., MacKinnon, D. P., & Khoo, S. T. (2003). Investigation of mediational
processes using parallel process latent growth curve modeling. Structural
Equation Modeling, 10, 23 8-262.

Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-ﬁt indexes for testing
measurement invariance. Structural Equation Modeling, 9, 233-255.

Christiansen, N., Villanova, P., & Mikulay, S. (1997). Political inﬂuence compatibility:
Fitting the person to the climate. Journal of Organizational Behavior, 18, 709-
730.

Dabos, G. E., & Rousseau, D. M. (2004). Mutuality and reciprocity in the psychological
contracts of employees and employers. Journal of Applied Psychology, 89, 52-72.

DeShon, R. P. (2004). Measures are not invariant across groups without error variance
homogeneity. Psychology Science, 46, 137-149.

DeShon, R. P., & Landis, R. S. (1997). The dimensionality of the Hollenbeck, Williams,
and Klein (1989) measure of goal commitment on complex tasks. Organizational
Behavior and Human Decision Processes, 70, 105-116.

Donavan, D. T., Fang, X., Bendapudi, N., & Singh, S. N. (2004). Applying interactional
psychology to salesforce management: A socialization illustration. Qualitative
Market Research: An International Journal, 7, 139-152.

Donovan, J. J ., & Radosevich, D. J. (1988). The moderating role of goal commitment on
the goal difﬁculty-performance relationship: A meta-analytic review and critical
analysis. Journal of Applied Psychology, 83, 308-313.

Edwards, J. R. (1991). Person-job ﬁt: A conceptual integration, literature review, and
methodological critique. In C. L. Cooper, & I. T. Robertson (Eds.), International
review of industrial and organizational psychology (pp. 283-357). Oxford,
England: John Wiley & Sons.

Edwards, J .R., Cable, D. M., Williamson, 1. 0., Lambert, L. S., & Shipp, A. J. (2006).
The phenomenology of ﬁt: Linking the person and environment to the subjective

experience of person-environment ﬁt. Journal of Applied Psychology, 91 , 802-
827.

Fiske, S. T., & Taylor, S. E. (1991). Social cognition. New York, NY: McGraw-Hill
Book Company.

French, J. R. P. Jr., Rogers, W., & Cobb, S. (1974). Adjustment as person-environment
ﬁt. In D. Coelho & J. B. Adams (Eds.), Coping and adaptation. New York: Basic
Books.

132

Fry, L. W., Vitucci, S., & Cedillo, M. (2005). Spiritual leadership and army
transformation: Theory, measurement, and establishing a baseline. Leadership
Quarterly, 16, 835-862.

Gerdes, H., & Mallinckrodt, B. (1994). Emotional, social, and academic adjustment of
college students: A longitudinal study of retention. Journal of Counseling and
Development, 72, 281-288.

Getzlaf, S. B., Sedlaceck, G. M., Kearney, K. A., & Blackwell, J. M. (1984). Two types
of voluntary undergraduate attrition: Application of Tinto’s model. Research in
Higher Education, 20, 257-268.

Golembiewski, R. T., Billingsley, K., & Yeager, S. (1975). Measuring change and
persistence in human affairs: Types of change generated by OD designs. Journal
of Applied Behavioral Science, 12, 133-15 7.

Goodman, S. A., & Svyantek, D. J. (1999). Person-organization ﬁt and contextual

performance: Do shared values matter? Journal of Vocational Behavior, 55, 254-
275.

Greenwald, A. G., & Farnham, S. D. (2000). Using the Implicit Association Test to
measure self-esteem and self-concept. Journal of Personality and Social
Psychology, 79, 1022-1038.

Hall, R. J ., Snell, A. F ., & Foust, M. S. (1999). Item parceling strategies in SEM:
Investigating the subtle effects of unmodeled secondary constructs.
Organizational Research Methods, 2, 233-256.

Haynes, T. L., Ruthig, J. C., Perry, R. P., Stupinsky, R. H., & Hall, N. C. (2006).
Reducing the academic risks of over-optimism: The longitudinal effects of

attributional retraining on cognition and achievement. Research in Higher
Education, 4 7, 755-779.

Helland, P. A., Stallings, H. J ., & Braxton, J. M. (2001-2002). The fulﬁllment of
expectations for college and student departure decisions. Journal of College
Student Retention, 3, 381-396.

Hertzog, C., Lindenberger, U., Ghisletta, P., & von Oertzen, T. (2006). On the power of
multivariate latent grth curve models to detect correlated change.
Psychological Methods, 11, 244-252.

Hollenbeck, J. R., & Klein, H. J. (1987). Goal commitment and the goal-setting process:

Problems, prospects, and proposals for future research. Journal of Applied
Psychology, 72, 212-220.

Hollenbeck, J. R., Klein, H. J ., O’Leary, A. M., & Wright, P. M. (1989). Investigation of

the construct validity of a self-report measure of goal commitment. Journal of
Applied Psychology, 74, 951-956.

133

Hollenbeck, J. R., Williams, C. R., & Klein, H. J. (1989). An empirical examination of

the antecedents of commitment to difﬁcult goals. Journal of Applied Psychology,
74, 18-23.

Hom, P. W., Griffeth, R. W., Palich, L. E., & Bracker, J. S. (1998). An exploratory
investigation into theoretical mechanisms underlying realistic job previews.
Personnel Psychology, 51, 421-451.

Ilgen, D. R., & Seely, W. (1974). Realistic expectations as an aid in reducing voluntary
re‘signations. Journal of Applied Psychology, 59, 452-455.

Irving, P. G., & Meyer, J. P. (1994). Reexamination of met-expectations hypothesis: A
longitudinal analysis. Journal of Applied Psychology, 79, 937-949.

Irving, P. G., & Meyer, J. P. (1995). On using direct measures of met expectations: A
methodological note. Journal of Management, 21, 1 159-1 175.

Jones, G. R. (1986). Socialization tactics, self-efﬁcacy, and newcomers’ adjustments to
organizations. Academy of Management Journal, 29, 262-279.

Joreskog, K., & Sbrbom, D. (1993). LISREL 8 user’s reference guide. Chicago: Scientiﬁc
Software.

Judge, T. A., & Bretz, R. D. (1993). Report on an alternative measure of affective
disposition. Educational and Psychological Measurement, 53, 1095-1104.

Judge, T. A., & Bretz, R. D. (1992). Effects of work values on job choice decisions.
Journal of Applied Psychology, 77, 261-271.

Judge, T. A., & Cable, D. M. (1997). Applicant personality, organizational culture, and
organization attraction. Personnel Psychology, 50, 359-394.

Judge, T. A., & Kristof-Brown, A. (2004). Personality, interactional psychology, and
person-organization ﬁt. In B. Schnedier & D. B. Smith (Eds.), Personality and
organizations (pp. 87-109). Mahwah, NJ: Lawrence Erlbaum Associates.

Kaufman, M. A., & Creamer, D. G. (1991). Inﬂuences of student goals for college on
freshman-year quality of effort and growth. Journal of College Student
Development, 32, 197-206.

Kenny, D. A. (2006). Mediation. Retrieved October 20, 2006 from
//davidakenney.net/cm/mediate.htm

Keman, M. C., & Lord, R. G. (1988). Effects of participative vs. assigned goals and
feedback in a multitrial task. Motivation and Emotion, 12, 75-86.

King, H., & Walsh, W. B. (1972). Change in environmental expectations and perceptions.
Journal of College Student Personnel, 13, 331-337.

134

Klein, H. J ., Wesson, M. J ., Hollenbeck, J. R., & Alge, B. J. (1999). Goal commitment
and the goal-setting process: Conceptual clariﬁcation and empirical synthesis.
Journal of Applied Psychology, 84, 885-896.

Klein, H. J., Wesson, M. J ., Hollenbeck, J. R., Wright, P. M., & DeShon, R. P. (2001).
The assessment of goal commitment: A measurement model meta-analysis.
Organizational Behavior and Human Decision Processes, 85, 32-55.

Klein, H., & Wright, P. M. (1994). Antecedents of goal commitment: An empirical
examination of personal and situational factors. Journal of Applied Social
Psychology, 24, 95-114.

Kline, R. B. (1998). Principles and practice of structural equation modeling. New York:
The Guilford Press.

Kohn, M. L., & Schooler, C. (1978). The reciprocal effects of the substantive complexity
of work and intellectual ﬂexibility: A longitudinal assessment. American Journal
of Sociology, 84, 24-52.

Koole, S. L., Dijksterhuis, A. P., & Van Knippenberg, A. D. (2001). What’s in a name:
Implicit self-esteem and the automatic self. Journal of Personality and Social
Psychology, 80, 669-685.

Kristof, A. L. (1996). Person-organization ﬁt: An integrative review of its

conceptualizations, measurement, and implications. Personnel Psychology, 49, 1-
49.

Kristof-Brown, A. L., & Stevens, C. K. (2001). Goal congruence in project teams: Does
the ﬁt between members’ personal mastery and performance goals matter?
Journal of Applied Psychology, 86, 1083-1095.

Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C. (2005). Consequences of
individuals’ ﬁt at work: A meta-analysis of person-job, person-organization,
person-group, and person-supervisor ﬁt. Personnel Psychology, 58, 281-342.

Latane, B., & Darley, J. M. (1970). The unresponsive bystander: Why doesn ’t he help?
New York: Appleton-Century-Crofts.

Latham, G. P., & Locke, E. A. (1991). Self-regulation through goal setting.
Organizational Behavior and Human Decision Processes, 50, 212-247.

Lauterbach, C. G., & Vielhaber, D. P. (1966). Need-press and expectation-press indices
as predictors of college achievement. Educational and Psychological
Measurement, 26, 965-972.

Lauver, K. J ., & Kristof-Brown, A. (2001). Distinguishing between employees’

perceptions of person-job and person-organization ﬁt. Journal of Vocational
Behavior, 59, 454-470.

135

Lewin, K. (1951). Field theory in social science. New York: Harper Torch Books.

Locke, E. A. (1968). Toward a theory of task motivation and incentives. Organizational
Behavior & Human Performance, 3, 157-189.

Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance.
Englewood Cliffs, NJ: Prentice Hall.

Locke, E. A., Latham, G. P., & Erez, M. (1988). The determinants of goal commitment.
Academy of Management Review, 13, 23-39.

Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. (1981). Goal setting and task
performance: 1969-1980. Psychological Bulletin, 90, 125-152.

Louis, M. R. (1980). Surprise and sense making: What newcomers experience in entering
unfamiliar organizational settings. Administrative Science Quarterly, 25, 226-251.

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002).
A comparison of methods to test mediation and other intervening variable effects.
Psychological Methods, 7, 83-104.

Maier, G. W., & Brunstein, J. C. (2001). The role of personal work goals in newcomers’
job satisfaction and organizational commitment: A longitudinal analysis. Journal
of Applied Psychology, 86, 1034-1042.

Major, D. A., Kozlowski, S. W. J ., Chao, G. T., & Gardner, P. D. (1995). A longitudinal
investigation of newcomer expectations, early socialization outcomes, and the

moderating effects of role development factors. Journal of Applied Psychology,
80, 418-431.

Mallinckrodt, B., & Sedlacek, W. E. (1987). Student retention and the use of campus
facilities by race. NASPA Journal, 24, 28-32.

March, J. G. & Simon, H. A. (1958). Organizations. Oxford, England: John Wiley.

Meglino, B. M., & Ravlin, E. C. (1998). Individual values in organizations: Concepts,
controversies, and research. Journal of Management, 24, 351-389.

Meglino, B. M., Ravlin, E. C., & Adkins, C. L. (1989). A work values approach to
corporate culture: A ﬁeld test of the value congruence process and its relationship
to individual outcomes. Journal of Applied Psychology, 74, 424-432.

Meyerson, D. E., & Scully, M. A. (1995). Tempered radicalism and the politics of
ambivalence and change. Organization Science, 6, 585-600.

Millsap, R. E., & Kwok, O. (2004). Evaluating the impact of partial factorial invariance
on selection in two populations. Psychological Methods, 9, 93-115.

136

Miner, A. (1987). Idiosyncratic jobs in formal organizations. Administrative Science
Quarterly, 32, 327-351.

Mischel, W. (1977). On the future of personality measurement. American Psychologist,
32, 246-254.

Mobley, W. H. (1977). Intermediate linkages in the relationship between job satisfaction
and employee turnover. Journal of Applied Psychology, 62, 237-240.

Moore, N. B., & Davidson, J. K., Sr. (2006). College women and personal goals:
Cognitive dimensions that differentiate risk-reduction sexual decisions. Journal of
Youth and Adolescence, 35, 574-586.

Moos, R. H., & Swindle, R. W., Jr. (1990). Person-environment transactions and the
stressor-appraisal-coping process. Psychological Inquiry, 1, 30-32.

Motowidlo, S. J. (2003). Job performance. In W. C. Borman, D. R. Ilgen, & R. J.
Klimoski, (Eds.) Handbook of psychology: Industrial and organizational
psychology, (V 01. 12, pp. 39-53) Hoboken, NJ: John Wiley & Sons, Inc.

Naylor, J. C., & Ilgen, D. R. (1984). Goal setting: A theoretical analysis of a motivational
technology. Research in Organizational Behavior, 6, 95-140.

Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of
social judgment. Englewood Cliffs, NJ: Prentice-Hall, Inc.

Operario, D., & Fiske, S. T. (2001). Ethnic identity moderates perceptions of prejudice:
Judgments of personal versus group discrimination and subtle versus blatant bias.
Personality and Social Psychology Bulletin, 27, 550-561.

O’Reilly, C. A., & Chatman, J. (1986). Organizational commitment and psychological
attachment: The effects of compliance, identiﬁcation, and internalization on
prosocial behavior. Journal of Applied Psychology, 71, 492-499.

O’Reilly, C. A., Chatman, J ., & Caldwell, D. F. (1991). People and organizational
culture: A proﬁle comparison approach to assessing person-organization ﬁt.
Academy of Management Journal, 34, 487-516.

Orr, J. M., Sackett, P. R., & Mercer, M. (1989). The role of prescribed and nonprescribed
behaviors in estimating the dollar value of performance. Journal of Applied
Psychology, 74, 34-40.

Pascarella, E. T., Smart, J. C., & Ethington, C. A. (1986). Long-term persistence of two-
year college students. Research in Higher Education, 24, 47-71.

Pascarella, E. T., & Terenzini, P. T. (1983). Predicting voluntary freshman year
persistence/withdrawal behavior in a residential university: A path analytic
validation of Tinto’s model. Journal of Educational Psychology, 75 , 215-226.

137

Podsakoff, P. M., MacKenzie, S. B., Lee, J-Y., & Podsakoff, N. P. (2003). Common
method biases in behavioral research: A critical review of the literature and
recommended remedies. Journal of Applied Psychology, 88, 879-903.

Porter, L. W., & Steers, R. M. (1973). Organizational, work, and person factors in

employee factors in employee turnover and absenteeism. Psychological Bulletin,
80, 151-176.

Premack, S. L., & Wanous, J. P. (1985). A meta-analysis of realistic job preview
experiments. Journal of Applied Psychology, 70, 706-719.

Reise, S. P., Widaman, K. F ., & Pugh, R. H. (1993). Conﬁrmatory factor analysis and
item response theory: Two approaches for exploring measurement invariance.
Psychological Bulletin, 114, 552-566.

Renn, R. W., Danehower, C., Swiercz, P. M, & Icenogle, M. L. (1999). Further
examination of the measurement properties of Liefer & McGannon’s (1986) goal
acceptance and goal commitment scales. Journal of Occupational and
Organizational Psychology, 72, 107-113.

Riedel, J. A., Nebeker, D. M., & Cooper, B. L. (1988). The inﬂuence of monetary
incentives on goal choice, goal commitment, and task performance.
Organizational Behavior and Human Decision Processes, 42, 155-180.

Roberson, L. (1990). Prediction of job satisfaction from characteristics of personal work
goals. Journal of Organizational Behavior, 11, 29-41.

Roberts, K.H., Hulin, C. L., & Rousseau, D. M. (1978). Toward an Interdisciplinary
Science of Organizations. San Francisco: Jossey-Bass.

Roth, P. L., Bobko, P., & Mabon, H. (2002). Utility analysis: A review and analysis at the
turn of the century. In N. Anderson, D. S. Ones, H. K. Sinangil, & C.
Viswesvaran, (Eds.) Handbook of industrial, work, and organizational
psychology, (Vol. 1: Personnel Psychology; pp. 383-328). Thousand Oaks, CA:
Sage Publications Ltd.

Ryan, A. M., & Schmit, M. J. (1996). An assessment of organizational climate and P-E
ﬁt: A tool for organizational change. The International Journal of Organizational
Analysis, 4, 75-95.

Rynes, S. L., & Gerhart, B. (1990). Interviewer assessments of applicant ﬁt: An
exploratory investigation. Personnel Psychology, 43, 13-35.

Salanick, G. R., & Pfeffer, J. (1977). An examination of need-satisfaction models of job
attitudes. Administrative Science Quarterly, 22, 427-456.

Schmidt, F. L., Ones, D. S., & Hunter, J. E. (1992). Personnel selection. Annual Review
of Psychology, 43, 627-670.

138

Schmitt, N. (1982). The use of analysis of covariance structures to assess beta and
gamma change. Multivariate Behavioral Research, I 7, 343-358.

Schmitt, N., Oswald, F. L., Friede, A. J ., Imus, A. L., & Merritt, S. M. (under review).
Perceived ﬁt with an academic environment: Attitudinal and behavioral outcomes.

Schneider, B. (1987). The people make the place. Personnel Psychology, 40, 437-453.

Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being:
Informative and directive functions of affective states. Journal of Personality and
Social Psychology, 45, 513-523.

Shaw, K. A. (1968). Accuracy of expectation of a university’s environment as it relates to
achievement, attrition, and change of degree objective. Journal of College Student
Personnel, 9, 44-48.

Sims, R. L., & Kroeck, K. G. (1994). The inﬂuence of ethical ﬁt on employee
satisfaction, commitment, and turnover. Journal of Business Ethics, 13, 939-947.

Sobel, M. E. (1982). Asymptotic conﬁdence intervals for indirect effects in structural
equation models. In S. Leinhardt (Ed.), Sociological methodology 1982 (pp. 290-
312). Washington, DC: American Sociological Association.

Sorbom, D. (1974). A general method for studying differences in factor means and factor
structures between groups. British Journal of Mathematical and Statistical
Psychology, 27, 229-239.

Steiger, J .H. (1990). Structural model evaluation and modiﬁcation: An interval
estimation approach. Multivariate Behavioral Research, 25 , 173-180.

Stewart, G. L., & Barrick, M. R. (2004). Four lessons learned from the person-situation
debate: A review and research agenda. In B. Schneider & D. B. Smith, (Eds.),
Personality and Organizations. Mahwah, NJ: Erlbaum, pp. 61-86.

Swindle, R. W., Jr., & Moos, R. H. (1992). Life domains in stressors, coping, and
adjustment. In B. W. Walsh, K. H. Craik, & R. H. Price (Eds.), Person-
environment psychology: Models and perspectives (pp. 1-33). Hillsdale, NJ:
Lawrence Erlbaum Associates.

Terborg, J. R. (1981). Interactional psychology and research on human behavior in
organizations. The Academy of Management Review, 6, 569-576.

Tinto, V. (1975). Dropout from higher education: A theoretical synthesis of recent
research. Review of Educational Research, 45 , 89-125.

Tubbs, M. E. (1993). Commitment as a moderator of the goaIAperformance relation: A
case for clearer construct deﬁnition. Journal of Applied Psychology, 78, 86-97.

139

Tubbs, M. E., & Dahl, J. G. (1991). An empirical comparison of self-report and
discrepancy measures of goal commitment. Journal of Applied Psychology, 76,
708-716.

Tubbs, M. E.,& Ekeberg, S. E. (1991). The role of intentions in work motivation:
Implications for goal-setting theory and research. Academy of Management
Review, 16, 180-199.

Turban, D. B., & Keon, T. L. (1993). Organizational attractiveness: An interactionist
perspective. Journal of Applied Psychology, 78, 184-193.

Vancouver, J. B., & Schmitt, N. W. (1991). An exploratory examination of person-

organization ﬁt: Organizational goal congruence. Personnel Psychology, 44, 333-
352.

Vandenberg, R. J ., & Lance, C. E. (2000). A review and synthesis of the measurement
invariance literature: Suggestions, practices, and recommendations for
organizational research. Organizational Research Methods, 3, 4-69.

Vandenberg, R. J ., & Self, R. M. (1993). Assessing newcomer’s changing commitments
to the organization during the ﬁrst six months of work. Journal of Applied
Psychology, 78, 557-568.

Van Maanen, J., & Schein, E. H. (1979). Toward a theory of organizational socialization.
In B. M. Staw (Ed.), Research in organizational behavior (Vol 1., pp. 209-264).
Greenwich, CT: JAI Press.

Verquer, M. L., Beehr, T. A., & Wagner, S. H. (2003). A meta-analysis of relations
between person-organiztion ﬁt and work attitudes. Journal of Vocational
Behavior, 63, 473-489.

Walther, E., Nagengast, B., & Traselli, C. (2005). Evaluative conditioning in social
psychology: Facts and speculations. Cognition & Emotion, 19, 175-196.

Wanous, J. P. (1977). Organizational entry: Newcomers moving from outside to inside.
Psychological Bulletin, 84, 601-618.

Wanous, J. P., Poland, T. D., Premack, S. L., & Davis, K. S. (1992). The effects of met
expectations on newcomer attitudes and behaviors: A review and meta-analysis.
Journal of Applied Psychology, 77, 282-297.

Wanous, J. P., & Colella, A. (1989). Organizational entry research: Current status and
future directions. In G. Ferris & K. Rowland (Eds.), Research in personnel and
human resource management (V o]. 7, pp. 59-120). Greenwich, CT: JAI Press.

Westerman, J. W., & Cyr, L. A. (2004). An integrative analysis of person-organization ﬁt
theories. International Journal of Selection and Assessment, 12, 252-261.

140

White, R. W. (1952). Lives in progress. New York: Dryden.

Willett, J. B., & Sayer, A. G. (1994). Using covariance structure analysis to detect
correlates and predictors of individual change over time. Psychological Bulletin,
116, 363-381.

Wright, P. M., O’Leary-Kelly, A. M., Cortina, J. M., Klein, H. J ., & Hollenbeck, J. R.
(1994). On the meaning and measurement of goal commitment. Journal of
Applied Psychology, 79, 795-803.

141

llllllllllllllWWIIl

063 4038

3
o
3

 

”2