 

 

 

 

 

 

(9%,. .54. m mar.
ﬁmwmmwkx 1:

L.
h

.5
.aa, ‘ ;. ..,..
rem.” .1.
h... u

, ,
.ﬁrﬁ a... ..

£3

a,
ravine. _
«an.

5

2.

4r
ﬁllglil.‘l
, .... I... .
ﬁxmunuxumwd
.. ....
43 £33,! an. 2.
Hawaii! .n ..
.. . .ﬁ. :tzamnn
A .3ﬁﬂ .‘l
{1. . ‘
.3 . ‘
:....... gain
I Luna." numb] «masomneWMMR
tn. 1 . .
f. .13». §% r .
A .. . . . , .
-.m.:.mu.m.fé
Irraﬁa
.32...-
. at}...
upsﬂve

,. . nut}. :
1.. 1....th
. .

“can! #m

t:
‘19: 7.510

 

(Tch
MlCl’ub . Jlate
Universnty

 

 

 

This is to certify that the

dissertation entitled

THE INFLUENCE OF PERSUASIVE MESSAGES
0N ATTITUDE AND SUBJECTIVE NORM:
A TEST OF THE THEORY OF REASONED ACTION

presented by
Lisa L. Massi Lindsey

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degreein 0mm nicat'i n

 

Major professor

Date 2 July 2003

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

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

DATE DUE DATE DUE DATE DUE
i ' r3 (1

lftz

JUN 0 5 2010

050510

 

6/01 c‘JClRC/DIIODuth5-9JS

THE INFLUENCE OF PERSUASIVE MESSAGES ON ATTITUDE AND
SUBJ ECT IVE NORM: A TEST OF THE THEORY OF REASONED ACTION

By

Lisa L. Massi Lindsey

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Communication

2003

ABSTRACT

THE INFLUENCE OF PERSUASIVE MESSAGES ON ATTITUDE AND
SUBJECTIVE NORM: A TEST OF THE THEORY OF REASONED ACTION

By

Lisa L. Massi Lindsey

The Theory of Reasoned Action (TRA) predicts that a person’s attitude
and subjective norm both impact behavioral intent which subsequently predicts
behavior (Ajzen & Fishbein, 1970, 1974, 1980; Fishbein & Ajzen, 1975). One
limitation to extant TRA research, however, is that it has not examined the impact
of persuasive messages on the TRA’s predictions. Given that attitudes and
subjective norms are subject to change, and given that people are exposed
frequently to messages designed to change their attitudes and subjective norms,
the current study examined the predicted relationships outlined by the TRA within
two systems of change: 1) when attitudes and 2) subjective norms were inﬂuenced
by a persuasive message. These predictions were tested by examining 276
undergraduate students’ eating behaviors (i.e., hit and vegetable consumption
and the number of meals skipped per day). Participants were assigned randomly
to one of three conditions: 1) a no-message control group, 2) an attitude message
group, and 3) a subjective norm message group. The results related to fruit and
vegetable consumption indicated that when respondents’ attitude or subjective
norm was inﬂuenced by a persuasive message the TRA’s predictions were not
consistent with the data. The test of the TRA when no change was present,

however, provided evidence that its predictions were consistent with the control

group data. These results were not replicated, however, with the second topic
(skipping meals). Speciﬁcally, the TRA’s predictions were not consistent with the
data ﬁom any group. Interestingly, the only model to predict accurately the
number of meals participants skipped every day was a simple causal string such
that attitudes impacted intentions which predicted behavior. The implications of
these results, limitations of the current investigation, and directions for future

research are discussed in detail.

ACKNOWLEDGMENTS

I have been very fortunate to have a large group of ﬁ'iends and family who have
supported me throughout my graduate career. I thank all of you and would like to
acknowledge individually those who have showed exceptional support:

To my dissertation committee, Frank Boster, Tim Levine, and Stan Kaplowitz,
thank you for your invaluable feedback and guidance throughout my program of study.
Frank, I am also eternally grateful for the opportunity to work as your editorial assistant
on Communication Monographs. To Sandi Smith, my irreplaceable advisor, thank you for
your dedication and guidance during my MSU career. I enjoyed working with all of you.

To my family, the Massi, Mangino, and Colley/Lindsey clans, I love you all and
thank you for your never-ending support.

To my friends at MSU, especially Dave, Jonathan, Kelli, Ken, Mike, Paul, Pete,
and Ryan, It’s been fun. Kelli, you are my last hope. Find out the truth about Dave and
report back to me.

To Rachel and J -Ching, thanks for the much needed distractions, the laughs, and
your incessant support.

Alas, Kimo, I continue in your footsteps. Thank you for your love and friendship.

To my brother, Michael, you have given me the courage, strength, and
determination to attain my goals. I love you and I miss you.

Most importantly, to my husband and life partner, Gary, your endless
encouragement, love, support, and unwavering belief in me gives me the strength to

accomplish anything. I love you with all of my heart. I am ﬁnally coming home.

iv

TABLE OF CONTENTS

LIST OF TABLES ..................................................... vii
LIST OF FIGURES .................................................... viii
INTRODUCTION AND RATIONALE ...................................... 1
REVIEW OF LITERATURE .............................................. 3
Theory of Reasoned Action .......................................... 3
Attitudes ................................................... 3

Subjective Norm ............................................ 4
Understanding the Theory of Reasoned Action ..................... 8

Examining the Theory of Reasoned Action When Change Occurs ...... 9

PILOT STUDY ........................................................ 12
Method ......................................................... 12
Participants ................................................ 12

Design ................................................... 12

Procedure ................................................. 13

Results ......................................................... l6
Instrumentation ............................................ 16

Message Ratings ............................. . .............. l6

Attitude Message Position .................................... l7

Subjective Norm Message Position ............................. 18

Summary of Pilot Results .................................... 20

MAIN EXPERIMENT ................................................... 21
Method ......................................................... 21
Participants ................................................ 21

Design ................................................... 22

Procedure ................................................. 22
Instrumentation ............................................ 25

Overview of Analyses ....................................... 29

Results ......................................................... 30
Instrumentation ............................................ 30

Evaluation of the Models ..................................... 32

Fruit and Vegetable Consumption Results ....................... 32

Skipped Meals Results ....................................... 41
DISCUSSION ................................................... 51
Findings and Implications .......................................... 51
Limitations ...................................................... 53
Directions for Future Research ...................................... 58

Conclusion ...................................................... 62

APPENDICES ......................................................... 64
Appendix A. Pilot Behavioral Questions ............................... 65
Appendix B. Message Realism Scale ................................. 66
Appendix C. Message Credibility Scale ............................... 67
Appendix D. Message Position Scale (Fruit and Vegetables Topic) .......... 68
Appendix E. Message Position Scale (Skipping Meals Topic) .............. 69
Appendix F. Perceptions of Advocated Familial Subjective Norm

(Fruit and Vegetables Topic) .................................. 70
Appendix G. Perceptions of Advocated Familial Subjective Norm

(Skipping Meals Topic) ...................................... 71
Appendix H. Perceptions of Advocated Close Friends Subjective Norm

(Fruit and Vegetable Topic) .................................. 72
Appendix I. Perceptions of Advocated Close Friends Subjective Norm

(Skipping Meals Topic) ...................................... 73
Appendix J. Attitude Scale (Fruit and Vegetables Topic) .................. 74
Appendix K. Attitude Scale (Skipping Meals Topic) ..................... 75
Appendix L. Perceptions of Familial Subjective Norm

(Fruit and Vegetables Topic) .................................. 76
Appendix M. Perceptions of Familial Subjective Norm

(Skipping Meals Topic) ...................................... 77
Appendix N. Perceptions of Close Friends Subjective Norm

(Fruit and Vegetables Topic) .................................. 78
Appendix 0. Perceptions of Close Friends Subjective Norm '

(Skipping Meals Topic) ...................................... 79
Appendix P. Attitude Message ...................................... 90
Appendix Q. Subjective Norm Message ............................... 81
Appendix R. Time One Survey ...................................... 82
Appendix S. Time Two Survey ...................................... 97
Appendix T. Fruit and Vegetables Belief Scale ......................... 109
Appendix U. Importance Scale ..................................... 110
Appendix V. Skipping Meals Belief Scale ............................ 111
Appendix W. Family Normative Beliefs (Fruit and Vegetables Topic) ...... l 12
Appendix X. Motivation to Comply with Family ....................... 113
Appendix Y. Family Normative Beliefs (Skipping Meals Topic) ........... 114
Appendix Z. Close Friends’ Normative Beliefs

(Fruit and Vegetables Topic) ................................. 1 15
Appendix AA. Motivation to Comply with Close Friends ................ 116
Appendix BB. Close Friends’ Normative Beliefs (Skipping Meals Topic) . . . . 117
Appendix CC. Behavioral Intent Scale (Fruit and Vegetables Topic) ........ 118
Appendix DD. Behavioral Intent Scale (Skipping Meals Topic) ........... 119

REFERENCES ....................................................... 120

vi

LIST OF TABLES

Table l .
Means and Standard Deviations (in parentheses) for Fruit and Vegetable

Consumption by Experimental Condition ...........................

Table 2
Means and Standard Deviations (in parentheses) for Skipping Meals Topic

by Experimental Condition ......................................

Table 3
Zero-order Correlations Used to Calculate Parameter Estimates in Model 1

Table 4
Zero-order Correlations Used to Calculate Parameter Estimates in Model 2

Table 5
Zero-order Correlations Used to Calculate Parameter Estimates in Model 3

Table 6
Zero-order Correlations Used to Calculate Parameter Estimates in Model 5

Table 7 <
Zero-order Correlations Used to Calculate Parameter Estimates in Model 7

Table 8
Zero-order Correlations Used to Calculate Parameter Estimates in Model 9

Table 9

Hypothetical TRA Data .........................................

vii

......... 31

......... 31

......... 33

......... 34

......... 39

......... 42

......... 45

......... 49

......... 60

LIST OF FIGURES

Figure 1 .
Model depicting the TRA when a persuasive message is aimed at attitudes
with path coefﬁcients corrected for attenuation due to measurement error. .......... 33

Figure 2
Revised model with path coefﬁcients corrected for attenuation due to
measurement error. . .' .................................................... 35

Figure 3
Model depicting the TRA when a persuasive message is aimed at subjective
norm with path coefﬁcients corrected for attenuation due to measurement error. ..... 37

Figure 4
Model depicting the TRA when no persuasive message is present with path
coefﬁcients corrected for attenuation due to measurement error. .................. 39

Figure 5
Model depicting the TRA when a persuasive message is aimed at attitudes
with path coefﬁcients corrected for attenuation due to measurement error. .......... 42

Figure 6
Revised model with path coefﬁcients corrected for attenuation due to
measurement error ....................................................... 43

Figure 7
Model depicting the TRA when a persuasive message is aimed at subjective
norm with path coefﬁcients corrected for attenuation due to measurement error. ..... 45

Figure 8
Revised model with path coefﬁcients corrected for attenuation due to
measurement error ....................................................... 46

Figure 9
Model depicting the TRA when no persuasive message is present with path
coefﬁcients corrected for attenuation due to measurement error. .................. 48

Figure 10

Revised model with path coefﬁcients corrected for attenuation due to
measurement error ....................................................... 49

viii

Figure 11a

Condition under which the TRA cannot assess directionality. .................... 60
Figure 11b
The reverse TRA. ................................ . ....................... 6O

ix

INTRODUCTION AND RATIONALE

The nature and direction of the relationship between attitudes and behavior has
been the subject of debate among scholars from varying ﬁelds of inquiry. In brief,
scholars have taken four differing positions with regards to the casual priority of attitudes
and behaviors: attitudes are predictive of behaviors (McGuire, 1976), behaviors cause
attitudes (Bern, 1972), attitudes and behaviors impact each other (Kelmen, 1974), and
attitudes and behaviors are related weakly (Wicker, 1969). Recent meta-analytic studies
have noted that a strong attitude-behavior relationship indeed exists (r = .79, Kim &
Hunter, 1993 a), and that behavioral intent mediates that relationship (Kim & Hunter,
1993b). A theory commonly employed to explain this relationship is the Theory of
Reasoned Action (TRA).

The TRA elucidates the attitude-behavior relationship such that a person’s attitude
(deﬁned in the TRA as the surnmative evaluation of a person’s beliefs) and subjective
norm (the sum of normative beliefs and a person’s motivation to comply with them) both
impact behavioral intent which subsequently predicts behavior (e.g., Ajzen & F ishbein,
1970, 1974, 1980). Moreover, the accuracy of the TRA predictions is strongly inﬂuenced
by the degree to which the attitude being assessed conceptually matches the behavior
being predicted. For example, meta-analytic results indicate that the average correlation
between attitudes and behavior is stronger when the attitude being assessed is relevant
(i.e., a conceptual match) to the behavior under scrutiny (mean r = .86, Kim & Hunter,
1993a).

One limitation to the body of TRA research, however, is that it has ignored the

inﬂuence persuasive messages may have on the attitude-behavior relationship.

Speciﬁcally, extant research has treated components of the TRA as static and has not
examined the effect of changes in attitude or subjective norm on the model’s predictions.
For example, Nucifora, Gallois, and Kashima (1993) examined the degree to which
condom use was predicted accurately by intentions to use condoms, and intentions were
predicted accurately by attitudes and subjective norm, as an application of the TRA to
AIDS-preventative behavior, but did not examine the effect of safe-sex or AIDS-
preventative messages on these predictions. To this end, the current study proposes a test
of the TRA that will focus on the impact of persuasive messages on the TRA’s
predictions.

It is important to provide such a test of the TRA for two reasons. First, attitudes
and subjective norms rarely are held constant; therefore, any static test of the TRA may
not elucidate fully the process by which attitudes and subjective norms affect behavior.
Second, people are presented continually with persuasive messages aimed to change their
attitudes, subjective norms, and behavior. Thus, it is important to understand how such
messages will affect the predictive power of the TRA. In order to better understand the
relationships proposed here, the TRA will be discussed in greater detail and a model will

be proposed.

REVIEW OF LITERATURE
The Theory of Reasoned Action

As noted previously, the TRA predicts that both a person’s attitude and subjective
norm inﬂuence the degree to which a person will intend to engage in a speciﬁc behavior
and that this intent subsequently predicts volitional behavior (e. g., Ajzen & Fishbein,
1970, 1974, 1980). In order to better understand these relationships, the attitude and
subjective norm components of the TRA must be addressed individually.

Attitudes

Although there is some debate regarding the origin and structure of attitudes (see
Hunter, Levine, & Sayers, 1976; Schwarz & Bohner, 2001) the TRA posits that an
attitude toward any given behavior is a function of how people evaluate their beliefs
about the behavior and the strength with which they hold those beliefs (F ishbein &
Ajzen, 1975; Hale, Householder, & Greene, 2002). Mathematically, one’s attitude toward
a behavior (A3) is equal to the sum of the evaluations (ei) and assessments of strength (5,)
one makes for all relevant beliefs about a focal behavior, and can be expressed as A3 =
Zeisi (F ishbein, 1967a, 1967b).

The beliefs on which an attitude is based are cognitions that link a given attribute
with a behavior (Hale et al., 2002). For example, the cognition “Eating a balanced diet is
healthy” is a belief that links an attribute (health) with a behavior (a decision to eat
regular, balanced meals). Belief evaluation is the degree to which a person judges the
attribute (health) as positive or negative, and belief strength refers to the degree of
certainty with which a person holds the belief. Therefore, if a person evaluates the belief

“Eating a balanced diet is healthy” as strongly positive (.90) and is certain that eating a

balanced diet will ensure good health (.80), the resulting attitude will be strongly
favorable toward healthy eating behaviors (.72). It should be noted here that this example
represents only the evaluation and strength of one belief about a focal behavior, where
people have a series of beliefs that are summed to form an attitude.

The resulting attitudes, according to the TRA, predict behavioral intention insofar
that people intend to act in ways consistent with their attitudes (Fishbein & Ajzen, 1975).
Consistent with this, Kim and Hunter (1993b) found that attitudes and intent were
strongly correlated (corrected mean r = .87, K = 92). Attitudes, however, are not the only
predictor of behavioral intentions according to the TRA, and it is important to understand
the predictive role of the subjective norm.

Subjective Norm

The TRA proposes that the subjective norm is a function of normative beliefs
about focal behaviors (how people perceive the expectations of others who are important
to them with regard to how they should act) and motivation to comply with the normative
belief (the perceived pressure people feel to act in accordance with others’ expectations).
Mathematically, one’s subjective norm (S) is equal to the sum of the normative beliefs
(m) and motivation to comply (mi), and can be expressed as S = 2mm, (Fishbein, 1967a,
1967b).

With regards to the eating behavior example used previously, one might have a
normative belief (e. g., “My parents think that eating a balanced diet is a good idea”) and
more or less motivation to comply with the normative belief (e. g., “As far as controlling
my eating habits, I don’t want to do what my parents say”). Therefore, if one strongly

perceives the normative belief “My parents think that eating a balanced diet is a good

idea” (.75) but has little motivation (.20) to act in accordance with the normative belief
(e. g., “I do not want to do what my parents say”), one’s subjective norm will not be in
favor of healthy eating (.15). Conversely, if one strongly and negatively perceives the
normative belief “My friends think that eating a balanced diet is necessary to lose
weight” (-.80) and perceives a great deal of pressure, either real or imagined, to act in
accordance with that normative belief (e.g., “I want to do what my friends say is a good
thing to do when it comes to weight management”) the motivation to comply will be high
(.90) and the subjective norm will be strongly out of favor with healthy eating behavior (-
.72). As noted previously with regard to attitudes, this example examines one’s
motivation to comply with only one normative belief where subjective norm is a sum of
one’s motivation to comply with any number of normative beliefs held by different
important others.

7 In order to understand the degree to which one’s normative beliefs about healthy
eating behaviors are based on our close relationships, an examination of the literature
focused on both familial and peer relationships as predictors of eating behaviors will be

discussed.

Familial predictors of eating behaviors. A great deal of research has sought to
determine the degree to which one’s familial relationships predict both healthy and
unhealthy eating habits. This line of research has found a number of factors which
contribute to such eating behaviors. Strong predictors of one’s eating habits are the
degree to which people perceive both parental pressure to diet and parental dissatisfaction
with one’s physical appearance (Haworth-Hoeppner, 2000; Moreno & Thelen, 1993;

Schur, Sanders, & Steiner, 2000; Striegel-Moore & Keamey-Cooke, 1994; Thelen &

Cormier, 1995). For example, Moreno and Thelen (1993) found that people who perceive
a great deal of parental pressure to lose weight are more likely to suffer ﬁom eating
disorders than those whose parents do not pressure them (see also Twamley & Davis,
1999, who found that familial pressure moderated the relationship between exposure to
thinness norms and eating pathology). Also, people with healthy eating behaviors
perceive their parents to be more caring and warm, less overprotective (Calarn, Waller,
Slade, & Newton, 1990), and more helping, trusting, and nurturing (Humphrey, 1987)
than people with eating disorders. Other predictors include defense styles and parental
bonding (Humphrey, 1986; Steiger, Van der F een, Goldstein, & Leichner, 1989),
attachment and separation difﬁculties (Armstrong & Roth, 1989; Humphrey, 1989;
O’Keamey, 1996), weight-related teasing and criticism by family (HaworthoHoeppner,
2000; Humphrey, 1987; Levine, Smolak, & Hayden, 1994), parental supervision (Young

& Fors, 2001), and family dysfunction (Scalf-Mclver & Thompson, 1989).

These predictors necessarily will affect one’s subjective norm. For example, if
one perceives that his or her parents value eating a healthy, balanced diet and is
motivated to comply with the parents’ pressure for healthy eating, one should develop a
subjective norm that favors eating behaviors that will result in following a balanced diet.
Conversely, one who is motivated to comply with parents who value idealized thinness
are likely to engage in unhealthy eating behaviors to attain the desired thinness (e. g.,
eliminating all fat from the diet, reducing caloric intake to dangerously low levels).

Although a number of familial predictors exist with regard to eating behaviors,
most researchers argue that a number of other inﬂuences must also be examined to

understand healthy eating behaviors and eating disorders (see Sanftner, Crowther,

Crawford, & Watts, 1996). One such alternate predictor is one’s peer relationships. The
inﬂuence of peer relationships on eating behaviors is discussed subsequently.

Peer inﬂuence and eating behaviors. Research has. shown that peer concerns with
weight and body shape serve as important modeling cues for weight management efforts
and that pe0p1e who are more committed to weight management are highly attuned to
similar behavior in their peers (Levine, Smolak, Moodey, et al., 1994). For example,
peers’ weight-loss and eating behaviors have been found to be substantial predictors of
people’s own behaviors (Paxton, Schutz, Wertheim, & Muir, 1999). Furthermore,
Vincent and McCabe (2000) found that direct inﬂuence of peers predicted eating
behaviors, and that the quality of those peer relationships, were less important in
predicting such behaviors. Oliver and Thelen (1996), however, found that likability was a
major predictor of eating concerns. Therefore, although the overall quality of peer
relationships may not be as important as the direct inﬂuence they can have on behaviors,
the more liked the peer, the more likely the peer’s inﬂuence with regard to behavior. For
example, a college student may be strongly inﬂuenced to adopt different eating habits
when subjected to ridicule from other students about her weight. This student likely does
not have a relationship of high quality with these ridiculing students, but she may be
motivated to comply with their normative beliefs about idealized body shapes
nonetheless. The student is more likely, however, to comply with the normative beliefs of
a person she holds in high regard.

It is also important to note that not all peer inﬂuence on eating behaviors is
negative. Whereas most research has focused on the effect that peer pressure has on the

development of eating disorders such as anorexia nervosa and bulimia nervosa,

Hausenblas & Carron (1998) found that college students report more positive peer
inﬂuences on their eating and dieting behaviors (e. g., support for following a balanced
diet) than negative inﬂuences (e. g., recommendations of binging and purging). Therefore,
one’s peer relationships, just as with familial relationships, will likely affect subjective
norms. For example, if one is motivated to comply with beliefs of a peer who values
eating a balanced diet, the resulting subjective norm will affect subsequent intentions and
behaviors to adopt healthy eating patterns.

Other inﬂuences on one’s subjective norm exist (e. g., exposure to idealized body
images in the media: Levine, Smolak, & Hayden, 1994; Levine, Smolak, Moodey, et al.,
1994; Polivy & Herman, 2002; and trait and social anxiety and depression: Evans &
Wertheim, 1998; Graber, Brooks-Gunn, Paikoﬂ', & Warren, 1994) and some people will
weigh their own attitudes more heavily than they will their subjective norm when
contemplating a given behavior. The TRA does brieﬂy address differential weight given
to attitudes and subjective norms, although some scholars believe this issue is addressed
inadequately (see Burgoon, Birk, & Coker, 1998). To better understand how attitudes and
subjective norms inﬂuence intentions and subsequent volitional behavior it is necessary
to now turn to an explanation of the TRA as a whole.

Understanding the Theory of Reasoned Action

The predicted TRA relationships among attitudes, subjective norm, and

behavioral intent have been traditionally understood as the mathematical function:
I = (A)W1 + (S)W2
where intent (I) is a function of one’s attitude toward the behavior (A) and subjective

norm related to the behavior (S) and the weight of each (W. and W2, respectively).

Moreover, behavioral intentions are posited to be the strongest predictor of behavior
(Fishbein & Ajzen, 197 5). Consistent with this, meta-analytic results indicate that
behavioral intentions and behavior are related strongly (corrected mean r = .82, K = 47,
Kim & Hunter, 1993b; see Hale et al., 2002 for a summary of all meta-analytic ﬁndings),
and research across a wide range of domains have provided evidence consistent with the
predictions made by the TRA.

It should be noted, however, that there is a substantial gap in existing TRA
research; namely, the TRA has been tested traditionally as a static model. That is to say,
existing attitudes and subjective norm were simply measured and change was not induced
in order to test systematically the effect of attitude or subjective norm change on
intentions and subsequent behavior. In order to determine the effectiveness of the TRA in
explaining the attitude-behavior relationship, it must also be studied when attitudes and
subjective norms are inﬂuenced by a persuasive message. To this end, change and the
TRA are discussed next.

Examining the Theory of Reasoned Action When Change Occurs

Introducing change can occur in a number of ways. One method communication
scholars can use is to aim persuasive messages at people’s attitudes and subjective norm.
For example, introducing a persuasive message can initiate change by impacting an
attitude by moving it from its current state to a more favorable or unfavorable attitude.
Changes in attitude and subjective norm may thereby result in changes in behavioral
intent and subsequent behavior. The effect of such persuasive messages on the TRA
predictions can be examined systematically by inﬂuencing change in either attitudes or

subjective norm. Each will be addressed subsequently.

Inﬂuencing change through attitudes. The linear discrepancy model (see Boster,
Fryrear, Mongeau, & Hunter, 1892; Boster, Mayer, Hunter, & Hale, 1980; Hunter et al.,
1976) indicates that attitude change as a result of a persuasive message is a ﬁrnction of
the impact of the message and the discrepancy between the original attitude and the
position advocated by the message. According to the linear discrepancy model,
arguments advocate a speciﬁc attitudinal position that can be arranged on the same
continuum as attitudes. Consequently, people will compare their own attitudes to the
attitude advocated in a persuasive message and attitudes should shift in the direction of
the message (Boster & Cruz, 2001). Mathematically, the linear discrepancy model
predicts that

AA = or(M - A0)
where attitude change (AA) is a function of the difference between one’s initial attitude
(A0) and the message (M) multiplied by the impact or persuasability of the message (or;
how much one changed their attitude). Given this, if one is exposed to a persuasive
message, and the message is successful, attitude change should occur (in the direction
outlined by the message).

Inﬂuencing change through subjective norms. Just as one can inﬂuence attitudinal
change with persuasive messages, so might subjective norms change as the result of a
message. Applying the logic of the linear discrepancy model to subjective norm, one
might expect that

AS = WM - So)
where change in subjective norm (AS) is a function of the difference between one’s initial

subjective norm (So) and the message (M) multiplied by the impact of the message ([3).

10

As noted previously, if one is exposed to a successful persuasive message changes in
subjective norm should occur (in the direction outlined by the message).

Given the predictions of the Theory of Reasoned Action and the rationale for how
change can occur as a result of persuasive messages, the current study proposes to
examine the predicted relationships outlined by the TRA within two systems of change.
First, the TRA will be examined when attitude change results from a persuasive message.
Second, the TRA will be examined when subjective norm changes as a result of a
persuasive message. These predictions will be tested by examining college students’
eating behaviors.

In order to assess the effectiveness of the persuasive messages used for the current
study, a pilot study was conducted to examine a number of dimensions including message
realism, message credibility, and advocated message position. The pilot study is

presented subsequently.

ll

PILOT STUDY
Method
Participants

Participants were 137 undergraduate students recruited from introductory
communication courses at a large Midwestern university. Respondents mean age was
21.23 years old (SD = 2.16), 1% were freshmen, 13% were sophomores, 44% were
juniors, and 42% were seniors. Women comprised 74.5% of the sample and 85.4% of the
respondents self-reported their ethnicity as Caucasian, 7.3% were African American,
3.6% were Hispanic American, 1.5% were Asian American, .7% were Native American,
.7% were of Middle Eastern descent, and .7% declined to state their ethnicity.
Additionally, 19.7% of the participants reported that they were currently living in
university dorms, l 1.7% lived in an on-campus apartment, 2.2% lived with their parents,
and 66.4% lived off campus by themselves or with roommates.

Participants were informed that they were being asked to participate in a survey
related to student life and the use of university facilities such as nutrition counseling and
student health clinics. Respondents received research credit for their participation.
Design

The pilot study utilized a single-factor, control group design such that participants
were assigned randomly to one of 3 conditions: 1) a control, no-message condition, 2) an
attitude-message condition in which participants were presented with a persuasive
message designed to result in favorable attitudes toward eating up to nine servings of
fruits and vegetables a day and not skipping meals, and 3) a subjective norm-message

condition in which participants were exposed to a persuasive message designed to result

12

in perceptions of subjective norms in favor of the same eating behaviors. Each message
was pilot tested in order to ensure I) that the position advocated by the messages was
rated highly on the same scale participants would use to measure their attitudes and
subjective norms, and 2) that the message position was substantially higher than the mean
attitude and subjective norm for the control group.

Procedure

Participants were ﬁrst asked to answer a number of background questions
including self-report measures of how many servings of fruit and vegetables they eat
every day and how many meals they skip a day. They were then instructed to read one of
the persuasive messages (attitude or subjective norm) and subsequently completed a
questionnaire. Participants in the control group completed only the questionnaire. All
pilot measures used Likert-type scales ranging from -3 (strongly disagree) to +3 (strongly
agree) unless otherwise noted.

All participants were initially asked to provide information regarding their current
behaviors on a number of open-ended questions (see Appendix A). These items provided
baseline information with regards to the behaviors of interest. Speciﬁcally, respondents
reported that they, on average, eat 1.43 servings of fruit per day (SD = 1.01), 1.69
servings of vegetables per day (SD = 1.14), skip 4.82 meals per week (SD = 3.23), and
skip .81 meals per day (SD = .56).

Respondents who were exposed to one of the persuasive messages (n = 97) were
asked to rate their perceptions of message realism and message credibility. Message

realism was assessed with ﬁve items and had a mean of 1.28 (SD = .96) and standardized

l3

item alpha ($161) was .81 for this scale (see Appendix B). Message credibility was
assessed with four items and had a mean of 1.27 (SD = 1.00, SIa = .89; see Appendix C).

Participants who were exposed to one of the persuasive messages were also asked
to rate their perceptions of the advocated position of the message they read. Speciﬁcally,
respondents who were exposed to the attitude message (n = 39) were asked to rate the
degree to which the message was in favor of 1) eating up to nine servings of mu and
vegetables every day and 2) skipping meals. The message position for fruit and vegetable
consumption was assessed with three items (see Appendix D) and had a mean of 1.72
(SD = 1.41, SIa = .83). Message position for skipping meals was assessed with three
items (see Appendix E) and had a mean of -2.07 (SD = 1.46, Sla = .94).

Approximately half of the respondents who were exposed to the subjective norm
message (n = 28) were asked to rate their perceptions of the family normative beliefs
presented in the message. Perceptions of the advocated family normative beliefs about
ﬁ'uit and vegetable consumption were measured with three items and had a mean of 1.89

(SD = .96, SIor = .92; see Appendix F). Perceptions of family normative beliefs about

skipping meals were measured with three items and had a mean of -2.00 (SD = 1.02, Slat
= .93; see Appendix G).

Approximately half of the respondents who were exposed to the subjective norm
message (n = 30) were asked to rate their perceptions of the close friends’ normative
beliefs presented in the message. Perceptions of the advocated close ﬁ'iends’ normative
beliefs about hit and vegetable consumption were measured with three items and had a

mean of 1.83 (SD = .81, S10: = .90; see Appendix H). Perceptions of close friends’

14

normative beliefs about skipping meals were measured with three items and had a mean
of -1.78 (SD = .92, S10: = .89; see Appendix 1).

Respondents in the control group (n = 40) completed scales to assess 1) their
attitude about eating up to nine servings of his and vegetables a day, 2) their attitude
about skipping meals, 3) their perceptions of their family’s normative beliefs about eating
up to nine servings of fruits and vegetables every day, 4) their perceptions of their
family’s normative beliefs about skipping meals, 5) their perceptions of their close
friends’ normative beliefs about eating up to nine servings of fruits and vegetables every
day, and 6) their perceptions of their close ﬁiends’ normative beliefs about skipping

meals.

Control participants’ fruit and vegetable attitudes were assessed with ﬁve items
and had a mean of .81 (SD = 1.39, SIa = .90; see Appendix J), and meal skipping
attitudes were assessed with ﬁve items and had a mean of -l .69 (SD = 1.16, SIa = .91;
see Appendix K). Perceptions of families’ normative beliefs about fruit and vegetable
consumption were assessed with four items and had a mean of .44 (SD = 1.42, Sla = .97;
see Appendix L), and families’ normative beliefs about skipping meals were assessed
with ﬁve items and had a mean of -.73 (SD = 1.28, SIa = .94; see Appendix M).
Perceptions of close friends’ normative beliefs about hit and vegetable consumption
were assessed with four items and had a mean of -.54 (SD = 1.19, SIa = .93; see
Appendix N), and close friends’ normative beliefs about skipping meals were assessed

with ﬁve items and had a mean of .45 (SD = 1.39, $10: = .96; see Appendix 0).

15

Results

As noted previously it was important that 1) each message was rated highly
(signiﬁcantly different ﬁom the midpoint of the scale) in terms of message realism and
message credibility, 2) that participants perceived the advocated message position as
favorable toward eating up to nine servings of hit and vegetables and unfavorable
toward skipping meals (signiﬁcantly different from the midpoint of the scale), and 3) that
the message position was signiﬁcantly different from the mean attitude and subjective
norm for the control group. The following analyses assess these issues.
Instrumentation

As noted previously, measures were comprised of seven-point, Likert-type items
on a scale ranging from -3 (Strongly Disagree) to +3 (Strongly Agree). Given that
speciﬁc items were speciﬁed a priori to measure only one factor, conﬁrmatory factor
analysis was employed to test the measurement model (Hunter & Gerbing, 1982). The
data were found to be consistent with the proposed factors. Internal consistency tests
showed that the errors calculated between items measuring the same construct were
within sampling error. Likewise, the parallelism test indicated that the errors calculated
between items measuring different constructs also were within sampling error.
Message Ratings

Message realism. For the attitude message, perceptions of realism had a mean of
1.69 (SD = .84) which was signiﬁcantly greater than the midpoint (zero) of the scale,
t(38) = 12.51, p < .001, two-tailed, r = .90. For the subjective norm message, perceptions
of realism had a mean of .99 (SD = .94) which was signiﬁcantly greater than the midpoint

(zero) of the scale, t(56) = 7.95, p < .001, two-tailed, r = .73. Given these results and the

16

magnitude of the effects, the probability that the observed means would have been as
different, or more different, from the midpoint of the scale is less than one in 1,000. Thus,
these data are strong grounds for doubting the viability of the null hypothesis (that
perceptions of message realism did not differ from the midpoint of the scale) and it is
rejected (see Abelson, 1995 for a more complete explanation of statistical signiﬁcance).

Message credibility. For the attitude message, perceptions of message credibility
had a mean of 1.74 (SD = .73) which was signiﬁcantly greater than the midpoint (zero) of
the scale, t(38) = 14.79, p < .001, two-tailed, r = .92. For the subjective norm message,
perceptions of message credibility had a mean of .97 (SD = 1.04) which was signiﬁcantly
greater than the midpoint (zero) of the scale, ((57) = 7.06, p < .001, two-tailed, r = .68.
Given these results and the magnitude of the effects, the probability that the observed
means would have been as different, or more different, ﬁ'om the midpoint of the scale is
less than one in 1,000. Thus, these data are strong grounds for doubting the viability of
the null hypothesis (that perceptions of message credibility do not differ from the
midpoint of the scale) and it is rejected.
Attitude Message Position

With regards to huh and vegetable consumption, participants perceived the
advocated position of the message as favorable toward eating up to nine servings of fruits
and vegetables every day (M = 2.75, SD = .72). Participants perceived that the message
position was signiﬁcantly above the midpoint (zero) of the scale, t(38) = 23.73, p < .001 ,
two-tailed, r = .97. Additionally, respondents perceived the message position to be
signiﬁcantly more favorable than the control group’s attitude toward eating up to nine

servings of fruit and vegetables a day (M = .58, SD = 1.18), t(77) = 9.89, p < .001, two-

17

tailed, r = .75. Given these results and the magnitude of the effects, these data are strong
grounds for doubting the validity of the null hypotheses (that perceptions of the
advocated position do not differ from the midpoint of the “scale or from baseline attitudes)
and they are rejected.

With regards to skipping meals, participants perceived the advocated position of
the message as unfavorable toward skipping meals (M = -2.88, SD = .51). Respondents
perceived that the message position was signiﬁcantly below the midpoint (zero) of the
scale, t(38) = -35.60, p < .001, two-tailed, r = -.99. Additionally, participants perceived
the message position to be signiﬁcantly less favorable than the control group’s attitude
toward skipping meals (M = -1.26, SD = 1.17), t(77) = -8.01, p < .001, two-tailed, r = -
.67. Given these results and the magnitude of the effects, these data are strong grounds
for doubting the validity of the null hypotheses (that perceptions of the advocated
position do not differ from the midpoint of the scale or from baseline attitudes) and they
are rejected.

Subjective Norm Message Position

With regards to ﬁ'uit and vegetable consumption, participants perceived the
overall advocated position of the subjective norm message as favorable toward eating up
to nine servings of fruits and vegetables every day (M = 1.01, SD = 1.32). Participants
perceived that the message position was signiﬁcantly above the midpoint (zero) of the
scale, t(56) = 5.77, p < .001, two-tailed, r = .61.

With regards to skipping meals, participants perceived the overall advocated
position of the message as unfavorable toward skipping meals (M = -1.53, SD = 1.63).

Respondents perceived that the message position was sigrriﬁcanﬂy below the midpoint

18

(zero) of the scale, t(57) = -7.1 1, p < .001, two-tailed, r = -.69. Given these results and the
magnitude of the effects, these data are grounds for doubting the validity of the null
hypotheses (that perceptions of the advocated position do not differ from the midpoint of
the scale or from baseline attitudes) and they are rejected.

Perceptions of family normative beliefs. With regards to fruit and vegetable
consumption, participants perceived that the message presented family normative beliefs
that were favorable toward eating up to nine servings of fruits and vegetables every day
(M = 1.89, SD = .96). Participants perceived that these normative beliefs presented in the
message were signiﬁcantly more favorable than control participants’ (M = .60, SD =
1.35) perceptions of their families’ normative beliefs, t(66) = 5.68, p < .001, two-tailed, r
= .57.

With regards to skipping meals, participants perceived the message presented
family normative beliefs that were unfavorable toward skipping meals (M = -2.00, SD =
1.02). Respondents perceived that these normative beliefs presented in the message were
signiﬁcantly less favorable than control participants’ (M = -.96, SD = 1.15) perceptions of
their families’ normative beliefs, t(66) = -4.61, p < .001, two-tailed, r = .49. Given these
results and the magnitude of the effects, these data are grounds for doubting the validity
of the null hypotheses (that perceptions of the advocated normative beliefs do not differ
from baseline perceptions of normative beliefs) and they are rejected.

Perceptions of close friends ’ normative beliefs. With regards to fruit and
vegetable consumption, participants perceived that the message presented close friend
normative beliefs that were favorable toward eating up to nine servings of fruits and

vegetables every day (M = 1.83, SD = .81). Participants perceived that these normative

l9

beliefs presented in the message were signiﬁcantly more favorable than control
participants’ (M = -.33, SD = 1.13) perceptions of their close ﬁ'iends’ normative beliefs,
t(68) = 12.01, p < .001, two-tailed, r = .82. b

With regards to skipping meals, participants perceived the message presented
family normative beliefs that were unfavorable toward skipping meals (M = -1.78, SD =
.92). Respondents perceived that these normative beliefs presented in the message were
signiﬁcantly less favorable than control participants’ (M = .39, SD = 1.38) perceptions of
their close friends’ normative beliefs, t(68) = -10.07, p < .001, two-tailed, r = -.77. Given
these results and the magnitude of the effects, these data are grounds for doubting the
validity of the null hypotheses (that perceptions of the advocated normative beliefs do not
differ from baseline perceptions of normative beliefs) and they are rejected.
Summary of Pilot Results

It was important that 1) each message was rated as highly realistic and credible, 2)
that participants perceived the advocated message position as favorable toward eating up
to nine servings of fruit and vegetables and unfavorable toward skipping meals
(signiﬁcantly different from the midpoint of the scale), and 3) that the message position
was signiﬁcantly different from the mean attitude and subjective norm for the control
group. The pilot results indicated that both messages were consistent these criteria and

were used subsequently in the main experiment.

20

MAIN EXPERIMENT
Method

Participants

Participants were recruited from undergraduate communication courses at a large
Midwestern university. Eight hundred seventy-two students were recruited from these
classes. Speciﬁcally, these 872 students received an email directly from the researcher
asking them to participate in the study. Of the 872 students who were recruited, 452
(51.8%) completed Survey 1. These 452 respondents received an email directly from the
researcher asking them to participate in the second survey 7-10 days later. Of the 452
participants who completed Survey 1, 380 (84.1%) completed Survey 2. These 380
respondents received an email directly from the researcher asking them to participate in
the third and ﬁnal survey 7-10 days later. Of the 380 participants who completed Survey
2, 334 (87.9%) completed Survey 3. Although 334 respondents completed all three
surveys, 25 were dropped ﬁom the main experiment because they had participated in the
pilot study, and 33 more were dropped because they submitted one or more surveys that
were largely incomplete. Therefore, all analyses are based on the remaining 276
participants.1

Respondents’ mean age was 20 years old (SD = 1.86), 29.7% were freshmen,
27.5% were sophomores, 27.2% were juniors, and 15.6% were seniors. Women
comprised 67.4% of the sample and 81.2% of the respondents self-reported their ethnicity
as Caucasian, 9.9% were African American, 2.9% were Hispanic American, 2.1% were
Asian American, 1.8% were of Middle Eastern descent, .7% were Native American, .7%

were Indian (from India), and .7% were Paciﬁc Islanders. Additionally, 52.5% of the

21

participants reported that they were currently living in university dorms, 14.5% lived with
their parents, and 33% lived off campus by themselves or with roommates.

Participants were informed that they were being asked to participate in a survey
related to student life and the use of university facilities such as nutrition counseling and
student health clinics. Respondents received research credit for their participation.
Design

This study utilized a single-factor, control group design such that participants
were assigned randomly to one of 3 conditions: 1) a control, no-message condition, 2) an
attitude-message condition in which participants were presented with a persuasive
message designed to result in favorable attitudes toward eating up to nine servings of
fruits and vegetables 3 day and not skipping meals (see Appendix P), and 3) a subjective
norm-message condition in which participants were exposed to a persuasive message
designed to result in perceptions of subjective norms in favor of the same eating
behaviors (see Appendix Q). These messages are the result of the pilot test reported
previously.

Procedure

Students enrolled in Communication comses were notiﬁed initially about the
study through an in-class announcement made by the researcher. Speciﬁcally, they were
told that the study entailed completing three different Student Life Surveys online.
Respondents were informed that they would earn class credit for their full participation.
The students were told to expect an email ﬁorn the researcher notifying them that the ﬁrst

survey was online and ready for them to complete. Those students who chose not to

22

participate voluntarily were provided with an alternative assignment worth equal credit
(determined by course instructors).

Participants received an email from the researcher indicating that the ﬁrst survey
was online. The email provided the URL of the website and informed participants that the
ﬁrst survey would remain online 24 hours a day for 3 days so that they could complete
the survey at a time that was most convenient for them. When participants logged on to
the website the server assigned automatically each respondent a random, personalized
number that was used to code each participant’s data so that each time a respondent
returned to the website their data was compiled by the web server and avoided the need
for personal identiﬁers to match surveys. After the ﬁnal survey, the server created a list
of email addresses of those respondents who participated in the study and the researcher
matched the list of emails generated by the computer to the class lists provided by the
respondents’ instructors in order to determine who earned the research credit. The email
list was in no way connected to respondents’ questionnaire responses. Therefore, all
identifying information was kept separate from participants’ self-report answers.

When participants went to the website for the ﬁrst survey (see Appendix R for the
Time One Survey) they ﬁrst read a brief introduction to the study and were asked to read
carefully the consent form. Upon completion of reading the consent form participants
were instructed that “By clicking the AGREE button you indicate your voluntary
participation in this study.” Respondents were offered the choice to click “AGREE” or “I
DO NOT AGREE.” If participants clicked “AGREE” they were directed to the ﬁrst
survey. If they clicked “I DO NOT AGREE” they were directed to a page that thanked

them for their time.

23

Once logged on to the ﬁrst survey, participants were asked to answer a series of
questions about their knowledge and use of university facilities, their eating habits, their
extra-curricular activities, study habits, and other general questions about their health and
lifestyles. Many of these questions were ﬁller items designed to avoid priming
respondents prior to reading a message. Next they were assigned randomly to one of
three message conditions. In one condition they received no message (control group). The
rest of the participants were assigned to either (1) a message designed to increase their
favorable attitudes toward eating up to 9 servings of fruit and vegetables a day and not
skipping meals, or (2) increase perceptions that their family and close friends are in favor
of those same healthy eating behaviors. Subsequently, participants were asked to answer
a series of questions designed to measure their attitudes and beliefs about eating fruits
and vegetables and skipping meals, their perceptions of their family’s and close friends’
attitudes and beliefs about the same topics, and participants’ behavioral intention
regarding these speciﬁc eating behaviors.

After respondents completed the questionnaire they were instructed to click on a
“SUBMIT” button to submit their completed survey. Once they clicked “SUBMIT” they
were directed to a page that thanked them for their time and reminded them that in 7-10
days they would receive an email notifying them that the second survey was online and
ready for them to complete. This email was sent to only those students who completed
the ﬁrst survey.

Approximately 7-10 days after the ﬁrst survey was taken offline participants
received an email message notifying them that the second survey was online and ready

for them to complete (see Appendix S for the Time Two Survey). The second survey

24

asked participants a series of follow-up questions regarding their eating habits, use of
university facilities, exercise, and other general behaviors within the last week. After
completing the second survey participants were instructed to click on the “SUBMIT”
button to submit their completed survey and were directed to a page that thanked them
for their time.

Instrumentation

All scales, unless otherwise noted, were similar to those designed for the pilot
study. Speciﬁcally, measures were comprised of seven-point, Likert-type items on a scale
ranging from -3 (Strongly Disagree) to +3 (Strongly Agree). The measurement reported
here is ﬁ'om the Time One Survey unless otherwise noted.

Attitudes. Participants’ attitudes toward eating up to nine servings of ﬁ'uits and
vegetables every day were measured two different ways. The ﬁrst was an attitude scale
comprised of four items (see Appendix J). The mean attitude toward eating up to 9
servings of fruit and vegetables a day was 1.01 (SD = 1.68) and standardized item alpha
was .96. The second method of measuring participants’ attitudes was a belief scale that
was consistent with Fishbein and Ajzen’s (1975) conceptualization. Speciﬁcally,
participants were asked ﬁrst to rate the degree to which they held 8 different beliefs about
eating up to 9 servings of fruit and vegetables everyday (see Appendix T) and
subsequently rated how important each belief was to them (see Appendix U). Importance
was scaled 0 (not at all important) to 7 (very important). Each belief was multiplied by
the importance ascribed to it and the resulting evaluative beliefs were summed to form
the second attitude measure (this scale ranged from -21 to +21). The mean attitude

toward eating up to 9 servings of fruit and vegetables every day on the F ishbein and

25

Ajzen belief scale was 5.53 (SD = 6.08, range = -12.0 to 21.0) and standardized item
alpha was .86.

Respondents’ attitudes toward skipping meals were also measured in two ways.
The ﬁrst was an attitude scale comprised of four items. The mean attitude toward
skipping meals was -1 .75 (SD = 1.12) and standardized item alpha was .93 (see Appendix
K). The second method of measuring participants’ attitudes was consistent with F ishbein
and Ajzen’s (1975) conceptualization and was comprised of eight items (this scale ranged
ﬁom -21 to +21; see Appendix V for the belief items and Appendix U for the importance
items). The mean attitude toward skipping meals on the F ishbein and Ajzen scale was
-9.79 (SD = 6.48, range = -21.0 to 7.0) and standardized item alpha was .83.

Subjective norm: Family. Participants’ familial subjective norms regarding ﬁ'uit
and vegetable consumption were measured two different ways. The ﬁrst was a subjective
norm scale comprised of four items (see Appendix L). The mean family subjective norm
toward eating up to 9 servings of fruit and vegetables a day was .48 (SD = 1.75) and
standardized item alpha was .93. The second method of measuring familial subjective
norms was consistent with Fishbein and Ajzen’s ( 1975) conceptualization. Speciﬁcally,
participants were asked ﬁrst to rate the degree to which they perceived their families held
8 different beliefs about eating up to 9 servings of fruit and vegetables everyday
(normative beliefs; see Appendix W) and subsequently rated how important it was for
them to do what their family thought they should do (motivation to comply; see
Appendix X). Importance was scaled 0 (not at all important) to 7 (very important). Each
normative belief was multiplied by the motivation to comply and the resulting scores

were summed to form the second familial subjective norm scale (this scale ranged ﬁ'om

26

-21 to +21). The mean familial subjective norm toward eating up to 9 servings of mu
and vegetables every day on the Fishbein and Ajzen scale was 1.77 (SD = 5.14, range =
-13.0 to 16.0) and standardized item alpha was .81. .

Respondents’ familial subjective norms about skipping meals were also measured
two different ways. The ﬁrst was a subjective norm scale comprised of four items (see
Appendix M). The mean family subjective norm toward eating up to 9 servings of fruit
and vegetables a day was .48 (SD = 1.7 5) and standardized item alpha was .93. The
second method of measuring familial subjective norms was consistent with Fishbein and
Ajzen’s (1975) conceptualization and included eight items (this scale ranged ﬁ'om ~21 to
+21; see Appendix Y for the belief items and Appendix X for motivation to comply). The
mean familial subjective norm toward skipping meals on the Fishbein and Ajzen scale
was -7.54 (SD = 6.18, range = -21.0 to 7.0) and standardized item alpha was .80.

Subjective norm: Close friends. Participants’ close ﬁiends subjective norm
regarding fruit and vegetable consumption was measured two different ways. The ﬁrst
was a subjective norm scale comprised of four items (see Appendix N). The mean close
friend subjective norm toward eating up to 9 servings of huh and vegetables a day was
-.68 (SD = 1.56) and standardized item alpha was .98. The second method of measuring
familial subjective norms was consistent with Fishbein and Ajzen’s (1975)
conceptualization. Speciﬁcally, participants were asked ﬁrst to rate the degree to which
they perceived their close friends held 8 different beliefs about eating up to 9 servings of
fruit and vegetables everyday (normative beliefs; see Appendix Z) and subsequently rated
how important it was for them to do what their close ﬁ'iends thought they should do

(motivation to comply; see Appendix AA). Importance was sealed 0 (not at all important)

27

to 7 (very important). Each normative belief was multiplied by the motivation to comply
and the resulting scores were summed to form the second close friend subjective norm
scale (this scale ranged from -21 to +21). The mean close friend subjective norm toward
eating up to 9 servings of fruit and vegetables every day on the Fishbein and Ajzen scale
was -.06 (SD = 5.30, range = -21.0 to 18.0) and standardized item alpha was .85.

Respondents’ close friends subjective norm about skipping meals was also
measured two different ways. The ﬁrst was a subjective norm scale comprised of four
items (see Appendix 0). The mean close friends subjective norm toward skipping meals
was .28 (SD = 1.53) and standardized item alpha was .97. The second method of
measuring close friend subjective norms was consistent with F ishbein and Ajzen’s (1975)
conceptualization and included eight items (this scale ranged hour -21 to +21; see
Appendix BB for belief items and Appendix AA for motivation to comply). The mean
close friend subjective norm toward skipping meals on the F ishbein and Ajzen scale was
-3.61 (SD = 4.92, range = —l9.0 to 9.0) and standardized item alpha was .83.

Intentions. Participants’ intent to eat up to 9 servings of fruit and vegetables a day
was measured with 4 items and had a mean of -.57 (SD = 1.50) and standardized item
alpha was .93 (see Appendix CC). Respondents’ intent to skip meals was measured with
3 items and had a mean of -1.01 (SD = 1.44) and standardized item alpha was .93 (see
Appendix DD). In order to be consistent with the position advocated by the message,
participants’ intent to increase fruit and vegetable consumption should be positive and
their intent to skip meals should be negative.

Behavior: Fruit and vegetable consumption. Fruit and vegetable consumption was

assessed at two points in time. First, it was measured at the beginning of Survey 1 prior to

28

any exposure to a persuasive message and was assessed with two open-ended questions
that asked “On average, how many servings of ﬁ'uit do you eat per day (one serving ~ 1
cup of fresh fruit)?” and “On average, how many servings of vegetables do you eat per
day, not including starchy vegetables like potatoes and corn (one serving ~ 1 cup of raw
vegetables)?” The answers to both items were summed to create an assessment of fruit
and vegetable consumption. The average number of ﬁ'uit and vegetable servings
consumed per day had a mean of 2.97 (SD = 2.45) at the time of Survey 1.
Approximately 7 -10 days later participants were asked the same two open-ended
questions in order to determine how many daily servings of ﬁ'uit and vegetables they had
eaten in the past week and results indicated that participants ate an average of 4.09
servings of fruit and vegetables (SD = 2.08).

Behavior: Skipping meals. The number of meals people skip per day also was
assessed at two points in time. First, it was measured at the beginning of Survey 1 prior to
any exposure to a persuasive message and was assessed with one open-ended question
that asked “On an average day, how many meals do you skip?” On average, participants
were skipping .77 meals per day (SD = .67) at the time of Survey 1. Approximately 7-10
days later participants were asked the same open-ended question in order to determine
how many meals they had been skipping in the past week. Results indicated that
participants skipped an average of .74 meals a day (SD = .70).

Overview of Analyses

Conﬁnnatory factor analysis (Hunter & Gerbing, 1982) was employed to test the

measurement model and path analytic techniques were used to test the Theory of

Reasoned Action within the two systems of change. The conﬁrmatory factor analysis

29

results will be presented ﬁrst, followed by a test of the TRA models. Because the current
study examined two behaviors, the results for fruit and vegetable consumption will be
presented ﬁrst and those for skipping meals will be presented subsequently.
Results

Instrumentation

Given that speciﬁc items were speciﬁed a priori to measure only one factor,
conﬁrmatory factor analysis was employed to test the measurement model (Hunter &
Gerbing, 1982). The data were not consistent with the Fishbein and Ajzen measurement
models. Speciﬁcally, the attitude belief scale, the familial normative belief scale, and the
close ﬁiends normative belief scale for both the fruit consumption and skipped meals
topics were not consistent with the proposed measurement models. Internal consistency
tests showed that the errors calculated between items measuring the same construct were
larger in magnitude than one would expect from sampling error alone. Additionally, the
tests of parallelism also showed that the errors calculated between items measuring
different constructs were larger in magnitude than one would expect from sampling error
alone. Therefore, these scales were all dropped from further analyses.2

The remaining four scales (attitude, family subjective norm, close ﬁiends
subjective norm, and behavioral intent) for each topic were found to be consistent with
the proposed factors. Internal consistency tests showed that the errors calculated between
items measuring the same construct were within sampling error. Likewise, the parallelism
test indicated that the errors calculated between items measuring different constructs also
were within sampling error. Therefore, these scales were retained for all further analyses

(see Tables 1 and 2 for a report of descriptive statistics by topic).3

30

Table 1

Means and Standard Deviations (in parentheses) for Fruit and Vegetable Consumption by Experimental
Condition

 

Family Close Friend

 

Subjective Subjective Consumption Consumption
Condition Attitude Norm Norm Intent Time 1 Time 2
Control Group .02 -.06 -1.06 -1. 15 2.74 4.15
(1.58) (1.75) (1.48) (1.36) (2.22) (5.44)
Attitude Message 1.63 .40 -.49 -.32 3.21 4.18
(1.41) (1.79) (1.55) (1.38) (2.20) (2.50)
Subjective Norm 1.49 1.13 -.49 -. 18 3.01 3.95
Message (1.01) (1.51L (1.60) (1.56) (2.86) (3.61)

 

Note. Attitude, Family Subjective Norm, Close Friend Subjective Norm, and Intent were scaled -3
(Strongly Disagree) to +3 (Strongly Agree). The measure of Consumption was open-ended.

Table 2

Means and Standard Deviations (in parentheses) for Skipping Meals Topic by Experimental Condition

 

Family Close Friend

 

Subjective Subjective Skipped Skipped
Condition Attitude Norm Norm Intent Time 1 Time 2
Control Group -1.81 -1.36 -.03 -1.53 .78 .72
(1.14) (1.42) (1.81) (1.61) (.71) (.73)
Attitude Message -1.85 -.99 .39 -1.49 .75 .78
(1.03) (1.70) (1.65) (1.47) (.66) (.58)
Subjective Norm -1.76 -l .27 .35 -1.26 .78 .71
Message (1.36) (1.54) (1.59) (1.47) Q65) (.78)

 

Note. Attitude, Family Subjective Norm, Close Friend Subjective Norm, and Intent were sealed -3
(Strongly Disagree) to +3 (Strongly Agree). The measure of Skipped Meals was open-ended.

Evaluation of the Models

To test each of the following models the least squares criterion was used to
estimate the parameters, parameter size was examined, and the ﬁt of the model was

assessed. Parameter size was determined in the path diagram by performing a simple

31

regression of each endogenous variable onto its causal antecedent and model ﬁt was
tested by comparing the estimated parameter size to the reproduced correlations (see
Hunter & Gerbing, 1982 for information on reproducing correlations in path analysis). To
the extent that the path coefﬁcients are substantial and the differences between parameter
estimates and reproduced correlations (errors) are attributable to sampling error, the
model is said to be consistent with the data. If errors are larger than what is expected
from sampling error, the model is said to be inconsistent with the data.
Fruit and Vegetable Consumption Results

Inﬂuencing change through attitudes. This ﬁrst model posited that when people
read a message persuading them to increase their ﬁ'uit and vegetable consumption they
are more likely to have favorable attitudes toward increased consumption (compared to a
control group). Additionally, consistent with the TRA, the more favorable one’s attitude
toward increased fruit and vegetable consumption, and the more favOrable one perceives
family and close ﬁiends to be toward the same behavior (subjective norm), the greater
one’s intention will be to increase consumption of ﬁ'uit and vegetables and will
subsequently engage in such behavior. The correlations employed to estimate the ﬁt of
the model parameters are presented in Table 3, and the path coefﬁcients are presented in

Figure 1.

32

Table 3

Zero-order Correlations Used to Calculate Parameter Estimates in Model I

 

1 2 3 4 5 6
1. Message [.00
2. Attitude .47" .96
3. Family Subjective Norm .14 .51" .98
4. Close Friends Subjective Norm .17“ .41 " .46" .98
5. Behavioral Intent .28" .57" .51" .42“ .93
6. Fruit &Vegetable Consumption -.03 -.02 .04 .08 .19“ 1.00

 

Note. Message was coded such that 1 = Attitude Message and 0 = Control Group. Standardized item alpha
appears in the diagonal. " indicates p < .01, two-tailed, and * indicates p < .05, two-tailed.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/"—‘ Message '48 Attitude
,___,
.13 .41
>— Family .25 .20 .
Subjective Intent F rurt & Veg
19 Norm Consumption
.47 .15
% Close
Friends
¥ Subjective
Norm

 

 

 

Figure 1. Model depicting the TRA when a persuasive message is aimed at attitudes with path coefﬁcients
corrected for attenuation due to measurement error.

One may observe from Figure 1 that all of the path coefﬁcients are in the
direction predicted but not all paths are ample. The coefﬁcient linking the message and
attitude was .48, P(.36 S p S .60) = .95, indicating that the message induction had a

substantial effect on attitudes toward eating up to 9 servings of fruit and vegetables per

33

day. Attitudes, in turn, affected behavioral intentions (path coefﬁcient = .40) such that the
more favorable participants’ attitude toward increasing ﬁ'uit and vegetable consumption,
the greater participants’ intentions to engage in the behavior, P(.26 S p S .54) = .95. The
coefﬁcient linking family subjective norm and intent was .25, P(.09 S p S .41) = .95,
indicating the more favorable participants perceive their family to be toward increasing
fruit and vegetable consumption, the more likely they are to intend to engage in such
behavior. The coefﬁcient linking behavioral intent and behavior was .20, P(.04 S p S .36)
= .95 demonstrating that intentions to eat more fruit and vegetables resulted in subsequent
increases in fruit and vegetable consumption. The coefﬁcient linking close friends
subjective norm and intent, however, was .15, P(-.01 _<_ p S .31) = .95 and was within
sampling error of zero. Although this path is not large in magnitude, the ﬁt of the overall

model was tested nonetheless.

The differences between predicted and obtained correlations for all unconstrained
bivariate relationships in the model were examined and three of the 7 differed
substantially from what was expected from sampling error. These largest errors were -.1 1
(between attitude and behavior), .34 (between attitude and close ﬁiend subjective norm)
and .46 (between attitude and family subjective norm). Further, the global test for
goodness of ﬁt indicated that the data were not consistent with the model, 380) = 27.84,
p < .001. Given that not all of the path coefﬁcients were large in magnitude, and that the
model and parameter estimates did not predict accurately the unconstrained correlations,
the data were judged to be inconsistent with the model. Consequently, a post hoc search

for an alternative model that ﬁt the data was undertaken.

34

 

 

 

 

.48 .60 .20

Message Attitude > Intent > Fruit & Veg

Consumption

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2. Revised model with path coefﬁcients corrected for attenuation due to measurement error.

Results indicated that the data were consistent with a revised model (see Figure 2).
Speciﬁcally, the coefﬁcient linking the message and attitude was .48, P(.36 S p S .60) =
.95, indicating that the message induction had a substantial effect on attitudes toward
eating up to 9 servings of fruit and vegetables per day. Attitudes predicted one’s intent to
eat up to nine servings of fruit and vegetables per day [.60, P(.50 S ,0 S .70) = .95] which
subsequently predicted behavior [.20, P(.04 S p S .36) = .95]. All path coefﬁcients were
ample and were in the direction predicted. Additionally, an examination of the
differences between predicted and obtained correlations for all the unconstrained
bivariate relationship revealed that the errors did not differ substantially from what was
expected from sampling error. Further, the global test for goodness of ﬁt indicated that
the data were consistent with the model, 352(3) = 2.26, p = .52. Given that the path
coefﬁcients were ample, and that the model and parameter estimates predicted accurately
the unconstrained correlations, the data were judged to be consistent with the alternative
model.

Inﬂuencing change through subjective norms. This second model posited that
when people read a message persuading them that their family and close ﬁiends want
them to increase their fruit and vegetable consumption they are more likely to perceive
familial and close ﬁiend subjective norms as favorable toward increased consumption

(compared to a control group). Additionally, consistent with the TRA, the more favorable

35

one’s attitude toward increased hit and vegetable consumption, and the more favorable
one perceives family and close friends to be toward the same behavior (subjective norm),
the greater one’s intention will be to increase consumption of mu and vegetables and
will subsequently engage in such behavior. The correlations employed to estimate the ﬁt
of the model parameters are presented in Table 4, and the path coefﬁcients are presented

in Figure 3.

Table 4

Zero-order Correlations Used to Calculate Parameter Estimates in Model 3

 

1 2 3 4 5 6
1. Message 1.00
2. Attitude .42" .96
3. Family Subjective Norm .34“ .52" .98
4. Close Friends Subjective Norm .19“ .38" .43” .98 '
5. Behavioral Intent .29" .55” .45“ .38" .93
6. Fruit &Vegetable Consumption -.07 -.04 .01 .06 .15 1.00

 

Note. Message was coded such that 1 = Subjective Norm Message and 0 = Control Group. Standardized
item alpha appears in the diagonal. ** indicates p < .01, two-tailed, and " indicates p < .05, two-tailed.

36

 

Attitude

 

 

 

     

 

.42 .16

. Fruit & Veg
Consumption

 

Family
Subjective
.3 Norm

 

 

 

 

 

 

 

\ Message

 

 

 

 

. 19 Close
Friends
Subjective
Norm

 

 

 

Figure 3. Model depicting the TRA when a persuasive message is aimed at subjective norm with path
coefﬁcients corrected for attenuation due to measurement error.

The OLS estimates shown in Figure 3 indicate that all of the path coefﬁcients are
in the direction predicted but not all paths are large in magnitude. The coefficient linking
the message and family subjective norm was .34, P(.20 S p S .48) = .95, indicating that
the message induction had a substantial effect on participants’ perceptions that their
family was in favor of eating up to 9 servings of fruit and vegetables per day. The path
coefﬁcient from message to close friends subjective norm was .19, P(.OS S p S .33) = .95,
also indicating that the message induction had an effect on participants’ perceptions that
their close friends were in favor of increased fruit and vegetable consumption. The
coefﬁcient linking attitude to intent was .43, P(.27 S p S .59) = .95 such that the more

favorable participants’ attitude toward increasing ﬁ'uit and vegetable consumption, the

greater participants’ intentions to engage in the behavior.

The coefﬁcients linking family subjective norm and intent [.17, P(-.01 S p S .35)

= .95], close friends subjective norm and intent [.15, P(-.01 S p S .31) = .95], and intent

37

and behavior [. 16, P(.00 S p _<_ .32) = .95] were within sampling error of zero. Although

these paths are not ample, the ﬁt of the overall model was tested.

The differences between predicted and obtained correlations for all unconstrained
bivariate relationships in the model were examined and ﬁve of the 7 differed substantially
from what was expected from sampling error. These errors were -.11 (between message
and behavior), -.11 (between attitude and behavior), .39 (between attitude and family
subjective norm), .31 (between attitude and close friend subjective norm), and .37
(between family and close friend subjective norms). Further, the global test for goodness
of ﬁt indicated that the data were not consistent with the model, x2(8) = 34.36, p < .001.
Given that not all of the path coefﬁcients were large in magnitude, and that the model and
parameter estimates did not predict accurately the unconstrained correlations, the data
were judged to be inconsistent with the model. A subsequent post hoc search for an
alternative model that ﬁt the data was undertaken, but the data were not consistent with
an alternative model.

The model with no change. In order to determine if the TRA could predict
accurately people’s fruit and vegetable consumption when no message was present, the
TRA was tested using only the control group. As noted previously, the TRA predicts that
the more favorable one’s attitude toward increased fruit and vegetable consumption, and
the more favorable one perceives family and close friends to be toward the same behavior
(subjective norm), the greater one’s intention will be to increase consumption of ﬁ'uit and
vegetables and will subsequently engage in such behavior. The correlations employed to
estimate the ﬁt of the model parameters are presented in Table 5, and the path

coefﬁcients are presented in Figure 4.

38

Table 5

Zero-order Correlations Used to Calculate Parameter Estimates in Model 4

 

1 2 3 4 5
1. Attitude .96
2. Family Subjective Norm .36" .98
3. Close Friends Subjective Norm .37" .38" .98
4. Behavioral Intent .43" .41” .37"I .93
5. Fruit &Vegetable Consumption .05 .05 .09 .23‘ I .00

 

Note. Standardized item alpha appears in the diagonal. “ indicates p < .01, two-tailed, and " indicates p <

 

 

 

 

 

 

 

 

 

 

 

 

.24 Fruit & Veg
Intent ' . Consumption

 

 

 

 

 

.05, two-tailed.
Attitude
.36 .29
Family
Subjective -25
.37 Norm
38
Close .1 8
Friends
x Subjective
Norm

 

 

 

Figure 4. Model depicting the TRA when no persuasive message is present with path coefﬁcients corrected

for attenuation due to measurement error.

One may observe from Figure 4 that all of the path coefﬁcients are in the

direction predicted but that the relationship between close friends subjective norm and

intent is not very large in magnitude. An examination of each link in the model revealed

that the coefficient linking attitude and intent was .29, P(.ll S p S .47) = .95, indicating

that the more favorable participants’ attitude toward increasing fruit and vegetable

39

consumption, the greater participants’ intentions to engage in the behavior. The path
coefﬁcient from family subjective norm to intent was .25, P(.05 .<_ p _<. .45) = .95,
indicating the more favorable participants perceive their family to be toward increasing
fruit and vegetable consumption, the more likely they are to intend to engage in such
behavior. The coefﬁcient linking close friends subjective norm and intent (.18) was not
within sampling error of zero, P(.02 S p S .34) = .95. This demonstrated that the more
favorable participants perceive their close ﬁiends to be toward increasing fruit and
vegetable consumption, the more likely they are to intend to engage in such behavior.
The coefﬁcient linking behavioral intent and behavior was .24, P(.06 S p S .42) = .95
indicating that intentions to eat more ﬁ'uit and vegetables resulted in subsequent increases

in fruit and vegetable consumption.

The differences between predicted and obtained correlations for all unconstrained
bivariate relationships in the model were examined and none differed substantially from
what was expected from sampling error. Further, the global test for goodness of ﬁt
indicated that the data were consistent with the model, x2(3) = .29, p = .96. Given that the
path coefﬁcients were ample, and that the model and parameter estimates predicted
accurately the unconstrained correlations, the data were judged to be consistent with the
model.

Summary of fruit and vegetable results. The predicted relationships outlined by
the TRA were examined when 1) attitudes were inﬂuenced by a persuasive message, 2)
subjective norms were inﬂuenced by a persuasive message, and 3) no message was
present. The results indicated that when people’s attitude or subjective norm is inﬂuenced

by a persuasive message the TRA’s predictions are not consistent with the data. The test

40

of the TRA when no change is present, however, provided evidence that its predictions
are consistent with the control group data. Therefore, the TRA’s predictive power appears
to lie in its ability to predict behavior when attitudes and subjective norms are measured
in their current state but not when they are changing. To provide a second test of these
relationships, an examination of skipping meals is reported subsequently.
Skipped Meals Results

Inﬂuencing change through attitudes. This ﬁrst model posited that when people
read a message persuading them to avoid skipping meals every day they are more likely
to have unfavorable attitudes toward skipping meals (compared to a control group).
Additionally, consistent with the TRA, the more unfavorable one’s attitude toward
skipping meals, and the more unfavorable one perceives family and close friends to be
toward the same behavior (subjective norm), the less intent one will have to skip meals
and will subsequently avoid such behavior. The correlations employed to estimate the ﬁt
of the model parameters are presented in Table 6, and the path coefﬁcients are presented

in Figure 5.

41

Table 6

Zero-order Correlations Used to Calculate Parameter Estimates in Model 5

 

1 2 3 4 S 6
1. Message 1.00
2. Attitude -.04 .93
3. Family Subjective Norm .11 .42” .96
4. Close Friends Subjective Norm .12 .26" .46" .97
5. Behavioral Intent .00 .62“ .51" .27'MI .93
6. Number of Meals Skipped per Day .04 .17“ .04 .06 .23“ 1.00

 

Note. Message was coded such that 1 = Attitude Message and 0 = Control Group. Standardized item alpha
appears in the diagonal. ” indicates p < .01, two-tailed, and " indicates p < .05, two-tailed.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/-"—‘ Message ”04 Attitude
.11 .61
>- Family .08 .24 ‘ Number of
12 f Subjective Intent Meals Skipped
' Norm per Day
.46 .09
Close
Friends
¥ Subjective
Norm

 

 

 

Figure 5. Model depicting the TRA when a persuasive message is aimed at attitudes with path coefﬁcients
corrected for attenuation due to measurement error.

One may observe ﬁ'om Figure 4 that all of the path coefﬁcients are in the
direction predicted but not all paths are ample. The coefﬁcient linking attitude and intent
was .61, P(.49 S pS .73) = .95, demonstrating that the less favorable one’s attitude
toward Skipping meals the less intent one had to Skip meals. The path between intent and

behavior was .24, P(.10 S p S .38) = .95, indicating that the less intent one had to skip

42

meals, fewer meals were actually skipped (in other words, they avoided the behavior).
The coefﬁcients linking the message and attitude [-.04, P(-.20 S p S .12) = .95], family
subjective norm and intent [.08, P(-.10 S p S .26) = .95], and close friends subjective
norm and intent [.09, P(-.07 S p S .25) = .95] all were within sampling error of zero.

Although these paths indicate that the data were not consistent with this model, the ﬁt of

the model was assessed subsequently.

The differences between predicted and obtained correlations for all unconstrained
bivariate relationships in the model were examined and two errors differed substantially
from what was expected from sampling error. These largest errors were .28 (between
attitude and close friends subjective norm) and .45 (between attitude and family
subjective norm). Further, the global test for goodness of ﬁt indicated that the data were
not consistent with the model, 380) = 22.25, p < .001. Given that not all of the path
coefﬁcients were large in magnitude, and that the model and parameter estimates did not
predict accurately the unconstrained correlations, the data were judged to be inconsistent
with the model. Consequently, a post hoc search for an alternative model that ﬁt the data

was undertaken.

 

 

 

. .67 «24 Number of
Attitude Intent Meals Skipped

per Day

 

 

 

 

 

 

 

 

 

 

 

Figure 6. Revised model with path coefficients corrected for attenuation due to measurement error.

Results indicated that the data were consistent with a revised model (see Figure

6). Speciﬁcally, attitudes toward Skipping meals predicted one’s intent to skip meals [.67,

43

P(.57 S p S .77) = .95] which subsequently predicted behavior [.24, P(.10 S p S .38) =
.95]. Both path coefﬁcients were ample and were in the direction predicted. Additionally,
an examination of the differences between predicted and obtained correlations for all the
unconstrained bivariate relationship revealed that the error did not differ substantially
from what was expected from sampling error. Further, the global test for goodness of ﬁt
indicated that the data were consistent with the model, 380) = .03, p = .99. Given that the
path coefﬁcients were ample, and that the model and parameter estimates predicted
accurately the unconstrained correlation, the data were judged to be consistent with the
alternative model. One Should note, however, that because these data are consistent with
the model in Figure 6 the data are also consistent with the reverse model and one Should
use caution when interpreting these results.

Inﬂuencing change through subjective norms. This second model posited that
when people read a message persuading them that their family and close ﬁiends want
them to avoid skipping meals they are more likely to perceive familial and close ﬁiend
subjective norms as unfavorable toward increased consumption (compared to a control
group). Additionally, consistent with the TRA, the more unfavorable one’s attitude
toward skipping meals, and the more unfavorable one perceives family and close friends
to be toward the same behavior (subjective norm), the lesser one’s intention will be to
Skip meals and will subsequently avoid such behavior. The correlations employed to
estimate the ﬁt of the model parameters are presented in Table 7, and the path

coefﬁcients are presented in Figure 7.

Table 7

Zero-order Correlations Used to Calculate Parameter Estimates in Model 7

 

l 2 3 4 5 6
1. Message 1.00
2. Attitude .01 . 93
3. Family Subjective Norm .03 .46“l .96
4. Close Friends Subjective Norm .12 .24“ .36“ .97
5. Behavioral Intent .07 .70‘”‘I .40“I .27“ .93
6. Number of Meals Skipped per Day -.01 .21 " .08 .06 .21" 1.00

 

Note. Message was coded such that 1 = Subjective Norm Message and 0 = Control Group. Standardized
item alpha appears in the diagonal. " indicates p < .01, two-tailed, and "‘ indicates p < .05, two-tailed.

 

Attitude

 

 

 

   

 

.01 .22

 

Number of
Meals Skipped

per Day

  

Intent

 

 

Family
Subjective
.03 Norm

 

 

 

 

 

 

 

 

\ Message

 

 

 

 

. 12 Close
Friends
Subjective
Norm

 

 

 

Figure 7. Model depicting the TRA when a persuasive message is aimed at subjective norm with path
coefﬁcients corrected for attenuation due to measurement error.

The OLS estimates Shown in Figure 7 indicate that all of the path coefﬁcients are
in the direction predicted but not all paths are ample. The coefﬁcient linking attitude and
intent was .71, P(.59 S p S .83) = .95, demonstrating that the less favorable one’s attitude
toward Skipping meals the less intent one had to Skip meals. The path between intent and

behavior was .22, P(.08 S p S .36) = .95, indicating that the less intent one had to skip

45

meals, fewer meals were actually skipped. The coefficients linking the message and
family subjective norm [.03, P(-.13 S pS .17) = .95], message and close friends
subjective norm [.12, P(-.02 S p S .26) = .95], family subjective norm and intent [.05, P(-
.13 S p S .23) = .95], and close ﬁiends subjective norm and intent [.09, P(-.07 S pS .25)
= .95] all were within sampling error of zero. Although these paths indicate that the data

were not consistent with this model, the ﬁt of the model was assessed subsequently.

The differences between predicted and obtained correlations for all unconstrained
bivariate relationships in the model were examined and three errors differed substantially
from what was expected ﬁ'om sampling error. These largest errors were .25 (between
attitude and close friends subjective norm), .37 (between family and close friends
subjective norms), and .49 (between attitude and family subjective norm). Further, the
global test for goodness of ﬁt indicated that the data were not consistent with the model,
x2(8) = 36.78, p < .001. Given that not all of the path coefﬁcients were large in
magnitude, and that the model and parameter estimates did not predict accurately the
unconstrained correlations, the data were judged to be inconsistent with the model.

Consequently, a post hoc search for an alternative model that ﬁt the data was undertaken.

 

 

 

. .75 .22
Attitude Intent Number of

Meals Skipped
per Day

 

 

 

 

 

 

 

 

 

 

 

Figure 8. Revised model with path coefﬁcients corrected for attenuation due to measurement error.

Results indicated that the data were consistent with a revised model (see Figure

8). Speciﬁcally, attitudes toward Skipping meals predicted one’s intent to Skip meals [.75,

46

P(.67 S p S .83) = .95] which subsequently predicted behavior [.22, P(.08 S pS .36) =
.95]. Both path coefﬁcients were ample and were in the direction predicted. Additionally,
an examination of the differences between predicted and obtained correlations for all the
unconstrained bivariate relationship revealed that the error did not differ substantially
from what was expected from sampling error. Further, the global test for goodness of ﬁt
indicated that the data were consistent with the model, x2(1) = .26, p = .61. Given that the
path coefﬁcients were ample, and that the model and parameter estimates predicted
accurately the unconstrained correlation, the data were judged to be consistent with the
altemative model. As noted previously, because these data are consistent with the model
in Figure 8 the data are also consistent with the reverse model and one should use caution
when interpreting these results.

The model with no change. In order to determine if the TRA could predict
accurately the number of meals people Skip a day when no message was present, the TRA
was tested using only the control group. AS noted previously, the TRA predicts that the
more unfavorable one’s attitude toward skipping meals, and the more unfavorable one
perceives family and close ﬁiends to be toward the same behavior (subjective norm), the
lesser one’s intention will be to Skip meals and will subsequently avoid such behavior.
The correlations employed to estimate the ﬁt of the model parameters are presented in

Table 8, and the path coefﬁcients are presented in Figure 9.

47

Table 8

Zero-order Correlations Used to Calculate Parameter Estimates in Model 9

 

1 2 3 4 5
1. Attitude .93
2. Family Subjective Norm .46" .96
3. Close Friends Subjective Norm .31" .36" .97
4. Behavioral Intent .66" .35" .25" .93
5. Number of Meals Skipped per Day .23“ -.02 .04 .24‘ 1.00

 

Note. Standardized item alpha appears in the diagonal. ** indicates p < .01, two-tailed, and * indicates p <

.05, two-tailed.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C Attitude
.46 .69
\ﬂ Family .02 .25 Number of
Subjective Intent Meals Skipped
.31 r'" Norm per Day
.36
Close .03
g Friends
Subjective
k Norm

 

 

 

Figure 9. Model depicting the TRA when no persuasive message is present with path coefficients corrected

for attenuation due to measurement error.

One may observe from Figure 9 that all of the path coefﬁcients are in the

direction predicted but that two paths are very small in magnitude. An examination of

each link in the model revealed that the coefficient linking attitude and intent was .69,

P(.51 S p S .87) = .95, indicating that the more unfavorable participants’ attitude toward

skipping meals, the lesser participants’ intentions to engage in the behavior. The path

coefficient from intent to behavior was .25, P(.05 S p S .45) = .95, such that the less

48

intent one had to skip meals, fewer meals were actually skipped. The OLS estimates also
produced a path coefﬁcient from family subjective norm to intent of .02, P(-.23 S p S
.27) = .95, and from close ﬁiends subjective norm to intent of .03, P(-.21 S p S .27) = .95,

which were within sampling error of zero indicating that the data were not consistent with

the model.

The differences between predicted and obtained correlations for all unconstrained
bivariate relationships in the model were examined and one error (-.1 1, between family
subjective norm and behavior) differed substantially from what was expected ﬁ‘om
sampling error. The global test for goodness of ﬁt indicated that the data were consistent
with the model, {(3) = .75, p =.86, however, given that two of the path coefﬁcients were
essentially zero, the data were judged to be inconsistent with the model. Consequently, a

post hoc search for an alternative model that ﬁt the data was undertaken.

 

 

. .71 .25 Number Of
Attitude Intent Meals Skipped

per Day

 

 

 

 

 

 

 

 

 

 

 

 

Figure 10. Revised model with path coefﬁcients corrected for attenuation due to measurement error.

Results indicated that the data were consistent with a revised model (see Figure
10). Speciﬁcally, attitudes toward Skipping meals predicted one’s intent to Skip meals
[.71, P(.59 S p S .83) = .95] which subsequently predicted behavior [.25, P(.05 S p S .45)
= .95]. Both path coefﬁcients were ample and were in the direction predicted.
Additionally, an examination of the differences between predicted and obtained

correlations for all the unconstrained bivariate relationship revealed that the error (.06)

49

did not differ substantially from what was expected ﬁ'om sampling error. Further, the
global test for goodness of ﬁt indicated that the data were consistent with the model,
12(1) = .17, p = .68. Given that the path coefﬁcients were'ample, and that the model and
parameter estimates predicted accurately the unconstrained correlation, the data were
judged to be consistent with the alternative model. Again, because these data are
consistent with the model in Figure 10 the data are also consistent with the reverse model
and one should use caution when interpreting these results.

Summary of skipped meals results. The predicted relationships outlined by the
TRA were examined when a persuasive message inﬂuenced attitudes, subjective norms,
and when no message was present. The results indicated that 1) unlike the effect of the
persuasive message with the fruit and vegetable consumption topic, the message was
unsuccessful in inﬂuencing attitudes or subjective norms, and 2) again unlike the ﬁ'uit
and vegetable topic, TRA’S predictions were not consistent with the data from any group.
Interestingly, the only model to predict accurately the number of meals participants
skipped every day was a Simple causal string such that attitudes impacted intentions
which predicted behavior. These results, and those ﬁ'om the hit and vegetable

consumption topic, will be explored more fully in the discussion.

50

DISCUSSION

The current study sought to test an extension of Ajzen and Fishbein’s (1970,
1974) Theory of Reasoned Action; namely, it provided a test of the predicted
relationships outlined by the TRA within two systems of change: 1) when attitude change
and 2) subjective norm change resulted from exposure to a persuasive message. This test
of the TRA was important because attitudes and subjective norms rarely are held constant
and because people often are presented with persuasive messages aimed to change their
attitudes, subjective norms, and behavior. Thus, it was important to understand how such
messages affect the predictive power of the TRA.
Findings and Implications

The ﬁrst test examined ﬁ'uit and vegetable consumption. The results indicated that
when attitudes or subjective norm were affected by a persuasive message, the predictive
power of the TRA was diminished. Speciﬁcally, the data were not cOnsistent with the
TRA when 1) attitude was impacted by a persuasive message or 2) subjective norms were
inﬂuenced by a message. In contrast to these results, an examination of the static model
(when no message inﬂuenced attitude or subjective norm) revealed that fruit and
vegetable consumption was predicted by respondents’ intentions, and intentions were
predicted by attitude and subjective norms. Hence, the data were consistent with the TRA
when no change was induced by a persuasive message. These results are interesting
theoretically because they delineate the predictive scope of the TRA. Although the TRA
may predict accurately intentions and subsequent behavior when attitudes and subjective

norms are measured at a given point in time, the theory may be limited potentially insofar

51

as the TRA may not be able to predict accurately intentions and behaviors when attitudes
or subjective norms are influenced by a persuasive message.

The ﬁrst test of the TRA revealed that the theory may be limited such that it may
not predict accurately behaviors when attitudes or subjective norms are inﬂuenced by a
persuasive message. An additional test of the TRA was conducted, however, in order to
determine if the ﬁndings from the ﬁrst test could be replicated. Those results are
discussed next.

The second test of the TRA examined the number of meals people skip every day.
Two striking results emerged from this examination: 1) although the persuasive messages
had some success inﬂuencing attitudes and subjective norms toward ﬁ'uit and vegetable
consumption, the effect of these same messages on attitudes and subjective norms toward
Skipping meals was miniscule (within sampling error of zero), and 2) the data were not
consistent with the TRA even though attitudes and subjective norm Were not inﬂuenced
by a persuasive message. In fact, the best predictor of Skipping meals was a simple causal
string such that behaviors were predicted by intentions, and intent was predicted by
attitude, regardless of the experimental condition. As noted previously, however, the
results were also consistent with the reverse model (attitude predicted by intent and intent
predicted by attitude). This possibility of reverse directionality will be discussed in more
detail later.

Although the ﬁndings are interesting theoretically, a number of limitations exist in

the current study. These limitations are discussed subsequently.

52

Limitations

A number of limitations to the current study exist. First, although the impact of
the persuasive subjective norm message on respondents’ perceptions of their family
subjective norm (r = .34) and their close friends subjective norm (r = .19) toward fruit
and vegetable consumption were ample, it should be noted that neither of these
inductions was very powerful. This may simply be an artifact of the fruit and vegetable
topic. Perhaps one’s subjective norm regarding the increased consumption of fruit and
vegetables is not susceptible to much change. A more likely explanation, however, lies in
the difﬁculty of creating a persuasive message aimed at subjective norms.

The persuasive message used in the main experiment was the subject of three
different modiﬁcations and pilot tests before it met adequately the criteria for inclusion in
the study. Two difﬁculties were discovered when designing the subjective norm message:
1) a message cannot advocate what a person’s actual family or ﬁiend believes (e.g.,
“Your parents want you to eat more fruits and vegetables,” or “Your ﬁiends think that
eating up to nine servings of ﬁ'uit and vegetables a day will improve your health”), and 2)
a message must use general claims about farrrilies and ﬁiends (e.g., “A recent survey
found that parents with college-aged children want their children to incorporate more
fruits and vegetables into their diet”). A message that tells its readers that their families or
friends are in favor of a speciﬁc behavior is not credible or believable. After all, the
message source does not know what the reader’s family believes, and how could it
(especially if the message source is not a family member)? Indeed, in focus groups

conducted to aid message design during the pilot phase, messages with direct claims

53

about participants’ families and ﬁiends resulted in anger and refusal to take the message
seriously.

By using more general claims, a message may gain credibility and realism, but
may fall short of inﬂuencing one’s subjective norm to a great magnitude because 1) no
evidence is presented to indicate how similar the generalized families in the message are
to a reader’s family and 2) a reader may not believe such generalized claims (especially
when one’s parents do not actually say “eat more fruit” or the parents’ behaviors are not
consistent with the message claims). Therefore, although the current study employed
generalized claims which were judged to be credible, realistic, and which advocated the
desired perceptions of subjective norm, the strength of the subjective norm message was
less than optimal and therefore may have provided a weak test of the model.

Second, one Should note that the persuasive ﬁ'uit and vegetable subjective norm
message had a substantial effect on participants’ attitudes (r = .42), in addition to its
impact on family (r = .34) and close ﬁiends (r = .19) subjective norms. This may have
been due to the evidence used in the subjective norm message. Speciﬁcally, the messages
were designed such that they were equivocal in many aspects; hence, many of the claims
and pieces of evidence in the subjective norm message were identical or similar to that
used in the persuasive attitude message. Although the focus of the subjective norm
message was to persuade participants that their families and close ﬁiends want them to
eat more fruits and vegetables, the nature of the evidence and claims may have resulted in
the message impacting attitudes as well. One should not conclude, however, that scholars
and practitioners should abandon the use of persuasive subjective norm messages in favor

of messages aimed only at attitudes because 1) as noted previously, more research is

54

needed in order to perfect subjective norm messages and 2) if subjective norm messages
do, indeed, affect attitudes positively, the messages may result in the desired behavior
change. Therefore, it is important to continue to examine’persuasive subjective norm
messages.

Third, an examination of mean fruit and vegetable consumption by condition (see
Table 1 for reference) indicated that participants in the control group increased their
consumption by 1.41 servings per day, compared to participants in the attitude and
subjective norm message groups who increased their consumption by .97 and .94
servings per day, respectively. Therefore, participants in the control group Showed
evidence of increased consumption without exposure to an experimental induction. It is
possible that exposure to the 54 questions about fruit and vegetable consumption during
the Time One survey may have served as a persuasive message. Speciﬁcally, reading and
answering questions about one’s attitude, intent, and what one’s fMily and close ﬁ'iends
think about ﬁuit and vegetable consumption may have likely had an effect on behavior. It
is also possible that such questions may have resulted in participants recalling actual
messages from their family and friends. Additionally, because control group participants
may have reasonably concluded that the researcher was interested in increased fruit and
vegetable consumption after exposure to the questions at Time One, demand effects may
have been present such that respondents answered the consumption behavior consistent
with that they believed the researcher would expect from them. Therefore, mere exposure
to the instrument may have resulted in increased fruit and vegetable consumption, or this

change in behavior could have been due to Simple demand effects.

55

Fourth, the attitude-behavior correlations obtained in the current study diverge
greatly from that those found in previous research and, most notably, Kim and Hunter’s
(1993a) meta-analysis. For example, the largest correlation obtained between attitudes
toward ﬁ'uit and vegetable consumption and behavior (r = .05), and that between attitudes
toward skipping meals and behavior (r = .23), deviate from meta-analytic ﬁndings (mean
r = .79). Measurement error is one possible explanation for the current ﬁndings. Because
the behaviors were assessed with Single items, it is likely that unreliability attenuated the
attitude-behavior correlations. Although it is not possible to assess the reliability of a
single item, therefore making it necessary to assume perfect measurement when testing a
model, such single-item measures are highly unreliable and can attenuate relationships. If
one assumed the reliability for the Single-item behavioral measures to be .40, the
corrected fruit and vegetable (r = .08) and Skipping meals (r = .38) attitude-behavior
correlations would be larger. The corrected relationships, however, remain substantially
weaker than the previous meta-analytic ﬁndings.

Restriction in range is another statistical artifact that may be responsible for
attenuating the attitude-behavior relationships. Speciﬁcally, the distributions for both
fruit and vegetable consumption (kurtosis = 20.62) and number of meals Skipped per day
(kurtosis = 25.52) were highly leptokurtic. That is, little variance was present with
regards to participants’ behavior. Given that such restriction in range can attenuate
relationships, it is likely that this, in addition to the measurement error associated with the
use of single items, may account for the divergent ﬁndings of the current study.

Fifth, the relationships found with the Skipping meals topic (that attitudes predict

intentions and intent predicts behavior) may be consistent with the reverse mode] because

56

both models are equivalent statistically (see Kim & Hunter, 1993a, 1993b). Given that the
current study examined these relationships using cross-sectional data it is impossible to
distinguish the direction of the attitude-behavior relationship found with regards to
skipping meals. Although one cannot be sure with the data available here, it is plausible
that respondents’ meal—skipping behavior may predict intentions and subsequent attitudes
toward meal skipping.

Recent research from the National Center for Health (2003) indicated that nearly
forty percent of people in the United States skip one meal every day, and that meal is
breakfast typically. People may have a routine, or habit, of skipping breakfast in order to
get to work or school on time, because they don’t feel hungry in the morning, or because
at a young age they Skipped breakfast before school and have continued that habit into
adulthood (N CH, 2003). This behavior may result ultimately in attitudes that are
favorable toward Skipping meals (e. g., “I think Skipping breakfast is okay because I Skip
it almost every day”) or unfavorable toward the behavior (e. g., “I eat breakfast every
morning so I think that eating breakfast is a good idea”). Because one cannot be certain
about the direction of the relationship found in the current investigation, one should be
cautious when interpreting these results.

In sum, the present investigation has a number of limitations. Namely, the current
study had problems related to 1) the persuasive subjective norm message induction, 2)
possible repeated exposure and demand effects, 3) the weak attitude-behavior
relationship found with both topics, and 4) the inability to decipher the directionality of

the attitude-behavior relationship with regards to the skipping meals topic. Given these

57

limitations, scholars should address these concerns in future research. Ideas for future
research are discussed next.
Directions for Future Research

Given that previous research has shown that one’s family can have an effect on
eating behaviors (Haworth-Hoeppner, 2000; Humphrey, 1987; Levine etal., 1994; Schur
et al., 2000), and that one’s peers and close ﬁiends also impact eating behaviors
(Hausenblas & Carron, 1998; Levine, Smolak, Moodey, et al., 1994; Paxton et al., 1999),
future research should focus on how to design an effective message to inﬂuence people’s
perceptions of subjective norms. Although the persuasive subjective norm message met
the criteria necessary for inclusion in the current study, future research should focus on
how to improve the induction. Speciﬁcally, future messages should be sure to include
claims that make the message seem less divergent ﬁom one’s experience in order to
strengthen the subjective norm induction (e. g., “Although your parents may not tell you,
and even though they may not eat enough fruit and vegetables themselves, your parents
would still want you to increase your consumption to ensure you are healthy”).

Additionally, one is likely inﬂuenced by people other than family and ﬁiends
(Sanftner et al., 1996). People are inﬂuenced by their romantic partners and co-workers,
for example, and future research should examine l) the degree to which these subjective
norms predict intentions and subsequent behavior and 2) message strategies to inﬂuence
such subjective norms. For example, one’s desire to act consistent with a romantic
partner’s beliefs may inﬂuence intentions and behavior to a greater magnitude than
ﬁiends or family members. Furthermore, this differential effect may be bound by the

speciﬁc behavior being examined. Therefore, future research should also examine which

58

subjective norms (e.g., fanrily, ﬁiends, partners, teachers) best predict different
behaviors.

Another line of research should aim to examine the degree to which the results of
the current study generalize to other topics. Future research Should test change and the
TRA model by focusing on other health and non-health topics (e.g., students’ time spent
studying every week, condom use, exercise behavior, workplace performance). Such tests
will help to determine the boundaries of the TRA’S predictive abilities and will provide
further evidence of the effectiveness (or ineffectiveness) of the TRA when attitudes and
subjective norms are inﬂuenced by persuasive messages.

Finally, scholars Should focus on providing a longitudinal test of the TRA. First,
because two models are plausible with regards to the Skipping meals topic, ﬁrture
research should examine these relationships longitudinally in order to decipher the
direction of the attitude-behavior relationship. A longitudinal investigation could address
the problems inherent in deciphering directionality of the attitude-behavior relationship.
Speciﬁcally, as Kim and Hunter (1993a, 1993b) noted, although it is commonly believed
that attitudes precede behavior, the reverse is also plausible. The TRA also allows for this
possibility; when the correlation between attitude and subjective norm is equal to the
product of the intention-attitude and intention-subjective norm correlations, the TRA
cannot distinguish the direction of the attitude-behavior relationship. For example, given
the data presented in Table 9 the TRA model is consistent with the data (see Figure 11a).
Additionally, the reverse TRA model is consistent with these same data (see Figure 11b).
Therefore, given certain data, the TRA may not be able to assess the directionality of the

attitude-behavior relationship.

59

Table 9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hypothetical T RA Data
1 2 . 3 4
1. Attitude
2. Subjective Norm .25
3. Behavioral Intent .50 .50
4. Behavior .25 .25 .50
\ .50
Behavioral _ .
Subjective /
\- Norm
Figure 1 I a. Condition under which the TRA cannot assess directionality.
‘ Attitude
.50
Behavior '50 Behavioral
Intentions
.50 Subjective
Norm
Figure 11b. The reverse TRA.

 

 

 

Note. If the obtained correlation between attitude and subjective norm is equal to the product of the
intention-attitude and intention-subjective norm correlations (here, r = .25) behaviors are equally as
predictive of attitudes as attitudes are of behavior.

Existing cross-lagged research has done little, however, to make a ﬁrm
determination of the directionality of the relationship. Although this body of research
appears to be consistent with the TRA such that attitudes and subjective norm predict

intentions and subsequent behavior (Bentler & Speckart, 1981; Kahle & Berman, 1979),

60

some evidence has shown that behaviors can have causal priority over attitudes and
subjective norms (Bentler & Speckart, 1981). Unfortunately, these ﬁndings are
downplayed and little has been done to address the degree to which the reverse TRA may
be true and the conditions under which this may occur.

Because the current investigation is limited insofar as it provides a test of the
TRA when persuasive messages impact components of the model but examines those
effects at only one point in time, this cross-sectional snapshot should not be the ﬁnal
investigation of the TRA models tested here. Future endeavors to understand the effect of
persuasive messages on the TRA’S predictions should also undertake a longitudinal
approach.

As noted previously, a few scholars have examined the TRA over time, but the
existing cross-lagged research has only tested the static model (e.g., Bentler & Speckart,
1981; Kahle & Berman, 1979). That is to say, existing attitudes, subjective norm, intent,
and behaviors were measured at more than one point in time, but persuasive message
inductions were not used in order to test systematically the longitudinal effect of attitude
or subjective norm change on intentions and subsequent behavior. A single-exposure,
persuasive message may inﬂuence attitudes or subjective norms immediately (as found in
the present study), but the long term implication of this inﬂuence on intentions and
behaviors, and the predictive power of the TRA, is not clear. Therefore, longitudinal
research is needed in order to determine if the TRA’s predictions are accurate over time.
It is possible that, consistent with the current investigation, the predictions are not
accurate at one point in time, but over time the theory may have increased predictive

power. This is an empirical question that should be addressed in future research.

61

Conclusion

The current study addressed core functional questions about inﬂuence and the
TRA. As noted previously, the answers to these questions clearly hold theoretical and
practical import. Although the ability of the present investigation to answer these
questions was limited insofar that message design was less than optimal in one condition
and the current analyses used only cross-sectional data, the ﬁndings reported here begin
to outline potential limitations with regards to the predictive power of the TRA. The
future research suggested here will address these limitations and provide an important,

expanded test of the TRA.

62

Footnotes

1The data collected for the current study were taken from a larger program of
research conducted over the course of approximately one month. Although all analyses
reported for the current study come from Survey 1 and Survey 2, only those respondents
who participated fully in all three surveys were included in the ﬁnal analyses. Therefore,
all participant information reported here refers to respondents who completed all three
surveys.

2Given that Fishbein and Ajzen (197 5) claim that attitudes are the sum of one’s
evaluative beliefs, they make an assumption that each of these beliefs is related linearly.
Therefore, Conﬁrmatory Factor Analysis was employed to test the attitude belief and
both subjective norm belief measurement models.

3 One should be careful to note that because the measurement used in the current
study differs from that advocated by Fishbein and Ajzen (1975) it may be a limitation

when comparing these results to other TRA studies.

63

APPENDICES

64

APPENDIX A

Pilot Behavioral Questions

1. On average, how many hours per week do you exercise?

2. On average, how many hours per week do you work?

3. On a scale of l to 10, how physically strenuous is your job?

not strenuous l 2 3 4 5 6 7 8 9 10 very strenuous

4. Do you smoke cigarettes? Yes No

5. On average, how many cigarettes do you smoke per day?

6. On average, how many servings of fruit do you eat per day (one serving ~ 1 cup of fruit)? __ ‘

7. On average, how many servings of vegetables do you eat per day, not including starchy vegetables like
potatoes and corn (one serving ~ 1 cup of vegetables)? "‘

8. On average, how many grams of fat do you eat per day?
9. Are you currently in a romantic relationship? Yes No

10. For how many months have you been with your current romantic partner? (For example, if you have
been with your partner for 2 years, you would answer 24).

l 1. In an average week, how many times do you skip meals?
12. On an average day, how many meals do you eat?

13. On an average day, how many snacks do you eat?

14. On an average day, how many meals do you skip? __ *

15. Do you live in an on-campus dorm? Yes No
16. Do you live in an on-campus apartment? Yes No
17. Do you live with your parent(s)? Yes No
18. Do you live off campus (separate from your parents)? Yes No
19. Do you play intercollegiate sports? Yes No

Note. " indicates questions about the behaviors of interest for the current study. All other items were ﬁller
items.

65

APPENDIX B

Message Realism Scale
1. This message was believable.
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree
2. This message was fake.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree
3. This message was exaggerated.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree
4. This message was plausible.
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree
5. This message was realistic.
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree

66

APPENDIX C

Message Credibility Scale

I. The information presented in the message is credible.
Strongly Disagree -3 -2 -1 0 +1

2. The message is reliable.
Strongly Disagree -3 -2 -1 0 +1

3. I think that the message is a believable one.
Strongly Disagree -3 -2 -1 0 +1

4. The information presented in the message is trustworthy.
Strongly Disagree -3 -2 -l 0 +1

67

+2

+2

+2

+2

+3

+3

+3

+3

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

APPENDIX D

Message Position Scale (Fruit and Vegetables Topic)

1. This message is in favor of eating up to 9 servings of fruit and vegetables a day.
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree

2. The message portrayed eating up to 9 servings of ﬁuit and vegetables a day as a positive thing.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

3. The message indicates that eating up to 9 servings of fruit and vegetables a day is bad. R
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree

68

APPENDIX E

Message Position Scale (Skipping Meals Topic)

1. This message is in favor of skipping meals.

Strongly Disagree -3 -2 -1 0 +1

2. The message portrayed Skipping meals as a positive thing.

Strongly Disagree -3 -2 -1 0 +1

3. The message indicates that Skipping meals is bad. R

Strongly Disagree -3 -2 -1 0 +1

69

+2

+2

+2

+3

+3

+3

Strongly Agree

Strongly Agree

Strongly Agree

APPENDIX F

Perceptions of Advocated Familial Subjective Norm (Fruit and Vegetables Topic)

1. According to this message, my family thinks that I should eat up to 9 servings of fruits and vegetables
every day.

Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree

2. This message indicates that it is important to my family that I eat up to 9 servings of fruits and
vegetables every day.

Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

3. According to this message, my family believes that it is important for me to eat up to 9 servings of

fruits and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

70

APPENDIX G

Perceptions of Advocated Familial Subjective Norm (Skipping Meals Topic)

1. According to this message, my family thinks that I should skip meals.
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree

2. This message indicates that it is important to my family that I skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

3. According to this message, my family believes that it is important for me to skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

71

APPENDIX H

Perceptions of Advocated Close Friends Subjective Norm (Fruit and Vegetable Topic)

4. According to this message, my close friends think that I should eat up to 9 servings of fruits and
vegetables every day.

Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

5. This message indicates that it is important to my close friends that I eat up to 9 servings of fruits and
vegetables every day.

Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree
6. According to this message, my close friends believe that it is important for me to eat up to 9 servings

of fruits and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

72

APPENDIX I

Perceptions of Advocated Close Friends Subjective Norm (Skipping Meals Topic)

4. According to this message, my close friends think that I should skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

5. This message indicates that it is important to my close friends that I skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

6. According to this message, my close friends believe that it is important for me to skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

73

APPENDIX J

Attitude Scale (Fruit and Vegetables Topic)

1. I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree -3 -2 -l 0 +1 +2 +3

2. I believe that I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3

3. It is important for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3

4. Every day I should eat up to 9 servings of ﬁ'uit and vegetables.
Strongly Disagree -3 -2 -l 0 +1 +2 +3

5. It is not crucial that I eat up to 9 servings of fruit and vegetables every day. R *
Strongly Disagree -3 -2 -1 0 +1 +2 +3

Note. " indicates item was deleted from measurement model for main experiment.

74

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

APPENDIX K
Attitude Scale (Skipping Meals Topic)

1. I support Skipping meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

2. I Should skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

3. I think that skipping meals is a good idea.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

4. I am in favor of skipping meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

5. It is not a good idea for me to Skip meals. R “
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

Note. * indicates item was deleted from measurement model for main experiment.

75

APPENDIX L

Perceptions of Familial Subjective Norm (Fruit and Vegetables Topic)

1. My family thinks that I should eat up to 9 servings of ﬁuit and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

2. It is important to my family that I eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

3. My family believes that it is important for me to eat up to 9 servings of ﬁ'uit and vegetables every day.
Strongly Disagree -3 -2 -l 0 +1 +2 +3 Strongly Agree

4. According to my family, every day I should eat up to 9 servings of fruit and vegetables.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

76

APPENDIX M

Perceptions of Familial Subjective Norm (Skipping Meals Topic)

1. My family thinks that it is acceptable Skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2 +3

2. It is okay with my family to skip meals.
Strongly Disagree -3 -2 -l 0 +1 +2 +3

3. According to my family, it is acceptable to skip meals.
Strongly Disagree -3 -2 -l 0 +1 +2 +3

4. It is not acceptable to my family to skip meals. R "'
Strongly Disagree -3 -2 -1 0 +1 +2 +3

5. My family believes that it is acceptable to skip meals.
Strongly Disagree -3 -2 -l 0 +1 +2 +3

Note. " indicates item was deleted from measurement model for main experiment.

77

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

APPENDIX N

Perceptions of Close Friends Subjective Norm (Fruit and Vegetables Topic)

1. My close ﬁiends think that I should eat up to 9 servings of fruits and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

2. It is important to my close friends that I eat up to 9 servings of fruits and vegetables every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

3. My close ﬁiends believe that it is important for me to eat up to 9 servings of fruits and vegetables
every day.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

4. According to my close ﬁiends, every day I should eat up to 9 servings of fruits and vegetables.
Strongly Disagree -3 -2 -1 0 +1 +2 +3 Strongly Agree

78

APPENDIX 0

Perceptions of Close Friends Subjective Norm (Skipping Meals Topic)

1. My close ﬁiends think that it is acceptable skip meals.
Strongly Disagree -3 -2 -l 0 +1 +2

2. It is okay with my close friends to Skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2

3. According to my close friends, it is acceptable to skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2

4. It is not acceptable to my close friends to skip meals. R ‘
Strongly Disagree -3 -2 -l 0 +1 +2

5. My close friends believe that it is acceptable to Skip meals.
Strongly Disagree -3 -2 -1 0 +1 +2

Note. " indicates item was deleted from measurement model for main experiment.

79

+3

+3

+3

+3

+3

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

APPENDIX P
Attitude Message
Adopting a Healthy Lifestyle
The Importance of Every 4 Hours

Often times it is all too easy for us to skip breakfast or lunch because we simply do not have the time or
money to stop and eat. Many people also skip meals to help them control their weight. Medical research
has shown that when people Skip meals it hurts them in many ways. Skipping meals slows down your
metabolism which makes your body hang on to excess fat and can affect hormone levels in both men and
women (for example, insulin) and lead to severe chronic illnesses like diabetes.

The National Institute of Health and the National Center for Health suggest that people should eat every 4
hours. Indeed, research has shown that you shouldn’t go more than 4 hours without eating. When you eat
every 4 hours you maintain a healthy metabolism that will help you get and stay ﬁt, normalize your
hormone levels, and help you prevent chronic illness. Always remember to eat every 4 hours and don ’t be
tempted to skip any meals!

The Importance of 9 a Day

Research from the Center for Disease Control and the National Institute of Health shows that 74%
Americans, and 85% of college students, do not get enough fruits and vegetables in their diets. Did you
know that every person needs to eat at least 5, and up to 9 servings of fruit and vegetables every day?
What constitutes a serving? One serving of fruit is approximately 1 cup of fresh fruit, and one serving of
vegetables is approximately 1 cup of raw vegetables (not including starchy vegetables like potatoes).

Research has shown that people who consume at least 4-5 servings of fruit and 4-5 servings of vegetables
every day as part of a balanced diet are at less risk for diseases such as diabetes, heart disease, cancer, and
high blood pressure. In addition to preventing diseases, eating up to 9 servings of fruit and vegetables every
day can help improve your overall health because they are rich in vitamins and antioxidants. So, make sure
that you always eat up to 9 servings of fruit and vegetables every day.

80

APPENDD( Q
Subjective Norm Message
Those Closest to You Want You to Adopt a" Healthy Lifestyle

The Importance of Every 4 Hours

Often times it is all too easy to skip breakfast or lunch because we simply do not have the time or money to
stop and eat. Many people also Skip meals to help them control their weight. Recent studies have shown
that 92% of family physicians believe that skipping meals can hurt you in many ways. Skipping meals
slows down your metabolism which makes your body hang on to excess fat and can affect hormone levels
in both men and women (for example, insulin) and lead to severe chronic illnesses like diabetes. To avoid
such health complications the National Institute of Health and the National Center for Health suggest that
people should eat every 4 hours.

In fact, a recent survey found that 83% of parents with college-age children want their children to eat every
4 hours, and do not want their children to skip meals. Also, studies done with college students indicate
76% think it is important to eat every 4 hours because when you maintain a healthy metabolism it will help
you get and stay ﬁt, normalize your hormone levels, and help you prevent chronic illness. We know that
our family and ﬁiends love us, care about us, and want us to be healthy. So, remember that most families
and college students are in favor of eating every 4 hours and don ’t want people to be tempted to skip
meals!

The Importance of 9 a Day

Research from the Center for Disease Control and the National Institute of Health shows that 74%
Americans, and 85% of college students, do not get enough ﬁ'uits and vegetables in their diets. Did you
know that every person needs to eat at least 5, and up to 9 servings of fruit and vegetables every day? What
constitutes a serving? One serving of fruit is approximame 1 cup of fresh ﬁuit, and one serving of
vegetables is approximately 1 cup of raw vegetables (not including starchy vegetables like potatoes). In a
recent survey, 96% of parents with college-age children said they were worried that their children were not
eating enough fruits and vegetables. In fact, most parents were in favor of eating up to 9 servings of fruit
and vegetables a day.

According to 93% of family physicians, people who consume at least 4-5 servings of ﬁuit and 4-5 servings
of vegetables every day as part of a balanced diet are at less risk for diseases such as diabetes, heart disease,
cancer, and high blood pressure. Research has also shown that, in addition to the many families who
support increased fruit and vegetable consumption, 80% of college students think it is important that their
friends eat 9 servings of fruit and vegetables a day. In addition to preventing diseases, most family
physicians assert that eating up to 9 servings of fruit and vegetables every day can help improve your
overall health because they are rich in vitamins and antioxidants. Again, remember that our ﬁiends and
family care about us and want us to live long and healthy lives. So, remember that most families and
college students believe that eating up to 9 servings of fruit and vegetables every day is important!

You should listen to your family and friends and adopt a healthy lifestyle by eating
more fruits and vegetables and by avoiding the opportunity to skip meals.

81

APPENDIX R

Time One Survey

(Participants ﬁrst logged on to the Introduction and Consent Form Page)

Student Life Survey

Thank you for your interest in the Student Life Survey. You will receive full credit for
your participation once you have completed all three surveys. Before you begin the
survey please read the consent form below.

Consent Form

The purpose of this study is to examine some of the issues surrounding student
activities, lifestyle, eating habits, and students’ use of university facilities. Full
participation in the entire study (all three surveys) will take 90 minutes or fewer.
Although participation in this study is not expected to produce discomfort or stress,
please note that you may refuse to answer certain questions or withdraw from the study at
any time without penalty. The information collected in this study will be treated with
strict conﬁdence. Your privacy will be protected to the maximum extent allowable by
law. Your responses are totally anonymous and will not be linked to your identity in any
way. If you have any questions about your role and rights as a participant in research, feel
free to contact Dr. Ashir Kumar, the Chairperson of the University Committee on
Research Involving Human Subjects at (517) 355-2180. The experimenter can answer
any questions you have about the study to help you choose whether or not to participate.
Contact Lisa Lindsey at (517) 432-1286 if you have any further questions or concerns
regarding this study.

Thank you,
Lisa Lindsey

By clicking the AGREE button you indicate your voluntary participation in this study.

 

 

AGREE I DO NOT AGREE

 

 

 

 

 

 

(If Ps clicked on AGREE they were directed to the survey. If PS clicked on I DO NOT AGREE they were
directed to a page thanking them for their time)

82

(If students clicked the I DO NOT AGREE button)

Thank You for Your Time

Only those people who agree to participate in the study
voluntarily may complete the survey.

83

(If students clicked the AGREE button they were directed to this survey)
T hank you for volunteering to participate in this survey.
General Instructions
In the questionnaire you are about to complete you will be asked to answer questions related to eating
habits, lifestyle, and your use of university facilities. In order to indicate your response to each question,

please click on the correct response or type your answer in the blank provided.

Please be sure to read all instructions very care/idly and to provide your honest answers. Remember, your
answers are completely anonymous and conﬁdential.

First, we ’d like you to answer some questions about yourselﬂ your experience with extra-curricular
activities, and your experience with services oﬂered on campus.

1. Do you live in an on-campus dorm? Yes No
2. Do you live with your parent(s)? Yes No
3. Do you live in an apartment (separate from your parents)? Yes No
4. Do you play intercollegiate sports? Yes No

5. In how many credits are you enrolled this semester?
6. On average, how many hours per night do you study for classes?

7. Did you know that Olin Health Center has a nutritionist available to work with students?
Yes No ‘

8. Have you ever met with the nutritionist at Olin Health Center?
Yes No

9. Did you know that Olin Health Center has a Life:Rx program that provides ﬁtness evaluations that test
strength, ﬂexibility, lung capacity, cardiovascular ﬁtness, and cholesterol?
Yes No

10. Have you ever had a Lifesz ﬁtness evaluation?
Yes No

l 1. On average, how many hours per week do you exercise?

12. On average, how many hours per week do you work?

13. On average, how many cigarettes do you smoke per day? __

14. On average, how many servings of fruit do you eat per day (one serving ~ 1 cup of fruit)? __

15. On average, how many servings of vegetables do you eat per day, not including starchy vegetables like
potatoes and corn (one serving ~ 1 cup of vegetables)?

16. In an average week, how many times do you skip meals?
17. On an average day, how many meals do you eat?

84

18. On an average day, how many snacks do you eat?

19. On an average day, how many meals do you skip?

 

Now we would like you to read the following message very carefully.
Please read the entire message because we will ask you questions about it later.

(the message to which participants were assigned would appear here)

(the contents of this box did not appear for participants in the control group)

 

 

 

Now we would like to ask you a series of questions about your eating habits.

Each of the questions below will ask you to rate your answer on a scale. The scale is from -3 to +3 and each
number represents the degree to which you agree or disagree with a statement.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

Please click on the answer that bat describes your opinion. For example, if you were to rate the weather
in Michigan on such a scale, the scale should be interpreted as follows:

Winter weather in Michigan is cold.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

If you snongly agree that the winter weather in Michigan is cold, then you would click on your answer as
follows:

Winter weather in Michigan is cold.
Strongly Disagree 0 O O O O O @ Strongly Agree
-3 -2 -l 0 +1 +2 +3

If you are neutral or undecided, then you neither agree nor disagree with the statement “Winter weather in
Michigan is cold,” and you would click on your answer as follows:

Winter weather in Michigan is cold.
Strongly Disagree 0 O 0 © 0 O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

If you moderately disagree with the statement “Winter weather in Michigan is cold,” then you would click
on your answer as follows:

Winter weather in Michigan is cold.
Strongly Disagree 0 © 0 O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

Again, please read each statement carefully and indicate the degree to which you agree or disagree with
them. Please answer honestly.

What are your opinions about fruit and vegetables?

85

l. I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

2. I believe that I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

3. It is important for me to eat up to 9 servings of ﬁrm and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. Every day I should eat up to 9 servings of fruit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

5. Eating up to 9 servings of fruit and vegetables a day will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

6. Eating up to 9 servings of fruit and vegetables a day will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. Eating up to 9 servings of fruit and vegetables a day will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

8. Eating up to 9 servings of fruit and vegetables a day will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O ' Strongly Agree
-3 -2 -1 0 +1 +2 +3

9. Eating up to 9 servings of fruit and vegetables a day will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

10. Eating up to 9 servings of fruit and vegetables 3 day will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

11. Eating up to 9 servings of fruit and vegetables a day will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

12. Eating up to 9 servings of fruit and vegetables a day will be expensive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

13. Eating up to 9 servings of fruit and vegetables a day will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

14. I intend to incorporate up to 9 servings of fruit and vegetables into my diet every day.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

86

15. I plan to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

16. Every day I aim to eat up to 9 servings of hit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

17. I expect that I will eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

18. I do not intend to eat up to 9 servings of ﬁuit and vegetables every day. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

19. Eating up to 9 servings of fruit and vegetables every day is easy for me.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

20. It is difﬁcult for me to eat up to 9 servings of fruit and vegetables every day. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

21. It is simple for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

22. It is feasible for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

23. I have control over the amount of fruit and vegetables I eat every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 —l 0 +1 +2 +3

24. I cannot control the amount of fruit and vegetables I eat every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

25. I am in charge of how many servings of fruit and vegetables I eat every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

26. I have the power to determine how many servings of fruit and vegetables I eat every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3
What do you think about skipping meals?
27. I support skipping meals.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

87

28. I Should Skip meals.
Strongly Disagree 0 O O
-3 -2 -1

OO
O

29. I think that skipping meals is a good idea.
Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

30. I am in favor of skipping meals.
Strongly Disagree 0 O O O O
-3 -2 -l 0

31. It is not a good idea for me to skip meals. R
Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

32. I believe that skipping meals will help me improve my appearance.
0 O O

Strongly Disagree 0 O O
-3 -2 -1 0 +1

33. I think that skipping meals will help me feel good about myself
0 O O O

Strongly Disagree O
-3 -2 -1 0 +1

34. Skipping meals will help me improve my overall health.

Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

35. I believe that skipping meals will help me reduce my risk of disease.

Strongly Disagree 0 O O O O
-3 -2 -1 0 + 1

36. I think that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O
-3 -2 -l 0 +1

37. Skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

38. I think that skipping meals will help me prevent cancer.

Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

39. Skipping meals will be expensive.

Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

40. Skipping meals will be convenient.

Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

41. I intend to skip meals.

Strongly Disagree 0 O O O
-3 -2 -1 0 +1

88

+2

0
+2

0
+2

0
+2

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

' Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

42. I plan to skip meals.

Strongly Disagree 0 O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

43. I aim to skip meals.

Strongly Disagree 0 O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

44. I expect that I will skip meals.

Strongly Disagree 0 O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

45. I do not intend to skip meals. R

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

46. Skipping meals is easy for me.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

47. It is difficult for me to skip meals. R

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

48. It is simple for me to skip meals.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

49. It is feasible for me to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0

50. I have control over the number of meals I skip.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

51. I cannot control the number of meals I skip.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

52. I am in charge of the number of meals I skip.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

53. l have the power to determine the number of meals I skip.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

How important are each of the following to you?

1. Improving my appearance.
Not Important At All 0 O O O O O O 0 Extremely Important
4

O l 2 3 5 6 7

89

2. Feeling good about myself.
Not Important At All 0 O O O
0 l

N
w
M

3. Improving my overall health.
Not Important At All 0 O O O
O l

M
Le)
LII
\l

4. Reducing my risk of disease.
Not Important At All 0 O O O
O l

N
L»
M

5. Losing weight.
Not Important At All 0 O O O
O 1 2

DJ
Us

6. Increasing my attractiveness.
Not Important At All 0 O O O

O l 2 3 5 6

7. Preventing cancer.

Not Important At All 0 O O O O O O O
O 1 2 3 4 5 6 7

8. Saving money.

Not Important At All 0 O O O O O O O
O l 2 3 4 5 6 7

9. Convenience.
Not Important At A11 0 O O O O O O
O 1 2 3 4

\IO

10. Eating up to 9 servings of fruits and vegetables a day.
Not Important At All 0 O O O O O O O
0 l 2 3 4 6 7

M

11. Skipping meals.
Not Important At All 0 O O O O O O O
O l 2 3 4 5 6 7

12. Generally speaking, doing what people important to me think I should do.

Not Important At All 0 O O O O O O O
O ‘ l 2 3 4 5 6 7

13. Generally speaking, doing what my family thinks I should do.
Not Important At All 0 O O O O O O O
O 1 2 3 4 5 6 7

14. Generally speaking, doing what my close friends think I should do.
Not Important At All 0 O O O O O O O
O l 2 3 4 5 6 7

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

When you answer the next set of questions, think about your family.

What are your family’s opinions about fruit and vegetable consumption?

90

1. My family thinks that I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

1. It is important to my family that I eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O ' O Strongly Agree
-3 -2 -1 0 +1 +2 +3

2. My family believes that it is important for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

3. According to my family, every day I should eat up to 9 servings of fruit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3
4. My family believes that eating up to 9 servings of fruit and vegetables a day will help me improve my
appearance.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3

5. My family thinks that eating up to 9 servings of fruit and vegetables a day will help me feel good about
myself
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3
6. According to my family, eating up to 9 servings of fruit and vegetables a day will help me improve my
overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3 '
7. My family believes that eating up to 9 servings of ﬁuit and vegetables a day will help me reduce my
risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3

8. My family thinks that eating up to 9 servings of fruit and vegetables 3 day will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3
9. According to my family, eating up to 9 servings of fruit and vegetables a day will help me be more
attractive.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3

10. My family thinks that eating up to 9 servings of fruit and vegetables a day will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

1 1. My family thinks that eating up to 9 servings of fruit and vegetables a day will be expensive.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

91

12. My family believes that eating up to 9 servings of fruit and vegetables a day will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

What are your family’s opinions about skipping meals?

1. My family thinks that it is acceptable skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. It is okay with my family to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. According to my family, it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. It is not acceptable to my family to skip meals. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. My family believes that it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

6. My family believes that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

7. My family thinks that skipping meals help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

8. According to my family, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

9. My family belive that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

10. My family thinks that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

11. According to my family, skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

12. My family thinks that skipping meals will help me prevent cancer.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 - l O + l +2 +3

92

13. My family thinks that skipping meals will be expensive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

14. My family believes that skipping meals will be convenient.
Strongly Disagree 0 O O O O O ' O Strongly Agree
-3 -2 -l 0 +1 +2 +3

When you answer the next set of questions, think about your close friends.
What do your close friends think about fruits and vegetables?

1. My close friends think that I should eat up to 9 servings of fruits and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

2. It is important to my close friends that I eat up to 9 servings of fruits and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

3. My close friends believe that it is important for me to eat up to 9 servings of fruits and vegetables
every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. According to my close friends, every day I should eat up to 9 servings of fruits and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. My close friends believe that eating up to 9 servings of fruits and vegetables a- day will help me
improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3
6. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me feel good
about myself.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. According to my close friends, eating up to 9 servings of fruits and vegetables a day will help me
improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3

8. My close friends believe that eating up to 9 servings of fruits and vegetables a day will help me reduce
my risk of disease.

Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3

9. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me lose
weight.

Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3

93

10. According to my close friends, eating up to 9 servings of fruits and vegetables a day will help me be
more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

11. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me prevent
cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

12. My close friends think that eating up to 9 servings of fruit and vegetables a day will be expensive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

13. My close friends believe that eating up to 9 servings of fruit and vegetables a day will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

What do your close friends think about skipping meals?

14. My close friends think that it is acceptable skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

15. It is okay with my close friends to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

16. According to my close friends, it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

17. It is not acceptable to my close friends to skip meals. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

18. My close friends believe that it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

19. My close friends believe that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

20. My close friends think that skipping meals will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

21. According to my close friends, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

22. My close ﬁiends believe that skipping meals will help me reduce my risk of disease.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

94

23. My close friends think that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

24. According to my close friends, skipping meals will help me be more attractive.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

25. My close friends think that skipping meals will help me prevent cancer.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

26. My close friends believe that skipping meals will be expensive.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

27. My close friends believe that skipping meals will be convenient.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

Now we would like to ask you a few final questions about yourself
1. How old are you?

2. What is your sex? Female Male

3. What is your year in school?

Freshman Sophomore Junior Senior Other

4. Are you an international student?

Yes No

5. Which of the following best describes your ethnic or racial background?
White/Caucasion Black/African American/African
Chicano/Latino/l-Iispanic Native American
Asian/Asian American Paciﬁc Islander
Indian (from India) Middle Eastern Descent

 

Please click on the SUBMIT button to submit your survey. SUBMIT

 

 

 

95

(After participants submitted Survey #1)
Thank you for completing the ﬁrst survey!
In 7-10 days you will receive an email informing you that the second survey is online.

If you are interested in learning more about the facilities available to
you on campus please click on the link below.

http://www.olin.msu.edu/navigation.php?id=99

(End Survey #1)

96

APPENDIX S

Time Two Survey

Thank you for returning to complete the second Student Life Survey.

Although this second survey is not as long as the ﬁrst one, it is still formatted such that you will just scroll
through all of the questions. Just keep scrolling down to answer all of the questions. In order to indicate

your response to each question, please click on the correct response or type your answer in the blank
provided.

Please be sure to read all instructions very carefully and to provide your honest answer. Remember, your
answers are completely anonymous and conﬁdential.

First, we’d like to ask you a few follow-up questions regarding your eating habits and use of
university facilities.

I.

10.

ll.

12.

During the past week, how many servings of fruit did you eat per day on average (one serving ~ 1
cup)?

During the past week, how many servings of vegetables did you eat per day on average (one serving ~
1 cup)?

During the past week, how many days did you exercise? _

During the past week, how many times did you skip a meal? _

During the past week, how many meals did you skip on an average day? _
During the past week, how many cigarettes did you smoke on an average day?

During the past week, did you make an appointment to see a nutritionist?
Yes No

During the past week, did you make a Life:Rx appointment?
Yes No

During the past week, did you actually see a nutritionist?
Yes No

During the past week, did you actually have a Lifesz evaluation?
Yes No

During the past week, did you see or hear any messages (other than those related to this survey) about
eating up to 9 servings of fruits and vegetables 3 day?

Yes No
During the past week, did you see or hear any messages (other than those related to this survey) about
skipping meals?

Yes No

Now we would like to ask you a series of questions about your eating Min.

97

Each of the questions below will ask you to rate your answer on a scale. The scale is from -3 to +3 and each
number represents the degree to which you agree or disagree with a statement.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2_ +3

Please click on the answer that best describes your opinion. Again, please read each statement careﬁrlly
and indicate the degree to which you agree or disagree with them. Please answer honestly.

What are your opinions about fruit and vegetables?

1. I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. I believe that I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

3. It is important for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. Every day I should eat up to 9 servings of fruit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

5. Eating up to 9 servings of fruit and vegetables a day will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3 ‘

6. Eating up to 9 servings of fruit and vegetables a day will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. Eating up to 9 servings of fruit and vegetables a day will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

8. Eating up to 9 servings of fruit and vegetables a day will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

9. Eating up to 9 servings of fruit and vegetables a day will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

10. Eating up to 9 servings of fruit and vegetables a day will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

11. Eating up to 9 servings of mu and vegetables a day will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

98

12. Eating up to 9 servings of fruit and vegetables a day will be expensive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 --2 -1 0 +1 +2 +3

13. Eating up to 9 servings of fruit and vegetables a day will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

14. I intend to incorporate up to 9 servings of fruit and vegetables into my diet every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

15. I plan to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

16. Every day I aim to eat up to 9 servings of fruit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

17. I expect that I will eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

18. I do not intend to eat up to 9 servings of ﬁuit and vegetables every day. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

19. Eating up to 9 servings of fruit and vegetables every day is easy for me. _
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

20. It is difﬁcult for me to eat up to 9 servings of fruit and vegetables every day. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

21. It is simple for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

22. It is feasible for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

23. I have control over the amount of fruit and vegetables I eat every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

24. I cannot control the amount of fruit and vegetables I eat every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

25. I am in charge of how many servings of fruit and vegetables I eat every day.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

99

26. I have the power to determine how many servings of fruit and vegetables I eat every day.

Strongly Disagree 0 O O O O
-3 -2 -l 0 +1

0
+2

0
+3

What do you think about skipping meals?

27 . I support skipping meals.
Strongly Disagree 0 O O O O
-3 -2 -l 0

28. I should skip meals.

Strongly Disagree 0 O O O
-3 -2 -1 0 +1

29. I think that skipping meals is a good idea.

Strongly Disagree 0 O O O O
-3 -2 -l 0 +1

30. I am in favor of skipping meals.

Strongly Disagree 0 O O O O
-3 -2 -l 0 +1

31. It is not a good idea for me to skip meals. R
Strongly Disagree 0 O O O O
-3 -2 -1 O

32. I believe that skipping meals will help me improve my appearance.

Strongly Disagree 0 O O O O
. -3 -2 -l 0 +1

0
+2

33. I think that skipping meals will help me feel good about myself.

Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

34. Skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

35. I believe that skipping meals will help me reduce my risk of disease.

Strongly Disagree 0 O O O O
-3 -2 -l 0 +1

36. I think that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O
-3 -2 -l 0 +1
37. Skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O
-3 -2 -1 0 +1

38. I think that skipping meals will help me prevent cancer.

Strongly Disagree 0 O O O O
-3 -2 -1 O + l

100

0
+2

0
+2

0
+2

0
+2

0
+2

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

39. Skipping meals will be expensive.

Strongly Disagree 0 O O O
-3 -2 -1 0

40. Skipping meals will be convenient.

Strongly Disagree 0 O O O
-3 -2 -1 0

41. I intend to skip meals.

Strongly Disagree 0 O O
-3 -2 -1 O

42. I plan to skip meals.

Strongly Disagree 0 O O
-3 -2 -1 O

43. I aim to skip meals.

Strongly Disagree 0 O O
-3 -2 -1 0

44. I expect that I will skip meals.

Strongly Disagree 0 O O O
-3 -2 -l O

45. I do not intend to skip meals. R

Strongly Disagree 0 O O O
-3 -2 -l O

46. Skipping meals is easy for me.

Strongly Disagree 0 O O O
-3 -2 -l 0

47. It is difﬁcult for me to skip meals. R

Strongly Disagree 0 O O O
-3 -2 -1 O

48. It is simple for me to skip meals.

Strongly Disagree 0 O O O
-3 -2 -l O

49. It is feasible for me to skip meals.
Strongly Disagree 0 O O O
-3 -2 -1 0

50. I have control over the number of meals I skip.
Strongly Disagree 0 O O O
-3 -2 -l O

51. I cannot control the number of meals I skip.
Strongly Disagree 0 O O O
-3 -2 -l O

52. I am in charge of the number of meals I skip.

Strongly Disagree 0 O O O
-3 -2 -1 O

101

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

Strongly Agree

53. I have the power to determine the number of meals I skip.
Strongly Disagree 0 O O O O O O
-3 -2 -1 0 +1 +2 +3

How important are each of the following to you?
1. Improving my appearance.

Not Important At All 0 O O O
O l 2

u
i‘O
MO
O‘O
Q0

2. Feeling good about myself.
Not Important At All 0 O O O O O O O
O l 2 4 6 7

DJ
tit

3. Improving my overall health.
Not Important At All 0 O O O
0 l 2

or
*0
MO
°‘O
\IO

4. Reducing my risk of disease.
Not Important At All 0 O O O O O O O
0 l 2 3 4 6 7

M

5. Losing weight.
Not Important At All 0 O O O O O O O
0 l 2 3 4 6 7

M

6. Increasing my attractiveness.
Not Important At All 0 O
O 1

NO
wO
J>0
MO
0‘0
\IO

7. Preventing cancer.
Not Important At All 0 O O O O O O O
0 2 4 6 7

H
b.)
KI!

8. Saving money.
Not Important At All 0 O O O O O O O
4 6 7

O
_-
N
W
M

9. Convenience.

Not Important At All 0 O O O O O O O
O l 2 3 4 5 6 7

10. Eating up to 9 servings of fruits and vegetables a day.

Not Important At All 0 O O O O O O O
0 1 2 3 4 5 6 7

ll. Skipping meals.
Not Important At All 0 O O O O O O O
O l 2 3 4 5 6 7

12. Generally speaking, doing what people important to me think I should do.

Not Important At All 0 O O O O O O O
0 l 2 3 4 5 6 7

102

Strongly Agree

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

Extremely Important

13. Generally speaking, doing what my family thinks I should do.
Not Important At All 0 O O O O O O 0 Extremely Important
0 1 2 3 4 5 6

\l

14. Generally speaking, doing what my close friends think I should do.
Not Important At A11 0 O O O O O O ' 0 Extremely Important
0 l 2 3 4 5 6 7

When you answer the next set of questions, think about your family.
What are your family’s opinions about fruit and vegetable consumption?

1. My family thinks that I should eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. It is important to my family that I eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. My family believes that it is important for me to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

4. According to my family, every day I should eat up to 9 servings of fruit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3
5. My family believes that eating up to 9 servings of fruit and vegetables a day will help me improve my
appearance. ~
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3

6. My family thinks that eating up to 9 servings of fruit and vegetables a day will help me feel good about
myself
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3
7. According to my family, eating up to 9 servings of fruit and vegetables a day will help me improve my
overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3
8. My family believes that eating up to 9 servings of fruit and vegetables a day will help me reduce my
risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3

9. My family thinks that eating up to 9 servings of fruit and vegetables a day will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3

10. According to my family, eating up to 9 servings of fruit and vegetables 3 day will help me be more
attractive.

Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3

103

1 1. My family thinks that eating up to 9 servings of fruit and vegetables a day will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

12. My family thinks that eating up to 9 servings of fruit and vegetables a day will be expensive. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

13. My family believes that eating up to 9 servings of fruit and vegetables a day will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

What are your family’s opinions about skipping meals?

14. My family thinks that it is acceptable skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

15. It is okay with my family to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

16. According to my family, it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

17. It is not acceptable to my family to skip meals. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3 .

18. My family believes that it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

19. My family believes that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

20. My family thinks that skipping meals help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

21. According to my family, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

22. My family believes that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

23. My family thinks that skipping meals will help me lose weight.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

104

24. According to my family, skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

25. My family thinks that skipping meals will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

26. My family thinks that skipping meals will be expensive. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

27. My family believes that skipping meals will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

When you answer the next set of questions, think about your close friends.
What do your close friends think about fruits and vegetables?

1. My close friends think that I should eat up to 9 servings of fruits and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. It is important to my close friends that I eat up to 9 servings of ﬁuits and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. My close friends believe that it is important for me to eat up to 9 servings of fruits and vegetables
every day. ,
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

4. According to my close friends, every day I should eat up to 9 servings of fruits and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3
5. My close friends believe that eating up to 9 servings of fruits and vegetables a day will help me
improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

6. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me feel good
about myself.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. According to my close friends, eating up to 9 servings of fruits and vegetables a day will help me
improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

105

8. My close friends believe that eating up to 9 servings of fruits and vegetables a day will help me reduce

my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3
9. My close friends think that eating up to 9 servings of fruits and vegetables 3 day will help me lose
weight.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3

10. According to my close ﬁ'iends, eating up to 9 servings of hits and vegetables a day will help me be
more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

11. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me prevent
cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

12. My close friends think that eating up to 9 servings of fruit and vegetables a day will be expensive. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

13. My close friends believe that eating up to 9 servings of fruit and vegetables a day will be convenient.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

What do your close friends think about skipping meals?

14. My close friends think that it is acceptable skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

15. It is okay with my close friends to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

16. According to my close friends, it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

17. It is not acceptable to my close friends to skip meals. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

18. My close friends believe that it is acceptable to skip meals.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

19. My close friends believe that skipping meals will help me improve my appearance.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 O + 1 +2 +3

106

20. My close ﬁiends think that skipping meals will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

21. According to my close friends, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O ‘ O Strongly Agree
-3 -2 -1 0 +1 +2 +3

22. My close friends believe that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

23. My close friends think that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

24. According to my close friends, skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

25. My close friends think that skipping meals will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

26. My close friends believe that skipping meals will be expensive. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

27. My close friends believe that skipping meals will be convenient.
Strongly Disagree 0 O O O O O O - Strongly Agree
-3 -2 -l 0 +1 +2 +3

 

Please click on the SUBMIT button to submit your survey. SUBMIT

 

 

 

107

(After participants submitted Survey #2)
Thank you for completing the second survey!
In 7 -10 days you will receive an email informing you that the ﬁnal survey is online.

If you are interested in learning more about the facilities available to
you on campus please click on the link below.

http://www.olin.msu.edu/navigation.php?id=99

(End Survey #2)

108

APPENDIX T

Fruit and Vegetables Belief Scale

1. Eating up to 9 servings of fruit and vegetables a day will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

2. Eating up to 9 servings of fruit and vegetables a day will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

3. Eating up to 9 servings of fruit and vegetables a day will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. Eating up to 9 servings of fruit and vegetables a day will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. Eating up to 9 servings of fruit and vegetables a day will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

6. Eating up to 9 servings of fruit and vegetables a day will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. Eating up to 9 servings of fruit and vegetables a day will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

8. Eating up to 9 servings of fruit and vegetables a day will be expensive. R
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

109

APPENDIX U

Importance Scale
How important are each of the following to you?

1. Improving my appearance.
Not Important At All 0 O O O O O O 0 Extremely Important
3 4 6 7

0 l 2

M

2. Feeling good about myself.
Not Important At All 0 O O O O O O 0 Extremely Important
0 1 2 4 7

b)
M
O‘

3. Improving my overall health.
Not Important At All 0 O O O O O O 0 Extremely Important
0 1 2 4 6 7

DJ
M

4. Reducing my risk of disease.
Not Important At All 0 O O
0 1 2

 

mo

0 O O 0 Extremely Important
4 6 7

LII

5. Losing weight.
Not Important At All 0 O O O O O O 0 Extremely Important
0 l 2 4 6

b)
M
\l

6. Increasing my attractiveness.
Not Important At A11 0 O
0 1

NO
040

O O O 0 Extremely Important
4 6 7 ‘

7. Preventing cancer.
Not Important At All 0 O O O O O O 0 Extremely Important
0 2 4 6 7

1.—
DJ

8. Saving money.
Not Important At All 0 O O O O O O 0 Extremely Important
4 6 7

O
....
N
us

110

APPENDIX V

Skipping Meals Belief Scale

1. I believe that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. I think that skipping meals will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. Skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. I believe that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. I think that skipping meals will help me lose weight.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

6. Skipping meals will help me be more attractive.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

7. I think that skipping meals will help me prevent cancer.
Strongly Disagree 0 O O O O O O , Strongly Agree
-3 -2 -l 0 +1 +2 +3

8. Skipping meals will be expensive. R

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

111

APPENDIX W

Family Normative Beliefs (Fruit and Vegetables Topic)

1. My family believes that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. My family thinks that skipping meals help me feel good about myself.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. According to my family, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. My family believes that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

5. My family thinks that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

6. According to my family, skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. My family thinks that skipping meals will help me prevent cancer.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

8. My family thinks that skipping meals will be expensive.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 + 1 +2 +3

112

APPENDIX X

Motivation to Comply with Family

How important is the following to you?
1. Generally speaking, doing what my family thinks I should do.

Not Important At All 0 O O O O O O 0 Extremely Important
0 1 2 3 4 5 6 7

113

APPENDIX Y
F arnily Normative Beliefs (Skipping Meals Topic)

1. My family believes that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. My family thinks that skipping meals help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. According to my family, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

. 4. My family believes that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. My family thinks that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

6. According to my family, skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. My family thinks that skipping meals will help me prevent cancer. ,
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

8. My family thinks that skipping meals will be expensive. R

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

114

 

APPENDIX Z

Close Friends’ Normative Beliefs (Fruit and Vegetables Topic)

1. My close friends believe that eating up to 9 servings of fruits and-vegetables a day will help me
improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

2. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me feel good
about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. According to my close friends, eating up to 9 servings of fruits and vegetables a day will help me
improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. My close friends believe that eating up to 9 servings of fruits and vegetables a day will help me reduce
my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -l 0 +1 +2 +3
5. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me lose
weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3
6. According to my close friends, eating up to 9 servings of fruits and vegetables a day will help me be
more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree

-3 -2 -1 0 +1 +2 +3

7. My close friends think that eating up to 9 servings of fruits and vegetables a day will help me prevent
cancer.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 + 1 +2 +3

8. My close friends think that eating up to 9 servings of fruit and vegetables a day will be expensive. R

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

115

APPENDIX AA

Motivation to Comply with Close Friends
How important is the following to you?

1. Generally speaking, doing what my close friends think I should do.
Not Important At All 0 O O O O O O 0 Extremely Important
0 l 2 3 4 5 6 7

116

APPENDIX BB

Close Friends’ Normative Beliefs (Skipping Meals Topic)

1. My close friends believe that skipping meals will help me improve my appearance.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. My close ﬁ-iends think that skipping meals will help me feel good about myself
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. According to my close friends, skipping meals will help me improve my overall health.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

4. My close friends believe that skipping meals will help me reduce my risk of disease.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

5. My close friends think that skipping meals will help me lose weight.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

6. According to my close friends, skipping meals will help me be more attractive.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

7. My close friends think that skipping meals will help me prevent cancer. ‘
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

8. My close friends believe that skipping meals will be expensive. R

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

117

APPENDD( CC

Behavioral Intent Scale (Fruit and Vegetables Topic)

1. I intend to incorporate up to 9 servings of fruit and vegetables into my diet every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

2. I plan to eat up to 9 servings of fruit and vegetables every day.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

3. Every day I aim to eat up to 9 servings of fruit and vegetables.
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. I expect that I will eat up to 9 servings of fruit and vegetables every day.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. I do not intend to eat up to 9 servings of fruit and vegetables every day. R "'

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

Note. * indicates item was deleted from measurement model in main experiment.

118

APPENDIX DD

Behavioral Intent Scale (Skipping Meals Topic)

1. I intend to skip meals.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

2. I plan to skip meals.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -1 0 +1 +2 +3

3. I aim to skip meals.

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

4. I expect that I will skip meals. "'

Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0 +1 +2 +3

5. I do not intend to skip meals. R "
Strongly Disagree 0 O O O O O O Strongly Agree
-3 -2 -l 0

Note. *indicates item was deleted from measurement model for main experiment.

119

 

REFERENCES

120

Abelson, R. P. (1995). Statistics as principled argument. Hillsdale, NJ: Lawrence
Erlbaum.

Ajzen, I., & Fishbein, M. (1980). Prediction of goal-directed behavior: Attitudes,
intentions, and perceived behavioral control. Journal of Experimental Social
Psychology, 22, 453-474.

Ajzen, I., & F ishbein, M. (1974). Factors inﬂuencing intentions and the intention-
behavior relation. Human Relations, 27, 1-15.

Ajzen, I., & Fishbein, M. (1970). The prediction of behavior from attitudinal and
normative variables. Journal of Experimental and Social Psychology, 6, 466-487.

Armstrong, J. G., & Roth, D. M. (1989). Attachment and separation difﬁculties in eating
disorders: A preliminary investigation. International Journal of Eating Disorders,
8, 141-155.

Bem, D. J. (1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in
experimental social psychology (V 01. 6, pp. xx-xx). New York: Academic Press.

Bentler, P. M., & Speckart, G. (1981). Attitudes “cause” behaviors: A structural equation
analysis. Journal of Personality and Social Psychology, 40(2), 226-238.

Boster, F. J., & Cruz, M. G. (2001). Persuading in the small group context. In J. P.
Dillard & M. Pfau (Eds), The persuasion handbook: Developments in theory and
practice (pp. 477-494). Thousand Oaks, CA: Sage.

Boster, F. J ., Fryrear, J. E., Mongeau, P. A., & Hunter, J. E. (1982). An unequal speaking
linear discrepancy model: Implications for the polarity shift. In M. Burgoon (Ed.),
Communication Yearbook 6 (pp. 395-418). Beverly Hills, CA: Sage.

Boster, F. J ., Mayer, M. E., Hunter, J. E., & Hale, J. L. (1980). Expanding the persuasive
arguments explanation of the polarity shiﬁ: A linear discrepancy model. In D.
Nimmo (Ed.) Communication Yearbook 4 (pp. 165-176). New Brunswick, NJ:
Transaction Books.

Calarn, R., Waller, G., Slade, P., & Newton, T. (1990). Eating disorders and perceived
relationships with parents. International Journal of Eating Disorders, 9, 479-485.

Graber, J. A., Brooks-Gunn, J ., Paikoff, R. L., & Warren, M. P. (1994). Prediction of

eating problems: An 8-year study of adolescent girls. Developmental Psychology,
30, 823-834.

Hausenblas, H. A., & Carton, A. V. (1998). Group inﬂuences on eating and dieting
behaviors in residence members. College Students Journal, 32. 585-589.

121

Haworth-Hoeppner, S. (2000). The critical shapes of body image: The role of culture and
family in the production of eating disorders. Journal of Marriage and the Family,
62, 212-227.

Humphrey, L. L. (19863). Family relations in bulirnic-anorexic and nondistressed
families. International Journal of Eating Disorders, 5, 223-232.

Humphrey, L. L. (1986b). Structural analysis of parent-child relationships in eating
disorders. Journal of Abnormal Psychology, 95, 395-402.

Humphrey, L. L. (1987). Comparison of bulimic-anorexic and nondistressed families
using structural analysis of social behavior. Journal of the American Academy of
Child and Adolescent Psychiatry, 26, 248-255.

Humphrey, L. L. (1989). Observed family interactions among subtypes of eating
disorders using structural analysis of social behavior. Journal of Counseling and
Clinical Psychology, 57, 206-214.

Hunter, J. E., & Gerbing, D. W. (1982). Unidimensional measurement, second order
factor analysis, and causal models. Research in Organizational Behavior, 4, 267-
320.

Kahle, L. R., & Berman, J. J. (1979). Attitudes cause behaviors: A cross-lagged panel
analysis. Journal of Personality and Social Psychology, 3 7(3), 315-321.

Kelman, H. C. (1974). Attitudes are alive and well and gainfully employed in the sphere
of action. American Psychologist, 29, 310-324.

Kim, M-S., & Hunter, J. E. (1993a). Attitude-behavior relations: A meta-analysis of
attitudinal relevance and topic. Journal of Communication, 43, 101-142.

Kim, M-S., & Hunter, J. E. (1993b). Relationships among attitudes, behavioral
intentions, and behavior: A meta-analysis of past research, part 2. Communication
Research, 20(3), 331-364.

Levine, M. P., Smolak, L. Hayden, H. (1994). The relation of sociocultural factors to
eating attitudes and behaviors among middle school girls. Journal of Early
Adolescence, 14, 471-490.

Levine, M. P., Smolak, L., Moodey, A. F., Shuman, M. D., & Hessen, L. D. (1994).

Normative developmental challenges and dieting and eating disturbances in
middle school girls. International Journal of Eating Disorders, 15, 11-20.

122

 

McGuire, W. J. (1976). The concept of attitudes and their relations to behaviors. In H. W.
Sinaiko & L. A. Broedling (Eds), Perspectives on attitude assessment: Surveys
and their alternatives (pp. 7-31). Champaign, IL: Pendleton.

Moreno, A., & Thelen, M. H. (1993). Parental factors related to bulimia nervosa.
Addictive Behaviors, 18, 681-689.

National Center for Health (2003). Fruits and vegetables: Eat 5 to 9 a day for better
health. Retrieved June 4, 2003 ﬁom http://5aday.gov/index-about.shtml.

Nucifora, J., Gallois, C., & Kashima, Y. (1993). Inﬂuences on condom use among
undergraduates: Testing the theory of reasoned action and planned behavior. In D.
J. Terry, C. Gallois, & M. McCamish (Eds), The theory of reasoned action: Its
application to AIDS-preventative behaviour (pp. 47-64). Tarrytown, NY:
Pergamon Press.

O’Keamey, R. (1996). Attachment disruption in anorexia nervosa and bulimia nervosa: A
review of theory and empirical research. International Journal of Eating
Disorders, 20, 115-127.

Oliver, K. K., & Thelen, M. H. (1996). Children’s perceptions of peer inﬂuence on eating
concerns. Behavior Therapy, 27, 25-39.

Paxton, S. J ., Schutz, H. K., Wertheim, E. H., & Muir, S. L. (1999). Friendship clique
and peer inﬂuence on body image concerns, dietary restraint, extreme weight-loss

behaviors, and binge eating in adolescent girls. Journal of Abnormal Psychology,
108. 255-266.

Polivy, J ., & Herman, C. P. (2002). Causes of eating disorders. Annual Review of
Psychology, 53, 187-213.

Sanﬁner, J. L., Crowther, J. H., Crawford, P. A., & Watts, D. D. (1996). Maternal
inﬂuence (or lack thereof) on daughters’ eating attitudes and behaviors. Eating
Disorders, 4, 147-159.

Scalf-Mclver, L., & Thompson, J. K. (1989). Family correlates of bulimic characteristics
in college females. Journal of Child Psychology, 45, 467-472.

Schur, E. A., Sanders, M., & Steiner, H. (2000). Body dissatisfaction and dieting in
young children. International Journal of Eating Disorders, 27, 74-82.

Schlegel, R. P., D’Avernas, J. R., Zanna, M. P., DeCourville, N. H., & Manske, S. R.

(1992). Problem drinking: A problem for the theory of reasoned action? Journal
of Applied Social Psychology, 22(5), 358-385.

123

Steiger, H., Van der Feen, J ., Goldstein, C., & Leichner, P. (1989). Defense styles and
parental bonding in eating-disordered women. International Journal of Eating
Disorders, 8, 131-140.

Streigel-Moore, R. H., & Keamey-Cooke, A. (1994). EXploring parents’ attitudes and
behaviors about their children’s physical appearance. International Journal of
Eating Disorders, 15, 377-385.

Thelen, M. H., & Cormier, J. F. (1995). Desire to be thinner and weight control among
children and their parents. Behavior Therapy, 26, 85-99.

Twarnley, E. W., & Davis, M. C. (1999). The sociocultural model of eating disturbance
in young women: The effects of personal attributes and family environment.
Journal of Social and Clinical Psychology, 18, 467-489.

Vincent, M. A., & McCabe, M. P. (2000). Gender differences among adolescents in
family, and peer inﬂuences on body dissatisfaction, weight loss, and binge eating
behaviors. Journal of Youth and Adolescence, 29, 205-221.

Wicker, A. W. (1969). Attitude versus action: The relationship of verbal and overt
behavioral responses to attitude. Journal of Social Issues, 25, 41-78.

Young, E. M., & Fors, S. W. (2001). Factors related to the eating habits of students in
grades 9-12. The Journal of School Health, 71, 483-488.

124

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

m[m1111MWW