Does the transtheoretical model need an attitude adjustment?

Pros and Cons and Attitude 1

Running head: DECISIONAL BALANCE AND ATTITUDE

Does the Transtheoretical Model Need an Attitude Adjustment?

Integrating Attitude with Decisional Balance as Predictors of Stage of Change for Exercise

Patricia J. Jordan, M.A.1 Claudio R. Nigg, Ph.D. 1

Gregory J. Norman, Ph.D.2 Joseph S. Rossi, Ph.D. 1

Sonya V. Benisovich, M.A. 3

1 Cancer Prevention Research Center, University of Rhode Island

2 Stanford Center for Research in Disease Prevention, Stanford University School of Medicine

3 Stanford University, School of Education

Reference:

Jordan, P. J., Nigg, C. R., Norman, G. J., Rossi, J. S., & Benisovich, S. V. (2001). Does the transtheoretical model need an attitude adjustment? Integrating attitude with decisional balance as predictors of stage of change for exercise. Psychology of Sport and Exercise, 3(1), 65-83.

This research was supported by Grants CA27821 and CA50087 from the National Cancer

Institute.

The authors gratefully acknowledge the contribution of Dr. Lisa L. Harlow.

Correspondence concerning this article should be addressed to: Patricia J. Jordan, Cancer

Prevention Research Center, University of Rhode Island, 2 Chafee Road, Kingston, RI, 02881,

U.S.A. Electronic correspondence may be sent via e-mail to: <[email protected]>.

Pros and Cons and Attitude 2

Abstract

Objectives. This study compared conceptually similar decision-making components from

the theory of planned behavior (attitude) and the transtheoretical model (pros and cons) to

determine the extent to which attitude towards exercise added to the prediction of stage of

exercise behavior beyond that of the pros and cons.

Method. A sample of college undergraduates (N=223) were given stage of change,

attitude and decisional balance measures regarding their exercise behavior. Structural equation

modeling (SEM) was used to test the underlying measurement structure of the decision-making

components, while a series of discriminant function analyses (DFAs) were performed using a

combination of pros, cons, and cognitive and affective attitudes as predictors of membership in

one of the five stages of change.

Results: SEM determined that a correlated four-factor model, which included pros and

cons and two attitude subscales, provided the best fit to the data. The DFAs revealed that the

addition of attitude components to pros and cons significantly increased the overall explained

variance across the stages of change from 32% to 56% and improved the predictive ability of pros

and cons from 31.2% to 48.2%.

Conclusions: Although conceptually related, pros, cons and attitude were not closely

linked at a construct-measurement level. Furthermore, the addition of attitude to pros and cons

increased the overall explained variance across stages of change and improved the predictive

ability of pros and cons alone. The measurement model and DFA results taken in combination

strongly suggest that the addition of cognitive and affective attitudes may strengthen the decision-

making aspect of the TTM.

Pros and Cons and Attitude 3

Does the TTM Need an Attitude Adjustment?

Integrating Attitude with Decisional Balance as Predictors of Stage of Change for Exercise

The most recent statistics available from the U.S. Department of Health and Human

Services (Murray & Lopez, 1996; National Center for Health Statistics, 1997) associate lack of

exercise with five of the top 10 U.S. population causes of death. Exercise is an associated risk

factor for diseases of the heart (the number one cause of death in 1995), cerebrovascular diseases

(number three), lung diseases (number four), diabetes (number seven), and suicide (number nine).

The U.S. government’s Healthy People 2010 initiative focuses on advances in preventive

therapies and reduction of risk factors associated with major causes of death and chronic illness,

including physical activity and fitness (U.S. Department of Health and Human Services, 2000).

This initiative identified physical activity as one of the leading indicators of preventive health

behaviors and emphasized the need for behavior health professionals to develop and provide more

effective interventions and behavior change programs designed to promote and maintain healthy

behaviors. However, the choice of an optimal program must ultimately be based on a better

understanding of the factors underlying behavior in any given situation (Godin, 1994).

A number of different theoretical frameworks have been applied to the study of exercise

behavior among various segments of the population (Godin, 1994). The theory of planned

behavior (Ajzen & Madden, 1986; Ajzen, 1991), for example, has been identified as one of the

most commonly applied behavior-prediction models in the exercise domain (McAuley &

Courneya, 1993), as well as the theory with the most support in this same area (Courneya, 1995;

Courneya, Estabrooks & Nigg, 1997).

Pros and Cons and Attitude 4

Another widely used behavior-change construct in the exercise domain is Prochaska and

DiClemente’s (1983) stages of change, the central organizing construct of the transtheoretical

model of behavior change (Prochaska & DiClemente, 1983, 1984; DiClemente, Prochaska,

Fairhurst, Velicer, Velasquez, & Rossi, 1991). Currently, the most popular stage model in health

psychology (Horwath,1999), the transtheoretical model integrates cognitive, affective and

behavioral processes and principles of change from leading theories of psychotherapy and health

psychology, including 10 processes of change, decisional balance and self-efficacy.

Although these two theories propose different determinants of behavior and behavior

change, they include conceptually similar variables—each theory places great importance on

decision-making (Dishman, 1994). Pros and cons from the transtheoretical model mirror the

behavioral beliefs from the theory of planned behavior that form the basis of attitude toward

behavior. In fact, there have been suggestions that decisional balance may not be a distinct

construct in this regard, as evidence by applications that treat it as an attitude-like construct (e.g.,

Rakowski et al., 1999).

The purpose of this study was to compare conceptually similar decision-making

components from the theory of planned behavior (attitude) and the transtheoretical model (pros

and cons) to determine if, and to what extent, attitude towards exercise added to the prediction of

transtheoretical model stage of exercise behavior beyond that of the pros and cons.

Development of a comprehensive behavior-change model is a process that assumes

continuous evaluation and improvement. Combining elements from other theoretical frameworks,

such as the theory of planned behavior, with the transtheoretical model can have the advantage of

including and examining different theoretically relevant variables (Courneya, 1995). This is not to

Pros and Cons and Attitude 5

suggest that the transtheoretical model is uncontested as the behavior change model of choice

(c.f., Joseph, Breslin, Skinner, 1999; Weinstein, 1998); however, it has captured broad interest in

research and practice and warrants further investigation. Such research also has the potential to

increase our understanding of what is important in promoting exercise behavior. Further, it can

evaluate whether two conceptually similar variables (e.g., decisional balance and attitude) address

real differences or whether they represent the same construct using different labels.

Theoretical Overview

The theory of planned behavior (Ajzen, 1991), an extension of Fishbein and Ajzen’s

(1975; Ajzen & Fishbein, 1980) theory of reasoned action, attempts to understand and

consequently predict behaviors not entirely under individual control. The theory of planned

behavior proposes that intention is the best predictor of behavior and is directly influenced by

attitude, subjective norm and perceived behavioral control. Attitude focuses on an individual's

positive or negative evaluation of a specific behavior, while the normative component reflects the

perceived social pressure felt by the individual to perform (or not) a particular behavior (Ajzen &

Fishbein, 1980). Perceived behavioral control assesses the presumed ease or difficulty of

performing a behavior and is thought to be an approximation of one’s actual situational control

(Ajzen & Fishbein, 1980; Terry & Hogg, 1996). The theory of planned behavior has been applied

to the study of exercise behavior by a number of investigators (e.g., Courneya & McAuley, 1995;

Dishman, 1994; Godin, 1993, 1994; McAuley & Courneya, 1993) and has been very helpful in

understanding the decision-making process underlying exercise behavior (Ajzen & Driver, 1992;

Godin, 1994).

Pros and Cons and Attitude 6

The strength of the transtheoretical model is its focus on the dynamic nature of health

behavior change (Marcus, Rakowski & Rossi, 1992). The stages of change were introduced as a

framework to describe the temporal aspect of the adoption and maintenance of healthy behaviors.

Individuals modifying a given behavior move through a series of stages, from precontemplation to

maintenance (Prochaska, DiClemente, & Norcross, 1992).

Precontemplation is the stage in which an individual has no intent to change behavior in

the near future, usually measured as the next six months. Individuals in the Contemplation stage

openly state their intent to change within the next six months. This has been deemed about as far

in advance as an individual can plan decisions (Prochaska & Marcus, 1994). Preparation is the

stage in which individuals intend to take steps to change, usually within the next 30 days. This

time frame reflects the immediacy of the intention to act. Preparers may have already made minor

adjustments in their thought patterns and behaviors, but typically do not reach the predetermined

Action criteria. The Action stage is one in which an individual has made overt, perceptible lifestyle

modifications for fewer than six months. For exercise, the highest likelihood for relapse occurs

within the first six months of starting a regular program (Dishman, 1994). Those in the

Maintenance stage are working to prevent relapse and consolidate gains secured during Action

(DiClemente et al., 1991; Prochaska et al., 1992; Prochaska, Redding, & Evers, 1997) and have

successfully continued exercising beyond six months (e.g., longer than a sport season).

Processes of change are those covert and overt activities individuals use to change

behavior (Prochaska, Velicer, DiClemente, & Fava, 1988), decisional balance represents an

individual’s assessment of the perceived importance of the advantages and disadvantages (pros

and cons) of performing a behavior (Velicer, DiClemente, Prochaska, & Brandenburg, 1985), and

Pros and Cons and Attitude 7

self-efficacy reflects confidence in one’s ability to perform (or not) the behavior in a given

situation (DiClemente, Prochaska, & Gibertini, 1985). Based on initial studies of smoking

cessation behavior, the transtheoretical model has been successfully applied to a wide variety of

health-related behaviors (Prochaska et al., 1994) and has demonstrated reliability in describing and

predicting behavior change. To date, more than 50 published studies have applied transtheoretical

model variables to exercise behavior in a variety of populations to both describe and increase

exercise behavior (Nigg, 1999).

Empirical Overview

Research using theory of planned behavior constructs has identified attitude as an

important behavior-change component (Courneya, 1995; Courneya, Estabrooks, & Nigg, 1997;

Courneya, Nigg, & Estabrooks, 1998; Godin, 1993, 1994). To obtain salient beliefs about a

person’s attitudes towards performing a behavior, questions are phrased in terms of the

advantages and disadvantages of performing this same behavior (Ajzen & Fishbein, 1980).

Courneya (1995; Courneya, Nigg, & Estabrooks, 1998) assessed attitude across the five stages of

change in older adults and found that attitude significantly discriminated most stages of change

pairs. From the information provided by Godin (1993, 1994), approximately 30% of the variance

in intention to exercise seems to be explained by the attitudinal component. Although attitude has

been traditionally conceived as a complex system, combining affective, cognitive and behavioral

elements (Rosenberg & Hovland, 1960), researchers often evaluate it as a unidimensional concept

(Ajzen & Fishbein, 1980). Recent literature, however, has become more concerned with the

distinction between the cognitive and affective dimensions of attitude as they relate to behavioral

decisions (Crites, Fabrigar & Petty, 1994; Eagly, Mladinic & Otto, 1994; Trafimow & Sheeran,

Pros and Cons and Attitude 8

1998). These researchers have concluded that, when assessed independently, cognitive and

affective attitude components do make discrete contributions to the overall attitude dimension.

Research applying the transtheoretical model utilizes decisional balance in a similar fashion

(Velicer et al., 1985). Based on Janis and Mann’s (1977) model of decision making, the

transtheoretical model incorporates an individual’s relative assessment of the benefits and costs of

changing a specific behavior. Originally conceived as an eight-component model — composed of

gains and losses to self and others, as well as approval or disapproval from self and others —

most studies centered around the transtheoretical model have been unable to replicate an eight-

factor or four-factor structure with any reliability (Prochaska et al., 1994). The most consistent

replication has been found using a simple two-factor structure: the pros and cons of changing

(Prochaska et al., 1994).

Pros and cons have been identified as consistently discriminating among individuals at

different stages of readiness for exercise (Gorely & Gordon, 1995; Marcus, Eaton, Rossi, &

Harlow, 1994; Marcus & Owen, 1992; Marcus, Rakowski, & Rossi, 1992; Marcus, Selby,

Niaura, & Rossi, 1992; Nigg & Courneya, 1998). A cross-sectional examination of 12 health

behaviors, including exercise, confirmed the importance of the relationship between the pros and

cons progress through the stages of change (Prochaska et al., 1994). For all 12 behaviors, the

cons of behavior change outweighed the pros for individuals in the precontemplation stage;

whereas, the reverse was found to be true for those in action and maintenance. Based on results

from Marcus, Rakowski and Rossi (1992), decisional balance in the realm of exercise explains

nearly 25% of variance across stages.

Pros and Cons and Attitude 9

Methods

Participants

Study participants (N=223) were recruited from undergraduate psychology classes at a

New England-area university in the U.S. The mean age of the sample was 19.8 years (SD=3.1),

with an average 13.5 years of education (SD=1.2); 69% were female; 95% unmarried; and 79%

reported good or very good health. The sample was composed of 89% White, 4% Latino-

American, 3% African-American, 3% Asian and 1% Other participants. Across stages of change,

the sample was composed of 8% in precontemplation (n=17), 11% in contemplation (n=24), 39%

in preparation (n=88), 12% in action (n=27), and 30% in maintenance (n=67).

Procedure

Institutional approval was obtained, as well as permission from course instructors, to have

participants complete a set of 11 questionnaires, including the measures noted below and

demographics information. Potential participants were supplied with all relevant information

pertaining to the study and their voluntary participation implied informed consent. Students were

given regular class sessions to complete the questionnaires, or were asked to complete the

questionnaires at home and return them by the following class session. Submission of a completed

questionnaire was rewarded with credit toward the department’s research participation

requirements. The entire set of questionnaires took approximately 30-40 minutes to complete.

Measures

Stages of Change for Exercise Behavior. Stages of exercise behavior were measured

using a five-item algorithm, where respondents were asked to answer “yes” or “no” to a series of

Pros and Cons and Attitude 10

five questions based on their current exercise behavior and future intentions to exercise (Reed,

Velicer, Prochaska, Rossi, & Marcus, 1997). Regular exercise was defined prior to the questions

as “any planned physical activity performed to increase physical fitness from three to five times

per week for a minimum 20 minutes per session at a level that increases one’s breathing rate and

causes one to break a sweat.” Subjects who did not exercise regularly and had no intention to

begin in the next six months were considered to be in the precontemplation stage. Those who did

not exercise regularly, but intended to start within the next six months, but not the next 30 days

were classified in the contemplation stage. Those who did not exercise regularly, but intended to

start in the next 30 days were placed in the preparation stage. Subjects who did exercise regularly,

but had not done so for a minimum of six months, were in action. Those who had been exercising

regularly for more than six months were grouped in the maintenance stage. Versions of this

measure have been found to be both reliable and valid (Courneya, 1995; Marcus, Banspach et al.,

1992; Marcus, Selby et al., 1992; Marcus & Simkin, 1993; Nigg & Courneya, 1998; Reed et al.,

1997).

Attitude was measured using six seven-point semantic differential scales, as suggested by

Fishbein & Ajzen (1975; Ajzen & Fishbein, 1980). Adjectives were presented without a negative

or positive sign preceding the numerical ratings. The statement preceding the adjective pairs was,

“I feel my participation in exercise at the present time is….” Three items tapped cognitive

properties of attitude (useful/useless, harmful/beneficial, wise/foolish) and three items tapped

affective properties of attitude (enjoyable/unenjoyable, pleasant/unpleasant, stressful/relaxing)

(e.g., Useful 3 - 2 - 1 - 0 - 1- 2 - 3 Useless). For analyses, the scales were translated (1 to 7), with

Pros and Cons and Attitude 11

lower scores corresponding to the negative adjective. Internal consistency (coefficient Alpha) for

each of the cognitive and affective subscales was .85.

Decisional Balance. Due to the fact that the existing decisional balance measure for

exercise had been developed on an adult worksite sample, the measure was re-constructed using

this sample of college students. Development of this measure followed the sequential method of

scale construction described by Jackson (1970, 1971) and Comrey (1988) and involved a

sequence of steps to ensure content and internal validity. An initial pool of approximately 100

statements (including Marcus et al.’s scale, 1992) representing Janis and Mann's (1977) original

eight-category conception of the benefits (pros) and costs (cons) of exercise was generated by a

group of regular exercisers and non-exercisers, including the researchers. The statements were

reviewed and revised to improve understanding, eliminate redundancies and a Q-sort was

conducted to ensure proper representation across the eight decisional-balance categories

recommended by Janis and Mann (1977). The final pool of items consisted of 69 statements,

including 38 pros and 31 cons of exercising. Subjects were asked to indicate on a Likert scale,

ranging from “not at all important” (1) to “extremely important” (5), the importance of each

statement when deciding whether or not to exercise in their leisure time.

A principal components analysis conducted on the 69 x 69 matrix of interitem correlations

for the decisional balance items yielded two components, labeled pros and cons, which was

consistent with the literature (Prochaska et al., 1994). The number of components to retain was

determined using Velicer’s (1976; Zwick & Velicer, 1986) minimum average partial (MAP)

correlation procedure and Horn’s (1965) parallel analysis.

Pros and Cons and Attitude 12

An item was retained on a component if it met the two following criteria: 1) it loaded .70

or greater on the target factor; and 2) it did not load higher than .20 on the second component.

Complex items and items that failed to meet the preceding guidelines were deleted. A total of 26

items, 15 pros and 11 cons, remained after the above procedure was completed. Ten items for

each component were selected based on strength of loading and additional qualitative assessment

to ensure breadth of construct. A more in-depth analysis of the structure of the decisional balance

item set upon which this research is based can be found elsewhere (Nigg et al., 1998).

The final decisional balance scale included 10 pros and 10 cons. Statement wording and

loadings for each of the items are provided in Table 1. Coefficient Alpha (internal consistency)

reliabilities were strong (pros, α=.95; cons, α=.94).

Insert Table 1 about here

Results

Standardization of Scores

To obtain a standard metric, T-scores (M=50, SD=10) were calculated for pros, cons and

the two attitude constructs across the five stages of change (see Table 2). Figure 1 illustrates the

relationship between these variables for the current sample. Consistent with the findings of

Prochaska and colleagues (1994), the cons of exercise adoption were higher than the pros in the

precontemplation stage but reversed positions between the contemplation and preparation stages.

Figure 1 also portrays the linear relationship of cognitive and affective attitude across stages, with

an increase of almost two standard deviations from precontemplation to maintenance.

Pros and Cons and Attitude 13

Insert Table 2 & Figure 1 about here

Multivariate analysis of variance (MANOVA) was used to determine mean differences

across the categorical stages of change for pros, cons and the two attitude subscales prior to

model testing (F(16,648)=10.68, p < .05, η2 = .16). Follow-up ANOVAs determined that pros

(F(4,218)=24.01, p < .05, η2 = .32), cognitive attitude (F(4,219)=20.97, p < .05, η2 = .29) and

affective attitude (F(4,215)=19.44, p < .05, η2 = .27) exhibited significant mean differences across

the stages of change. Results are shown in Table 3.

Insert Table 3 about here

Correlations

Correlations between constructs were calculated to help inform subsequent measurement

model testing (see Table 4). Only the pros and attitude subscales were significantly related.

Insert Table 4 about here

Measurement Model Testing

To test the underlying measurement structure of the decision-making components, a series

of competing models were specified and estimated using structural equation modeling (SEM) with

EQS software (Bentler, 1989). Maximum likelihood estimation was used for all analyses. To

reduce the number of manifest variables and facilitate model testing, item parcels were created for

the pros and cons scales, whereby randomly selected sets of items on each scale were combined

and averaged to form a total of three composite manifest indicators for each latent construct. Item

parceling has been recommended for several reasons, including increased reliability, decreased

Pros and Cons and Attitude 14

idiosyncratic variance, the tendency for the item parcels to be more normally distributed and,

ultimately, a reduction in the ratio of measured variables to subjects (Marsh, Richards, Johnson,

Roche, & Tremayne, 1994). The six items that formed the attitude scale remained as single-item

indicators when cognitive and affective attitude constructs were examined independently, but item

pairs were created when the six attitude items were used to form a single attitude construct. This

ensured a ratio of three manifest variables to each latent construct for each analysis.

The following models were investigated:

1. Null Model. The null model assumed that there were no latent factors and that the items

were completely independent. This model was not thought of as a serious representation of the

data, but is a useful baseline against which other models may be compared.

2. One-Factor Model. This model hypothesized one general unspecified factor for decision

making. Evidence for this model would have indicated no differentiation among the decisional

balance and attitudinal constructs, lending support to a hypothesis that the only differences

between the two constructs were conceptual in nature and that existing measures may not be able

to discriminate these differences..

3. Uncorrelated Two-Factor Model. This model specified two uncorrelated factors: one

for decisional balance, composed of pros and cons; and one for attitude, composed of the

cognitive and affective subscales. No previous literature has found support for conceiving pros

and cons as a single factor; however, this model was intended as a means for comparison.

4. Uncorrelated Three-Factor Model. This model indicated a discrimination of the three

factors (pros, cons and attitude) and that the constructs are best represented as independent

aspects of decision making. Evidence for this model would support the hypothesis that these three

Pros and Cons and Attitude 15

decision-making components were unrelated — evidence that the three constructs are unique,

both conceptually and quantitatively.

5. Correlated Three-Factor Model. This model is similar to the previous model, however,

the pros and attitude factors were allowed to correlate, based on the results shown in Table 4.

This model represented the measurement model with three distinct but somewhat related factors.

6. Uncorrelated Four-Factor Model. This model discriminated pros, cons and the two

attitude components (affective and cognitive) and represented them as unrelated decision-making

constructs. Evidence for this model would support an hypothesis of the uniqueness of cognitive

and affective attitudes, as well as their distinction from the two decisional balance constructs.

7. Correlated Attitude Subscales Model. This model was based on the correlations among

the four decision-making constructs, which indicated that the two attitude subscales factors were

highly correlated, whereas the other correlations were relatively small.

8. Correlated Four-Factor Model. Resembling the four uncorrelated factors model, this

model allowed the pros factor and the two attitude subscales to correlate with each other, based

on the results shown in Table 4.

To assess the fit of the specified models three different fit indices were calculated for each

model. These included: Chi-square/degrees of freedom ratio (χ2/df); Comparative Fit Index (CFI);

and average absolute standardized residuals (AASR). For the χ2/df ratio, values less than 2.0

indicate good overall model fit (Marsh, Balla, & McDonald, 1988). A χ2/df ratio above 5.0

indicates unacceptably large differences between the parameterized model and the data. The CFI

is a measure of the proportion of the variation and covariation explained by the specified model

(Bentler, 1990). The index is bounded by 0 and 1.0, with 1.0 indicating a perfect fit. CFI values

Pros and Cons and Attitude 16

greater than .90 indicate excellent model fit. The AASR is an index of the discrepancy between

the specified model and the data (Bentler, 1990). Generally, values of .05 or less indicate good

model fit with 0 indicating a perfect fit. It is important to note that there is no single accepted

index of model fit. Use of several measures of fit, representing a variety of conceptual approaches

to the problem of assessing model adequacy, provides some protection against the possibility of

model misspecification (Marsh et al., 1988). Evaluation of all the models that were tested is

summarized in Table 5.

Insert Table 5 about here

Structural equation modeling revealed that Model 8 (correlated four-factor model)

provided the best fit to the data; however, Model 7 (the correlated attitude subscales model) also

provided a reasonable fit. A chi-square difference test revealed that one or both of the paths

removed to form the latter model were significant, χ2(2)=31.26, p<.001. Based on these results

and the overall poor fit of the three-factor models, the remaining analyses were conducted on all

four factors. The final model is depicted in Figure 2.

Insert Figure 2 about here

Predicting Stage of Change

A series of discriminant function analyses (DFAs) were performed using a combination of

pros, cons and cognitive and affective attitudes as predictors of membership in one of the five

stages of change. The first DFA used only pros and cons as predictors for the five stages of

change; the second DFA used only cognitive and affective attitudes as predictors; the third DFA

Pros and Cons and Attitude 17

used all four scales as predictors. For each DFA missing data appeared to be random throughout

the groups and predictors, so listwise deletion was employed and the analyses were carried out on

all remaining cases.

Predictors in the first DFA were pros and cons. Only the first discriminant function was

significant, χ2(8) = 89.3, p < .001, accounting for 32.2% of the total variance. Using procedures

described by Huberty (1984), classification rates for each stage of change were examined and

found significantly greater than chance for the precontemplation (z = 9.73, p < .001),

contemplation (z = 3.54, p < .001), and action (z = 8.05, p < .001) stages (see Figure 3).

However, the pros and cons correctly classified only 31.2% of the total cases, which was not

significantly greater than the chance rate of 27.7% (z = 1.17, ns).

Insert Figure 3 about here

The second DFA used cognitive and affective attitudes as predictors for the five stages of

change. Again, only the first discriminant function was significant, χ2(8) = 85.5, p < .001,

accounting for 30.5% of the total variance. The predictors correctly classified 37.5% of the cases

(z = 3.19, p < .01). Classification rates for the individual stages of change were significantly

greater than chance only for the precontemplation (z = 10.54, p < .001) and maintenance

(z = 6.93, p < .001) stages (see Figure 3).

All four scales were included as predictors in the third DFA. Both the first, χ2(16) =

150.6, p < .001, and second, χ2(9) = 32.9, p < .001, discriminant functions were significant,

respectively accounting for 42.2% and 13.4% of the total variance. Pros, cons and cognitive and

affective attitudes together correctly classified 48.2% of the cases (z = 6.72, p < .001).

Pros and Cons and Attitude 18

Classification rates for the stages of change were significantly greater than chance for the

precontemplation (z = 10.43, p < .001), contemplation (z = 4.18, p < .001), action (z = 6.85,

p < .001), and maintenance (z = 4.89, p < .001) stages (see Figure 3).

Chi-square difference tests indicated that: 1) the increase in variance accounted for by the

addition of the cognitive and affective attitude to pros and cons was statistically significant, χ2(8)

= 61.3, p < .001; and 2) the increase in variance accounted for by the addition of pros and cons to

cognitive and affective attitude was also statistically significant, χ2(8) = 65.1, p < .001

(Tabachnick & Fiddell, 1996).

Discussion

The stage distribution from the current study revealed a large percentage of individuals in the

preparation stage, and a lower percentage in precontemplation and contemplation, compared to other

college samples (Pinto & Marcus, 1995; Lee, Nigg, DiClemente, & Courneya, 2000). Owing to the

voluntary nature of the study, non-exercisers may not have wanted to complete a questionnaire about

exercise, therefore, study participants may be a more motivated portion of the sampled population. In

addition, although it was made clear at the beginning of each questionnaire that only leisure time

exercise should be considered, this point could have been misinterpreted. This seemingly

disproportionate representation of the stages is not considered a major limitation, as the focus of this

study was to investigate variable relationships and not supply accurate estimates of the population

prevalence in each stage. However, it is suggested that in the future more representative sampling

should be employed (e.g., random digit dialing).

Development of the decisional balance measure resulted in an internally reliable two-

component scale, consistent with other literature that has assessed the pros and cons of exercise

Pros and Cons and Attitude 19

(Marcus, Eaton, Rossi, & Harlow, 1994; Marcus, Rakowski, & Rossi, 1992; Nigg & Courneya,

1998; Prochaska et al., 1994). Some items from the present study were the same as those used by

Marcus and colleagues (1992; Marcus & Owen, 1992); however, the content of the final scale

differed somewhat from the scale developed on a working adult sample, reflecting value

differences placed on individual benefits or costs of exercising by the two samples. It is essential

that breadth of construct is present in a scale attempting to address the decision making

component, especially since conceptual frameworks (Janis & Mann, 1977) point to broader

interpretations. The developed scale represents seven of the eight categories described by Janis

and Mann (1977). The only category unaccounted for is approval from others. More importantly

both underlying mechanisms of the decision making process, self-persuasion and emotional

inoculation described by Janis and Mann (1977), are represented. This may indicate that the

structure underlying the decision making categories may be reflect the process, not different

categories of pros or cons. Therefore, the eight categories may be useful to guide interventions

whereas for assessment only the two processes are necessary.

Across the stages, the pros increased significantly from precontemplation to preparation.

The intersection of the pros and cons (the decisional balance point) was between contemplation

and preparation. The pros and decisional balance point replicated previous findings (e.g.,

Prochaska et al., 1994; Marcus, Rakowski, & Rossi, 1992). However, the cons did not differ

across the stages, which may be due to a combination of: 1) the homogeneous sample; 2) the

educated nature of the sample (college sample with an average of 13.5 years education); or 3)

power limitations in the early stages. This considered, the scale outperformed the decisional

balance measure used by Marcus, Rakowski and Rossi (1992), explaining approximately 35% of

Pros and Cons and Attitude 20

the variance across stages in a college sample versus 25% in the working adult sample. Of course,

this difference may be due to the difference in samples as much as to the change in items, but does

suggest that measurement sensitivity can be enhanced when scale development is targeted to

specific populations.

The two attitude subscales also performed similarly compared to previous studies. The

combined attitude components accounted for 33% of the variance across stage, replicating

findings of other exercise literature (Godin, 1993, 1994). The relationship of attitude to exercise

stage of change was linear across the stages of change for both subscales.

One interesting finding in this regard involves the pattern of results illustrated in Figure 3.

The DFA classification analyses revealed that decisional balance was predictive of membership in

the precontemplation, contemplation, and action stages. This is consistent with previous results

which suggest that the pros and cons are primarily useful in predicting movement in the earlier

stages of change (Prochaska, DiClemente, Velicer, Ginpil, & Norcross, 1985; Prochaska, Velicer,

Guadagnoli, Rossi, & DiClemente (1991); Prochaska et al., 1994; Velicer et al., 1985). In this

context, it is therefore not too surprising that the overall DFA classification rate for all five stages

of change was not statistically significant. Use of the two attitude variables alone as predictors

accounted for about the same amount of total variance as use of decisional balance alone, but was

successful at predicting membership only in the precontemplation and maintenance stages. The

combination of the decisional balance and attitude variables was much more successful than either

construct set alone, accounting for more than 50% of the total variance and successfully

predicting membership in all stages of change except for the preparation stage. Failure of any the

scales to predict preparation stage membership may be due to the fact that, by definition, this

Pros and Cons and Attitude 21

stage might be expected to be relatively less stable and more transitory, representing a transition

phase from contemplation to action (DiClemente et al., 1991). In general, precontemplation and

maintenance are considered to be the most stable stages of change whereas the middle stages —

contemplation, preparation, and action — are thought of as relatively less stable. These middle

stages were better predicted by pros and cons than by attitudes. These results could indicate that

the middle stages represent a time when attitudes are being formed or modified (and decisions are

being made) and are, therefore, too variable or inconsistent to be predictive of stage of change.

The results from the univariate post hoc tests shown in Table 3 also provide some support for this

hypothesis. While the pros differentiate the two early stages of change from the others, they do

not show any mean differences between the later three stages. The two attitude subscales do not

appear to differentiate the early stages of change, but do show mean differences between the

middle and later stages of change.

Although conceptually related, pros, cons and the two attitude subscales were relatively

distinct factors, indicating that the constructs were not multicollinear or redundant. In addition,

while the definition of attitude in the theory of planned behavior is conceptually and semantically

similar to pros and cons, they are not closely linked at a construct-measurement level. This was

supported by the final DFA, which illustrated that the addition of the attitude variables to pros and

cons increased the overall explained variance across stages of change by more than 23% (from

32.2% to 55.6%) and improved the predictive ability of pros and cons alone by 17% (from 31.2%

to 48.2%). The measurement model and DFA results taken in combination strongly suggest that

the addition of cognitive and affective attitudes may strengthen the decision-making aspect of the

TTM.

Pros and Cons and Attitude 22

A possible explanation for the difference in these constructs is that the decisional balance

item set was developed based on the eight categories outlined by Janis and Mann (1977): gains

and losses to self, which address all cognitive effects expected to result from the decision with

regards to individual utilitarian objectives; gains and losses to important others, which targets

individuals or groups with whom the individual is affiliated; self-approval and disapproval, which

includes one’s basic morals and values; and approval and disapproval of important others, which

targets the basic morals and values of individuals or groups with whom the individual is

associated. The attitude items, in contrast, addressed cognitive and affective components of

exercising. Although all eight dimensions of the Janis and Mann model were not represented in

the final pros and cons scales, the two constructs (decisional balance and attitudes) were not

developed from the same conceptual basis.

The difference in these constructs can perhaps be understood from an expectancy-value

model perspective. Two individuals who associate the same set of consequences with performing

a given behavior may hold different attitudes toward the behavior if they evaluate the

consequences differently. By the same token, people who have different sets of salient beliefs may

nonetheless have the same attitude. That is to say, one or more of a person’s beliefs can change

and yet his attitude may remain the same (Ajzen & Fishbein, 1980).

There may even be evidence in the disparity of the two response-style formats. A bipolar

scale of attitude may not correspond to that of the unipolar Likert format used for decisional

balance (Fishbein & Ajzen, 1975). In addition, scales providing a continuum of values may

indicate that the researcher has a bipolar conceptualization of the respective dimension, whereas

scales that present only positive values may indicate a unipolar conceptualization (Streiner &

Pros and Cons and Attitude 23

Norman, 1995). Although the attitude scales did not include positive or negative values,

respondents may have interpreted the endpoints as “negative” or “positive,” in order to

disambiguate the numeric values (Schwarz & Hippler, 1995; Schwarz, Knaüper, Hippler, Noelle-

Neuman, & Clark, 1991).

Some limitations need to be considered when interpreting these results. The cross-

sectional nature of the study does not allow for the inference of causality. The small cell sizes in

precontemplation and contemplation limit the power of the more advanced statistical analyses

used in this study. The assessments utilized self-report data and included no objective information

on actual exercise behavior. To address the ultimate goal of this line of research — behavior

change and associated health benefits — appropriate objective outcome measures should be

included.

Future research in this area should replicate the attitude variables in conjunction with the

pros and cons across the stages of change to determine if they add meaningfully to the

understanding of exercise behavior across different sample populations. In addition, it is important

to assess the relationship between attitudes and each stage of change to ascertain whether

attitudes are predictive of movement between all adjacent stage pairs or only certain ones. This

type of analysis is also recommended for the TTM variables and should be conducted

longitudinally. Furthermore, longitudinal investigations would allow an accurate assessment of the

temporal order of the constructs addressed in this investigation. For example, pros and cons may

precede attitudes leading to behavior, or pros and cons and attitudes may directly influence

behavior at the same level. Longitudinal designs can also ascertain whether the addition of attitude

Pros and Cons and Attitude 24

to the transtheoretical framework is predictive of stage movement and, ultimately, exercise

behavior.

This study demonstrates the importance of investigating other constructs and variables in

conjunction with the TTM in order to test the completeness of the approach and as part of an

ongoing evolutionary process. It is also important to recognize the value of integrating elements

of successful behavior-change models to produce the most effective outcomes. Although the

integration of attitudes into the TTM would require a more complete set of theoretical variables

(e.g., to what extent does attitude add to the TTM when processes of change and self-efficacy are

also included), these initial results seem worthy of further investigation. Furthermore, researchers

working with other change models should perhaps explore the integration of TTM variables into

their work in order to enhance efficacy and, ultimately, the development of interventions to

increase exercise adoption.

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human

Processes, 50, 179-211.

Ajzen, I., & Driver, B. l. (1992). Application of the theory of planned behavior to leisure

choice. Journal of Leisure Research, 24, 207-224.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.

Englewood Cliffs: NJ: Prentice-Hall.

Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behavior: Attitudes,

intentions and perceived behavioral control. Journal of Experimental and Social Psychology, 22,

453-474.

Pros and Cons and Attitude 25

Bentler, P. M. (1989). EQS: Structural equations program manual. Los Angeles, CA:

BMDP Statistical Software.

Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological

Bulletin, 107, 256-259.

Cohen, J. (1977). Statistical power analysis for the behavioral sciences. New York: Academic

Press.

Courneya, K. S. (1995). Understanding readiness for regular physical activity in older

individuals: An application of the theory of planned behavior. Health Psychology, 14, 80-87.

Courneya, K. S., Estabrooks, P. A., & Nigg, C. R. (1997). Predicting change in exercise

stage over a three-year period: An application of the theory of planned behavior. Avante, 3,

1-13.

Courneya, K. S., & McAuley, E. (1995). Cognitive mediators of the social influence-

exercise adherence relationship: A test of the theory of planned behavior. Journal of Behavioral

Medicine, 18, 499-515.

Courneya, K. S., Nigg, C. R., & Estabrooks, P. A. (1998). Relationships among the

theory of planned behavior in older persons over a three-year period. Psychology and Health, 13,

355-367.

Comrey, A. L. (1988). Factor-analytic methods of scale development in personality and

clinical psychology. Journal of Consulting and Clinical Psychology, 56, 754-761.

Crites, S. L., Fabrigar, L. R., & Petty, R. E. (1994). Measuring the affective and cognitive

properties of attitudes: Conceptual and methodological issues. Personality and Social Psychology

Bulletin, 20, 619-634.

Pros and Cons and Attitude 26

DiClemente, C. C., Prochaska, J. O., Fairhurst, S. K., Velicer, W. F., Velasquez, M. M.,

& Rossi, J. S. (1991). The process of smoking cessation: An analysis of precontemplation,

contemplation, and preparation stages of change. Journal of Consulting and Clinical Psychology,

59, 295-304.

DiClemente, C. C., Prochaska, J. O., & Gibertini, M. (1985). Self-efficacy and the stages

of self-change in smoking. Cognitive Therapy and Research, 9, 181-200.

Dishman, R. K. (1994). The measurement conundrum in exercise adherence research.

Medicine and Science in Sports and Exercise, 26, 1382-1390.

Eagly, A. H., Mladinic, A. O., & Otto, S. (1994). Cognitive and affective bases of

attitudes toward social groups and social policies. Journal of Experimental Social Psychology, 30,

113-137.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction

to theory and research. Reading, MA: Addison-Wesley.

Godin, G. (1993). The theories of reasoned action and planned behavior: Overview of

findings, emerging research problems and usefulness for exercise promotion. Journal of Applied

Sport Psychology, 5, 141-157.

Godin, G. (1994). Theories of reasoned action and planned behavior: Usefulness for

exercise promotion. Medicine and Science in Sports and Exercise, 26, 1391-1394.

Gorely, T., & Gordon, S. (1995). An examination of the transtheoretical model and

exercise behavior in older adults. Journal of Sport and Exercise Psychology, 17, 312-324.

Horn, J. I. (1965). A rationale and test for the number of factors in factor analysis.

Psychometrika, 30, 179-185.

Pros and Cons and Attitude 27

Horwath, C. C. (1999). Applying the transtheoretical model to eating behavior change:

Challenges and opportunities. Nutrition Research Review, 12, 281-317.

Huberty, C. J. (1984). Issues in the use and interpretation of discriminant analysis.

Psychological Bulletin, 95, 156-171.

Jackson, D. N. (1970). A sequential system for personality scale development. In C. D.

Spielberger (Ed.), Current topics in clinical and community psychology, Vol. 2 (pp. 61-96). New

York: Academic Press.

Jackson, D. N. (1971). The dynamics of structured personality tests: 1971. Psychological

Review, 78, 229-248.

Janis, I. L., & Mann, L. (1977). Decision making: A psychological analysis of conflict,

choice and commitment. New York: Free Press.

Joseph, J., Breslin, C., & Skinner, H. (1999). Critical perspectives on the transtheoretical

model and stages of change. In J. A. Tucker, D. M. Donovan, & G. A. Marlatt (Eds.), Changing

addictive behavior: Bridging clinical and public health strategies. New York: Guilford.

Lee, C. (1993). Attitudes, knowledge, & stages of change: A survey of exercise patterns

in older Australian Women. Health Psychology, 12, 476-480.

Lee, R. E., Nigg, C. R., DiClemente, C. C., & Courneya, K. S. (2000). Validating the

stages of change for exercise behavior with adolescents. Annals of Behavioral Medicine, 22, S80.

Marcus, B. H., Banspach, S. W., Lefebvre, R. C., Rossi, J. S., Carleton, R. A., & Abrams,

D. B. (1992). Using the stages of change model to increase the adoption of physical activity

among community participants. American Journal of Health Promotion, 6, 424-429.

Pros and Cons and Attitude 28

Marcus, B. H., Eaton, C. A., Rossi, J. S., & Harlow, L. L. (1994). Self-efficacy, decision-

making, and stages of change: An integrative model of physical exercise. Journal of Applied

Social Psychology, 24, 489-508.

Marcus, B. H., & Owen, N. (1992). Motivational readiness, self-efficacy and decision-

making for exercise. Journal of Applied Social Psychology, 22, 3-16.

Marcus, B. H., Rakowski, W., & Rossi, J. S. (1992). Assessing motivational readiness and

decision making for exercise. Health Psychology, 11, 257-261.

Marcus, B. H., Selby, V., Niaura, R., & Rossi, J. S. (1992). Self-efficacy and the stages of

exercise behavior change. Research Quarterly for Exercise and Sport, 63, 60-66.

Marcus, B. H., & Simkin, L. R. (1993). The stages of exercise behavior. Journal of Sports

Medicine and Physical Fitness, 33, 83-88.

Marsh, H. W., Balla, J. R. & McDonald, R. P. (1988). Goodness-of-fit indexes in

confirmatory factor analysis: The effect of sample size. Psychological Bulletin, 103, 391-410.

Marsh, H. W., Richards, G. E., Johnson, S., Roche, L., & Tremayne, P. (1994). Physical

self-description questionnaire: Psychometric properties and a multitrait-multimethod analysis of

relations to existing instruments. Journal of Sport & Exercise Psychology, 16, 270-305.

McAuley, E., & Courneya, K. S. (1993). Adherence to exercise and physical activity as

health-promoting behaviors: Attitudinal and self-efficacy influences. Applied and Preventive

Psychology, 2, 65-77.

Murray, C. J. L., & Lopez, A. D. (Eds.). (1996). The global burden of disease: A

comprehensive assessment of mortality and disability form diseases, injuries and risk factors in

1990 and projected to 2020. Cambridge, MA: Harvard University Press.

Pros and Cons and Attitude 29

National Center for Health Statistics. (1997). Top ten U.S. population causes of death.

Monthly Vital Statistics Report, 45(11S2)(23), 40-43.

Nigg, C. R. (1999). Predicting, explaining and understanding adolescent exercise behavior

using longitudinal and cross-sectional approaches. Unpublished doctoral dissertation, University

of Rhode Island, Kingston, RI.

Nigg, C. R., & Courneya, K. S. (1998). Transtheoretical model: Examining adolescent

health behavior. Journal of Adolescent Health, 22, 214-224.

Nigg, C. R., Rossi, J. S., Norman, G. J. & Benisovich, S. V. (March, 1998). Structure of

decisional balance for exercise adoption. Poster presented at the 19th annual meeting of Society

of Behavioral Medicine, New Orleans, LA.

Pinto, B. M., & Marcus, B. H. (1995). A stages of change approach to understanding

college students' physical activity. Journal of American College Health, 44, 27-29.

Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self-change of

smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology,

51, 390-395.

Prochaska, J. O., & DiClemente, C. C. (1984). Self-change processes, self-efficacy and

decisional balance across five stages of smoking cessation. In Advances in cancer control – 1983.

New York: Alan R. Liss, Inc.

Prochaska, J. O., DiClemente, C. C., & Norcross, J. (1992). In search of how people

change. American Psychologist, 47, 1102-1114.

Prochaska, J. O., DiClemente, C. C., Velicer, W. F., Ginpil, S., & Norcross, J. C. (1985).

Predicting change in smoking status for self-changers. Addictive Behaviors, 10, 395-406.

Pros and Cons and Attitude 30

Prochaska, J. O., Redding, C. A., & Evers, K. E. (1997). The transtheoretical model and

the stages of change. In K. Glanz, F. Marcus Lewis, and B. K. Rimer, Health behavior and health

education: theory, research and practice (2nd ed.) (pp. 60-84). San Francisco, CA: Jossey-Bass.

Prochaska, J. O., & Marcus, B. H. (1994). The transtheoretical model: Applications to

exercise. In R. K. Dishman (Ed.), Advances in Exercise Adherence (pp. 161-180). Champaign,

IL: Human Kinetics.

Prochaska, J. O., Velicer, W. F., DiClemente, C. C., & Fava, J. (1988). Measuring

processes of change: Applications to the cessation of smoking. Journal of Consulting and Clinical

Psychology, 56, 520-528.

Prochaska, J. O., Velicer, W. F., Guadagnoli, E., Rossi, J. S., & DiClemente, C. C.

(1991). Patterns of change: Dynamic typology applied to smoking cessation. Multivariate

Behavioral Research, 26, 83-107.

Prochaska, J. O., Velicer, W. F., Rossi, J. S., Goldstein, M. G., Marcus, B. H., Rakowski,

W., Fiore, C., Harlow, L. L., Redding, C. A., Rosenbloom, D., & Rossi, S. R. (1994). Stages of

change and decisional balance for 12 problem behaviors. Health Psychology, 13, 39-46.

Rakowski, W., Andersen, R. M., Stoddard, A. M., Urban, N., et al. (1997). Confirmatory

analysis of opinions regarding pros and cons of mammography. Health Psychology, 16, 433-442.

Reed, G., Velicer, W. F., Prochaska, J. O., Rossi, J. S. & Marcus, B. H. (1997). What

makes a good staging algorithm: Examples from regular exercise. American Journal of Health

Promotion, 12, 57-66.

Rosenberg, M. J., & Hovland, C. I. (1960). Cognitive, affective and behavioral

components of attitude. In M. J. Rosenberg & C. I. Hovland (Eds.), Attitude organization and

Pros and Cons and Attitude 31

change: An analysis of consistency among attitude components. New Haven, CT: Yale University

Press.

Schwarz, N., & Hippler, H.-J. (1995). The numeric values of rating scales: A comparison

of their impact in mail surveys and telephone interviews. International Journal of Public Opinion

Research, 7, 72-74.

Schwarz, N., Knaüper, B., Hippler, H.-J., Noelle-Neuman, E., & Clark, L. (1991). Rating

scales: Numeric value may change the meaning of scale labels. Public Opinion Quarterly, 55, 570-

582.

Streiner, D. L., & Norman, G. R. (1995). Health measurement scales: A practical guide to

their development and use (2nd ed.). New York: Oxford University Press.

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New

York: HarperCollins.

Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship:

A role for group identification. Personality and Social Psychology Bulletin, 22, 776-793.

Trafimow, D., & Sheeran, P. (1998). Some tests of distinction between cognitive and

affective beliefs. Journal of Experimental Social Psychology, 34, 378-397.

U.S. Department of Health and Human Services. (2000, 12 January). Healthy People 2010

Fact Sheet [Online]. Office of Disease Prevention and Health Promotion. Available:

http://www.health.gov/healthypeople/ [2000, 7 February].

Velicer, W. F. (1976). Determining the number of components from the matrix of partial

correlations. Psychometrika, 41, 321-327.

Pros and Cons and Attitude 32

Velicer, W. F., DiClemente, C. C., Prochaska, J. O., & Brandenburg, N. (1985).

Decisional balance measure for assessing and predicting smoking status. Journal of Personality

and Social Psychology, 48, 1279-1289.

Weinstein, N. D., Lyon, J. E., Sandman, P. M., & Cuite, C. L. (1998). Experimental

evidence for stages of health behavior change: The precaution adoption process model applied to

home radon testing. Health Psychology, 17, 445-453.

Zwick, W. R., & Velicer, W. F. (1986). A comparison of five rules for determining the

number of components to retain. Psychological Bulletin, 99, 432-442.

Pros and Cons and Attitude 33

Table 1

Statement Wording and Factor Loadings for Decisional Balance Scale

Statement Wording Factor Loading

PROS

I would feel more confident if I exercised regularly. .81

I would feel proud of myself if I exercised. .84

I would feel less stressed if I exercised. .80

Exercising puts me in a better mood for the rest of the day. .82

I would feel more comfortable with my body. .81

I would have more self-respect if I exercised. .77

Regular exercise would help me have a more positive outlook on life.

.83

I would look better if I exercised. .81

I feel better about myself. .82

I feel a sense of accomplishment after I exercise. .82

CONS

I feel uncomfortable or embarrassed in exercise clothes. .71

My friends do not like when I exercise. .81

My friends and/or family would say I spend too much money on exercise-related things.

.73

There is too much I would have to learn to exercise. .78

I am afraid to find that I am not good at exercising. .76

After exercising, my family and friends have to wait for me to shower.

.73

Others make fun of me when I exercise. .77

Exercise puts an extra burden on my significant other. .78

Others say I look stupid in my exercise outfit. .84

I would feel selfish if I exercised regularly. .74

Pros and Cons and Attitude 34

Table 2

Standardized Scores for Pros, Cons and Attitude by Stage of Exercise Adoption

Stage of Adoption

Precontemplationa Contemplationb Preparationc Actiond Maintenancee

M SD M SD M SD M SD M SD

Pros 33.45 7.98 43.41 10.22 52.10 8.20 55.84 7.39 52.08 8.07

Cons 50.30 9.13 49.32 9.10 48.20 7.63 52.10 11.26 51.17 12.48

Cognitive 38.19 7.97 42.60 10.13 48.67 9.98 53.59 6.31 55.99 6.13

Affective 40.49 11.76 44.17 10.34 47.69 9.87 53.29 7.03 56.51 5.39

an=16. bn=24. cn=87. dn=27. en=66.

Pros and Cons and Attitude 35

Table 3

Analysis of Variance and Tukey Follow-up Results for Pros, Cons and Attitude by Stages of

Exercise Adoption

Source df F η2 Tukey’s HSD*

Pros 4, 218 24.00** 0.32 PC < C < PR, A, M Cons 4, 218 1.47 0.02 N/A Cognitive Attitude 4, 219 20.97** 0.29 PC, C < PR, A, M

PR < M Affective Attitude 4, 219 20.15** 0.27 PC < PR, A, M

C, PR < A, M

Note: * = p < .05; ** = p < .001. PC=precontemplation; C=contemplation; PR=preparation; A=action; M=maintenance.

Pros and Cons and Attitude 36

Table 4

Correlations Between Pros, Cons and Attitude Constructs

Correlations

Cons

Cognitive Attitude

Affective Attitude

Pros -.032 .313* .267* Cons — -.015 -.002 Cognitive Attitude — — .780* Affective Attitude — — —

Note: * p < .01

Pros and Cons and Attitude 37

Table 5

Fit Indices for Tested Models

Model χ2 df χ2/df CFI AASR

Model 1: Null 2306.47 66 39.95 — — Model 2: One Factor 1532.27 54 28.38 0.34 0.12 Model 3: Uncorrelated Two-Factor 1013.40 54 18.77 0.57 0.12 Model 4: Uncorrelated Three-Factor 368.78 54 6.83 0.86 0.09 Model 5: Correlated Three-Factor 356.75 53 6.73 0.86 0.06 Model 6: Uncorrelated Four-Factor 252.10 54 4.67 0.91 0.12 Model 7: Correlated Attitude Subscales 174.20 53 3.29 0.95 0.08 Model 8: Correlated Four-Factor 142.94 51 2.80 0.96 0.04

Note: df=degrees of freedom; CFI=comparative fit index; AASR=average absolute standardized residuals

Pros and Cons and Attitude 38

Figure Captions

Figure 1. T-Scores for pros, cons and cognitive and affective attitude by stage of exercise

adoption.

Figure 2. Structural equation modeling results of Model 8 (Correlated Four-Factor).

Figure 3. Percent stage of change membership correctly predicted by decisional balance and

attitude variables in discriminant function classification analyses.

Pros and Cons and Attitude 39

30

35

40

45

50

55

60

PC C P A M

Stages of Change

T-S

core

s (M

ean=

50)

Pros

ConsCognitiveAffective

Pros and Cons and Attitude 40

Note: *** p < .001

Pros and Cons and Attitude 41

70.6

33.331.0

63.0

7.6

75.0

12.5

21.3

14.3

68.7

75.0

37.5 36.8

55.6 57.6

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

PC C PR A M

Stages of Change

% M

embe

rshi

p P

redi

cted

Series1

Series2

Series3

Pros & Cons

Cognitive & Affective Attitudes Pros & Con + Cognitive & Affective Attitude