Weekly variability in outcome expectations: Examining associations with related physical activity experiences during physical activity initiation

Abstract

Little is known about how outcome expectations change after physical activity initiation and whether changes are associated with physical activity experiences. In a diary study, physically inactive adults (N = 102) initiated an exercise regimen and reported their experiences daily (e.g. progress toward goals) and corresponding outcome expectations weekly (e.g. how much progress they expect this week). Average levels (between-person effects) for eight experiences (ps < .01) and deviations from the average levels (within-person effects) for three experiences (ps < .05) were associated with changes in outcome expectations. The findings demonstrate that outcome expectations for exercise vary over time and are associated with people’s subjective experiences.

Keywords

beliefs exercise exercise behavior experience health behavior

The decision to engage in regular physical activity is influenced, in part, by the outcomes the person expects will result. Expectations that regular physical activity will lead to weight loss, improved mood, and fewer future health problems serve to motivate the person’s behavioral efforts. Outcome expectations are the beliefs that specific positive and negative outcomes will result from a given behavior with positive expectations serving as incentives and negative expectations as disincentives for the behavior (Bandura, 1997). Conceptualizations of these beliefs are included as central constructs in most models of health behavior (cf. Williams et al., 2005). Although evidence for the influence of outcome expectations in physical activity is not entirely consistent, findings have shown that holding favorable outcome expectations (e.g. exercise will improve my mood) is predictive of greater motivation (Gellert et al., 2012), higher self-efficacy (Resnick et al., 2000), and higher physical activity levels (Williams et al., 2005). There is also evidence that differences between people’s initial outcome expectations and their achievement (i.e. outcome realizations) are predictive of future activity levels (Sears and Stanton, 2001; Wilcox et al., 2006). This evidence suggests that between-person differences in outcome expectations are important in explaining interindividual variability in physical activity behavior. Yet, little is known about how outcome expectations change over time.

Outcome expectations are theorized to be dynamic and influenced by experiences with the relevant behavior (Bandura, 1997; Williams, 2010). Consider someone who is about to initiate physical activity and has the expectation that her mood will improve. If she perceives her mood has not changed after a few weeks of regular physical activity, it is possible that her expectations for how physical activity will affect her mood will also change. Because proximal expectations provide motivation for immediate behavior (Ajzen, 1991), a recently changed outcome expectation (e.g. “exercise will not improve my mood”) is more relevant to immediate motivation and behavior than the expectation formed prior to initiating regular exercise. There is some evidence demonstrating that outcome expectations can change within the first few weeks after initiating physical activity (Cramp and Brawley, 2009; Papandonatos et al., 2011), but little is known about factors that influence these changes in outcome expectations. The objective of this study is to elucidate how outcome expectations change after physical activity initiation.

Physical activity experiences and outcome expectations

According to theoretical work on behavior change, people are sensitive to changes in their experiences during behavior change initiation (Rothman et al., 2009, 2011). Evidence across different behavioral domains supports this theoretical proposition. Social-cognitive factors such as satisfaction (Baldwin et al., 2009a, 2009b) and self-efficacy (Anderson et al., 2010; Gwaltney et al., 2005a) have been shown to be systematically associated with people’s behavior change experiences. For example, Baldwin et al. (2009a) found that people’s experiences (e.g. positive feedback from others’ frequency of cravings) with smoking cessation systematically covaried over time with their satisfaction. In addition, Gwaltney et al. (2005a) have shown that variability in daily self-efficacy is associated with people’s cessation experiences such as negative mood and momentary urges to smoke. Moreover, this daily variability has been shown to predict the onset of smoking relapse in adults (Gwaltney et al., 2005b) and adolescents (Van Zundert et al., 2010).

It is possible that outcome expectations for physical activity are likewise sensitive to shifts in experiences with recently initiated activity. Although theoretical models suggest that people actively monitor their behavioral, psychological, and physiological experiences during a behavior change (Leventhal et al., 2008; Rothman et al., 2011), the models do not specify which experiences people monitor. In considering this issue, it is important to note that people expect a variety of positive and negative experiences from physical activity, including positive emotions (Lox et al., 2000), enjoyment (Kendzierski and DeCarlo, 1991), sense of accomplishment, weight control, and improving attractiveness (Cash et al., 1994), as well as negative experiences such as body soreness. It is plausible that experiences associated with these expectations result in changes in expectations. To date, however, no study has examined whether experiences with physical activity are related to changes in the specific outcome expectations tied to those experiences.

Physical activity interventions, and the theoretical models that guide them (Lally and Gardner, 2011; Nigg et al., 2008; Schwarzer, 2008), implicitly assume that relevant constructs, such as outcome expectations, change within-person. However, research to date has focused on between-person differences in outcome expectations and, as a result, within-person associations have gone largely unexamined (Dunton and Atienza, 2009). Between-person variability refers to interindividual differences. If the associations between physical activity experiences and outcome expectations are due to between-person variability, then the associations are due to differences in the average levels of experiences that vary between individuals. For example, people who tend to report positive experiences with physical activity may also have a tendency to report positive changes in outcome expectations. Within-person variability refers to intraindividual differences that fluctuate across situations. For example, if a person’s positive experiences differ from her own average level during a week, this fluctuation in the positive experiences may be associated with changes in her outcome expectations for the upcoming week, regardless of any between-person differences. Identifying changes in outcome expectations that are due to within-person variability would clarify which outcome expectations are most sensitive to fluctuations in people’s experiences with physical activity.

Current study

Using a diary study in a sample of previously inactive adults, we investigated how experiences with physical activity are associated with changes in their outcome expectations during the first 4 weeks of physical activity initiation. Specifically, we explored whether changes in nine different outcome expectations are associated with between- and/or within-person variability in specific physical activity experiences that are related to the outcome expectations.

Methods

Participants

Participants (N = 119) were volunteers recruited from the Dallas or Fort Worth area through public advertisements who reported being physically inactive. Participants who did not complete the baseline outcome expectations questionnaire or any outcome expectations measures after the baseline session were not included in the analyses (N = 17).

The sample (N = 102) was primarily female (76.5%), racially and ethnically diverse (34.3% Hispanic, 40.3% non-Hispanic White, 16.7% non-Hispanic Black, 4.9% Asian, and 2.9% Other), had a mean age of 33.5 years (standard deviation (SD) = 12.3 years, range = 18–61 years), and had a mean body mass index (BMI) of 27.8 (SD = 5.9) with similar numbers of normal weight (BMI < 25.0; 38.2%), overweight (BMI = 25.0–29.9; 24.5%), and obese (BMI ≥ 30.0; 36.3%) individuals. Nearly all participants (91.1%) had some education beyond high school, and 70.6 percent reported an annual income less than US$50,000. There were no significant differences between participants included in the analyses (N = 102) and those who were not included on gender, race, BMI, education, or income. However, those who were not included were significantly older (M = 40.2 years vs M = 33.5; F(1, 117) = 4.51, p = .04). As a result, we included age as a covariate in the models. There were no differences between the two groups in baseline outcome expectations.

Procedure

Eligibility screen

Interested individuals were screened for eligibility via telephone or online questionnaire. Eligibility criteria included the following: (a) being insufficiently active (<60 minutes/week of moderate-to-vigorous intensity activity across multiple days over the past month; US Department of Health and Human Services (USDHHS), 2008b), assessed using the physical activity items from the Behavioral Risk Factor Surveillance System (BRFSS; Centers for Disease Control and Prevention (CDC), 2009); (b) having a BMI < 40 and absence of cardiovascular, pulmonary, or metabolic disease to avoid enrolling those for whom physical activity may pose a medical risk (American College of Sports Medicine (ACSM), 2009); (c) Internet access at home; (d) access to exercise equipment and/or a location to exercise; and (e) report an interest in initiating regular activity. Respondents who met eligibility criteria were scheduled for a baseline assessment.

Baseline session

A research assistant (RA) consented participants and provided instructions about initiating physical activity consistent with public health guidelines using evidence-based information to facilitate the instructions (USDHHS, 2008a, 2008b). Participants made plans to exercise 150 minutes of moderate-intensity activity per week and completed a series of questionnaires that included outcome expectations for the upcoming week.

Daily diaries

Beginning the day after the baseline session, participants completed daily diaries for 28 days using an online questionnaire (Qualtrics, Inc.) that included self-reported physical activity and physical activity experiences. Participants were sent an email with a link to the questionnaire with instructions to complete it prior to retiring for the night. However, participants could complete it before 12:00 p.m. the following day for the response to be considered timely (Gable and Poore, 2008), verified via electronic time stamps. Only timely responses were included in analyses. Participants who did not complete a diary on time were sent an email to inquire about any difficulties. There were 2856 possible daily diaries during the study, and participants completed 80.0 percent (2285 diaries) of the daily assessments on time. This rate is similar to other studies that have used daily assessments in physical activity (e.g. Dunton et al., 2009, 76% completion rate). The mean number of days completed was 22.4 (SD = 5.99, median = 24), and the range was 3–28.

Weekly questionnaires

Participants also completed weekly assessments sent on Days 7, 14, 21, and 28 via online questionnaire. In these questionnaires, participants reported their outcome expectations for the upcoming week using the same outcome expectation items assessed during the baseline session. Participants were instructed to complete it prior to retiring for the night; however, they could complete it within 48 hours for the response to be considered timely, verified via electronic time stamps. Only timely responses were included in analyses. Individuals who did not complete the weekly questionnaire on time were contacted the following day and asked to complete the questionnaire as soon as possible. There were 408 possible weekly diaries during the 4-week study period, and participants completed 92.2 percent (376 diaries) of them on time. The mean number of weekly questionnaires completed was 3.69 (SD = 0.77, median = 4), and the range was 1–4. Participants were compensated up to US$120 for participating.

Weekly telephone contact

At the end of each week, an RA called participants to discuss physical activity plans for the upcoming week and addressed any questions or concerns, including incomplete questionnaires. Previous work has shown that brief telephone contact results in improved physical activity adherence (Castro and King, 2002).

Measures

Physical activity experiences (daily)

In the daily diaries, participants reported various behavioral, psychological, and physiological experiences that are common in regular physical activity. Given that current theoretical models do not specify which experiences people monitor during physical activity, we chose these specific items for analyses because they included both positive- and negative-valenced experiences and captured an array of experiences associated with regular physical activity (Reasons for Exercise Inventory, Cash et al., 1994; Physical Activity Enjoyment Scale, Kendzierski and DeCarlo, 1991; Physical Activity Affect Scale, Lox et al., 2000). Rather than conducting factor analyses for all 28 days, we conducted exploratory analyses, using principal axis factor analysis and oblique rotation (oblimin), on 4 days (Days 1, 8, 15, and 22) to investigate whether items loaded on the same factor consistently (see Baldwin et al., under review). At each time point, we excluded items that loaded on multiple factors or did not load on any factor; items in multi-item scales were consistent in their factor loadings across the four time points. The analyses resulted in the following experience scales: (a) positive experiences, (b) positive affect, (c) negative affect, (d) tranquility, (e) fatigue, and (f) perceived progress toward goals. Despite not loading on any factor, we also examined 3 single items that are relevant to physical activity initiation: (g) perceptions of exercise as a chore, (h) body soreness, and (i) quality of sleep. For analyses, we aggregated the experience variables across the week in which they were assessed to create week-level physical activity experiences to place them on the same metric as outcome expectation measures.

Positive experiences

Five items assessed positive experiences with physical activity (e.g. how much they currently enjoy exercising). Participants responded using a Likert-type scale ranging from 0 (not at all) to 8 (a great deal). Reliability was high for each week-level scale (αs = 0.90–0.92).

Positive/negative affect and tranquility/fatigue

We used scales to assess affect-related daily experiences. The positive affect items asked, “How upbeat/energetic/enthusiastic did you feel today?” The negative affect items asked, “How miserable/discouraged/crummy did you feel today?” The tranquility items asked, “How relaxed/peaceful/calm did you feel today?” Finally, the fatigue items asked, “How tired/worn-out/fatigued did you feel today?” Reliability was high for each week-level scale: positive affect (αs = 0.95–0.98), negative affect (αs = 0.92–0.95), tranquility (αs = 0.95–0.98), and fatigue (αs = 0.95–0.96).

Progress toward goals

Five items assessed experiences regarding progress toward goals: weight-related, improving physical fitness, physical attractiveness, overall health, and preventing health problems. Participants responded using a Likert-type scale ranging from 0 (not at all) to 8 (a great deal). Reliability was high for each week-level scale (αs = 0.98).

Exercise as a chore, body soreness, and quality of sleep

We assessed each of the following experiences with an individual item: (a) perceiving exercise as a chore (“How much do you currently feel like exercising is a chore?”), (b) body soreness (“How sore did your body feel today?”), and (c) quality of sleep (“How would you rate the quality of sleep you are currently getting at night?”). Participants responded using a Likert-type scale ranging from 0 (not at all) to 8 (a great deal/extremely) for perceiving exercise as a chore and body soreness and a scale ranging from −4 (very poor) to +4 (very good) for quality of sleep because this item had a clear neutral point.

Outcome expectations

In the weekly questionnaires, participants reported their outcome expectations for the upcoming week. All outcome expectation items matched with a corresponding experience item. For example, one question in the positive experience scale asked, “How much of a sense of accomplishment do you feel today about the fact that you are exercising regularly?” and participants responded on a Likert-type scale ranging from 0 (not at all) to 8 (a great deal). The corresponding outcome expectation asked, “Over the next week, how much do you expect exercising will affect the sense of accomplishment you feel?” and participants responded on the same 0 (not at all) to 8 (a great deal) scale. Scales for the outcome expectations matched the categories for the experiences scales. Reliability of the outcome expectation scales for each of the 4 weeks was adequate: positive experiences (α = 0.74–0.87.), positive affect (α = 0.90–0.94), negative affect (α = 0.83–0.93), tranquility (α = 0.80–0.97), fatigue (α = 0.86–0.93), and perceived progress toward goals (α = 0.89–0.96).

Covariates

We included daily physical activity minutes as a covariate to examine the associations between physical activity experiences and changes in outcome expectations controlling for any effect physical activity may have on the associations. For similar reasons, we also included age and race as covariates. Physical activity was reported as the total duration of daily moderate- and vigorous-intensity activity (≥10 minutes). Items were modified from the BRFSS (CDC, 2009) to reflect daily rather than weekly reporting. We converted self-reported minutes of moderate- and vigorous-intensity activity into metabolic equivalent task (MET) minutes (Ainsworth et al., 2000) by multiplying moderate-intensity minutes by 4.5 (the midpoint of the range for moderate-intensity activity (3.0 and 5.9 METs)) and vigorous-intensity minutes by 7.5 (1.5 METs greater than the minimum value for vigorous-intensity; see Morrow et al., 2011). We aggregated daily MET minutes across the week to create a week-level measure of MET minutes. We dichotomized the race variable (0 = non-Hispanic Caucasian, 1 = other), and age was a continuous covariate.

Results

We employed a two-level hierarchical linear modeling (HLM) approach using IBM SPSS Version 21.0 to examine whether changes in the specific outcome expectations were related to the experiences with physical activity. Longitudinal HLM allows the examination of relationships between variables both within individuals (Level 1) and between individuals (Level 2), allows all participants to be included in the analyses regardless of missing data, and accounts for the correlation between repeated measures within individuals over time. We analyzed nine separate models to examine each outcome expectation.

Variability in outcome expectations over time

To verify whether outcome expectations are variable from week to week, we tested whether the Level-1 variance of the null model (ϵ_ij) that included only the intercept was significant. We expected that ϵ_ij would be significantly different from 0, indicating that outcome expectations are not static from week to week. ϵ_ij was significant in each model (ps < .001), indicating that all nine outcome expectations varied from week to week.

Between- and within-person effects of physical activity experiences

The data consist of multiple assessments across time (Level 1) nested within individuals (Level 2). To examine the effect of experiences on outcome expectations, each experience variable was partitioned into a person-centered mean (i.e. average level of experiences over 4 weeks; EXPERm_i) and a deviation from this mean each week (i.e. within-person fluctuation in the experience; EXPERdev_ij) to analyze the between- and within-person effects of physical activity experiences (Hedeker and Gibbons, 2006). We predicted each outcome expectation from the previous week’s corresponding physical activity experience components (e.g. between- and within-person components of positive experiences predicting outcome expectations for positive experiences the next week) in separate models for each category of experiences. An unstructured covariance matrix for the errors of the repeated assessments was used for all models because its deviance statistic was the smallest of various covariance structures we examined, including autoregressive, compound symmetry, and Toeplitz covariance structures (Singer and Willett, 2003). We used maximum likelihood (ML) for model estimation because it produces accurate estimates of the fixed effects and variance components when the data are not limited by a small sample (Raudenbush and Bryk, 2002).

The Level 1 models also included the previous week’s outcome expectations (PrevWkOE_ij−1) to control for the possibility that the experiences might be related to current outcome expectations merely because both are related to previous week’s outcome expectations. Physical activity minutes (within-person deviation) were included as a covariate. The Level 1 portion of the HLM model was OE_ij = b_0i + b_1i × EXPERdev_ij + b_2i × PrevWkOE_ij−1 +b_3i × PAdev_ij + ϵ_ij, where i represents each individual subject; j represents the 4-weekly exercise diary assessments (Week 1–4); b_0i represents the intercept for individual i; b_1i, b_2i, and b_3i represent the slope of the change in the outcome for subject i across deviations of experiences, levels of previous week’s outcome expectations, and deviations of physical activity minutes, respectively; and ϵ_ij represents the error in predicting outcome j for individual i. Variables were standardized to put them on the same metric (Heck et al., 2010). Random effects were dropped if nonsignificant.

The Level 2 portion of the model accounts for the differences between individuals in their intercept and slope. To investigate the between-person effects of experience, the person-centered mean of each experience (EXPERm_i) was included in the Level 2 equation for b_0i to determine their relations to the outcome (Hedeker and Gibbons, 2006). Race, age, and the average level of physical activity minutes over the 4 weeks (PAm_i) were included as covariates. The Level 2 portion of the HLM model was b_0i = γ₀₀ + γ₀₁ × Race_i + γ₀₂ × Age_i + γ₀₃ × PAm_i +γ₀₄ × EXPERm_i + u_0i.

The analyses indicated there were significant between-person effects for most experiences on outcome expectations (Table 1). Eight of the experiences had a significant between-person effect on outcome expectations; tranquility (p = .54) was the exception. In all cases, average levels of the experiences people reported were positively associated with their outcome expectations. For example, higher average levels of positive experiences across the 4 weeks of the study were associated with higher expectations for positive experiences. People who differed by 1 SD in their average level of positive experiences had a .414 SD difference in their outcome expectations for positive experiences.

Table 1.

Multilevel model examining between- and within-person associations with outcome expectations during the first 4 weeks of initiating physical activity.

Model/experience variable	Coefficient	Standard error	z value	p value
Positive exercise experiences
Between-person association	.414	.035	11.97	<.001
Within-person association	.050	.052	.955	.341
Positive affect
Between-person association	.155	.041	3.76	.003
Within-person association	.079	.057	1.38	.174
Negative affect
Between-person association	.161	.049	3.26	.009
Within-person association	.103	.047	2.18	.033
Tranquility
Between-person association	.014	.022	.619	.539
Within-person association	.072	.077	.937	.351
Fatigue
Between-person association	.175	.042	4.20	<.001
Within-person association	.046	.044	1.04	.300
Progress toward goals
Between-person association	.333	.026	12.69	<.001
Within-person association	.072	.044	1.64	.102
Perceiving exercise as a chore
Between-person association	.362	.022	16.40	<.001
Within-person association	.105	.048	2.20	.03
Body soreness
Between-person association	.201	.039	5.18	<.001
Within-person association	.180	.041	4.40	<.001
Quality of sleep
Between-person association	.136	.030	4.51	<.001
Within-person association	−.040	.066	−0.61	.54

The analyses indicated that within-person deviations from average levels significantly predicted outcome expectations for three of the nine experiences: negative affect, body soreness, and perceptions of exercise as a chore (Table 1). The within-person increases in these experiences were associated with increases in the respective outcome expectations. For example, a 1 SD increase in negative affect above one’s own average level was associated with a .103 SD increase in the expectation for negative affect during the upcoming week.

Discussion

The findings provide the first evidence to date demonstrating that outcome expectations for physical activity are systematically associated with people’s subjective experiences during the first few weeks of physical activity initiation. Understanding how outcome expectations change is important because expectations developed during a behavior change are more relevant to immediate motivation and behavior than expectations formed prior to initiating the change (e.g. proximal expectations provide motivation for immediate behavior; Ajzen, 1991; see also Gwaltney et al., 2005b). The findings are consistent with theories characterizing outcome expectations as dynamic (Bandura, 1997) and with findings in other health domains suggesting that subjective experiences influence social cognitive factors that guide the behavior change (e.g. Anderson et al., 2010; Baldwin et al., 2009a, 2009b; Gwaltney et al., 2005a).

In demonstrating that outcome expectations for physical activity are responsive to people’s ongoing experiences, our findings extend what is known about ongoing decision-making in physical activity (Dunton and Atienza, 2009) by elucidating a process through which outcome expectations change. Controlling for the previous week’s outcome expectations, higher average levels of eight of the physical activity experiences (between-person effects) resulted in higher levels of outcome expectations tied to these experiences. Additionally, decreases from an individual’s average level of negative affect, body soreness, and perceiving exercise as a chore in a given week (within-person effects) resulted in lower expectations in these experiences for the upcoming week. The fact that only three of the experiences had significant within-person effects suggests that certain outcome expectations are sensitive to week-to-week changes in people’s experiences. Interestingly, the three experiences with significant within-person effects were all negatively valenced (negative affect, body soreness, and perception of exercise as a chore).

One explanation why outcome expectations are sensitive to within-person changes in negative, but not positive, experiences with physical activity is that there may be a negativity bias toward outcome expectations during physical activity initiation. Negativity bias is the tendency to pay attention to, learn from, and use negative information far more than positive information (Baumeister et al., 2001). Before people initiate regular physical activity, they are motivated by the positive outcomes that the new behavior will afford, and those expectations help to motivate their initial efforts (Brassington et al., 2002). Our findings suggest people are more attuned to ongoing increases and decreases in their negative experiences when reassessing their outcome expectations. However, a negativity bias during physical activity initiation does not generalize across all relations with experiences. For example, satisfaction is associated with both positive and negative physical activity experiences (Baldwin et al., under review). Future research is needed to clarify the extent to which a negativity bias exists during physical activity initiation.

The findings also have important implications for physical activity interventions and research. First, it is important to recognize that expectations during physical activity initiation are changing from week to week, and those expectations tied to negative experiences are particularly sensitive to weekly fluctuations in the experiences. Thus, helping people process negative physical activity experiences may be an important target for interventions to buffer their effect on negative outcome expectations. Future research and interventions should consider how these changing outcome expectations facilitate or impede continuing motivation for physical activity. The findings also suggest that to accurately capture the influence of outcome expectations, multiple assessments of outcome expectations are needed. Using a single baseline assessment of outcome expectations to predict a distal outcome is likely a misguided test of the effect of outcome expectations. Finally, our findings also suggest that positive and negative outcome expectations should be considered separately, rather than jointly, in order to clarify their influence on physical activity-related outcomes.

Limitations

There are several factors that may limit the study results. First, outcome expectations were measured during the first 4 weeks after initiating physical activity, and it is unclear whether the pattern of results would be consistent over a longer period of time. Second, future studies should clarify the role of outcome expectation change in the initiation of other health behaviors. For example, it would be important to clarify whether changes in outcome expectations for weight loss are observed during the first few weeks given that the primary outcome (amount of weight lost) typically takes more than a few weeks to attain. Third, although the outcome expectation categories we examined captured a broad array of experiences and were based on existing scales of experiences with physical activity (Cash et al., 1994; Kendzierski and DeCarlo, 1991; Lox et al., 2000), the categories were not exhaustive. There may be other expectations that are also sensitive to changes during the first few weeks after initiating regular activity (e.g. being able to cope with stress). Future research could examine the effects of other specific outcome expectations. Fourth, we enrolled participants who self-selected into the study who already held favorable expectations toward physical activity, and who otherwise they would not have likely self-selected to enroll. Thus, these results are limited to those who were previously inactive, but also demonstrate a desire to initiate regular activity. Fifth, most of participants in the analyses were women (77%) and were well educated (91% with some education beyond high school). Although these characteristics are not uncommon in physical activity studies, and the sample did include people of diverse ethnic or racial backgrounds and with wide range of baseline BMI, the sample is not entirely representative of the general population, and thus, the conclusions that can be drawn are limited accordingly. Finally, chronic health conditions may pose additional barriers to maintaining physical activity (e.g. arthritis pain), and future research is needed to explore the relationship between disease-related factors (e.g. pain acceptance) and outcome expectations (Gyurcsik et al., 2011).

Conclusions

The findings demonstrate that outcome expectations for physical activity are a dynamic construct varying from week to week and are associated with people’s ongoing subjective experiences. Negative outcome expectations, in particular, are sensitive to fluctuations in people’s experiences during physical activity initiation. This evidence is consistent with how outcome expectations are conceptualized theoretically (Bandura, 1997) and has important implications for assessment of outcome expectations in physical activity research.

Footnotes

Acknowledgements

Portions of this research were presented at the annual meeting of the Society of Behavioral Medicine, April 2012, New Orleans, LA, USA.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Ainsworth

Haskell

Whitt

. (2000) Compendium of physical activities: An update of activity codes and MET intensities. Medicine and Science in Sports and Exercise 32: S498–S516.

Ajzen

(1991) The theory of planned behavior. Organizational Behavior and Human Decision Processes 50: 179–211.

American College of Sports Medicine (ACSM) (2009) Guidelines for Exercise Testing and Prescription (8th edn). Philadelphia, PA: Lippincott Williams & Wilkins.

Anderson

Winett

Wojcik

. (2010) Social cognitive mediators of change in a group randomized nutrition and physical activity intervention: Social support, self-efficacy, outcome expectations and self-regulation in the guide-to-health trial. Journal of Health Psychology 15(1): 21–32.

Baldwin

Loehr

. (under review) Elucidating satisfaction with physical activity: An examination of the day-to-day associations between experiences with physical activity and satisfaction during physical activity initiation (manuscript under review).

Baldwin

Rothman

Jeffery

(2009a) Satisfaction with weight loss: Examining the longitudinal covariation between people’s weight-loss-related outcomes and experiences and their satisfaction. Annals of Behavioral Medicine 38: 213–224.

Baldwin

Rothman

Hertel

. (2009b) Longitudinal associations between people’s cessation-related experiences and their satisfaction with cessation. Psychology & Health 24: 187–201.

Bandura

(1997) Self-Efficacy: The Exercise of Control. New York: Freeman.

Baumeister

Bratslavsky

Finkenauer

. (2001) Bad is stronger than good. Review of General Psychology 5: 323–370.

10.

Brassington

Atienza

Perczek

. (2002) Intervention-related cognitive versus social mediators of exercise adherence in the elderly. American Journal of Preventive Medicine 23: 80–86.

11.

Cash

Novy

Grant

(1994) Why do women exercise? Factor analysis and further validation of the reasons for exercise inventory. Perceptual and Motor Skills 78: 539–544.

12.

Castro

King

(2002) Telephone-assisted counseling for physical activity. Exercise and Sport Sciences Reviews 30: 64–68.

13.

Centers for Disease Control and Prevention (CDC) (2009) Behavioral Risk Factor Surveillance System Survey Questionnaire. Atlanta, GA: US Department of Health and Human Services/Centers for Disease Control and Prevention.

14.

Cramp

Brawley

(2009) Sustaining self-regulatory efficacy and psychological outcome expectations for exercise: Effects of a group-mediated cognitive behavioural intervention. British Journal of Health Psychology 14: 595–611.

15.

Dunton

Atienza

(2009) The need for time-intensive information in healthy eating and physical activity research: A timely topic. Journal of the American Dietetic Association 109: 30–35.

16.

Dunton

Atienza

Castro

. (2009) Using ecological momentary assessment to examine antecedents and correlates of physical activity bouts in adults age 50+ years: A pilot study. Annals of Behavioral Medicine 38: 249–255.

17.

Gable

Poore

(2008) Which thoughts count? Algorithms for evaluating satisfaction in relationships. Psychological Science 19: 1030–1036.

18.

Gellert

Ziegelmann

Schwarzer

(2012) Affective and health-related outcome expectancies for physical activity in older adults. Psychology & Health 27: 816–828.

19.

Gwaltney

Shiffman

Sayette

(2005a) Situational correlates of abstinence self-efficacy. Journal of Abnormal Psychology 114: 649–660.

20.

Gwaltney

Shiffman

Balabanis

. (2005b) Dynamic self-efficacy and outcome expectancies: Prediction of smoking lapse and relapse. Journal of Abnormal Psychology 114: 661–675.

21.

Gyurcsik

Brawley

Spink

. (2011) Is level of pain acceptance differentially related to social cognitions and behavior? The case of active women with arthritis. Journal of Health Psychology 16(3): 530–539.

22.

Heck

Thomas

Tabata

(2010) Multilevel and Longitudinal Analysis using SPSS. New York: Routledge/Taylor & Francis.

23.

Hedeker

Gibbons

(2006) Longitudinal Data Analysis. New York: Wiley.

24.

Kendzierski

DeCarlo

(1991) Physical activity enjoyment scale: Two validation studies. Journal of Sport & Exercise Psychology 13: 50–64.

25.

Lally

Gardner

(2011) Promoting habit formation. Health Psychology Review 5: 1–22.

26.

Leventhal

Weinman

Leventhal

. (2008) Health psychology: The search for pathways between behavior and health. Annual Review of Psychology 59: 477–505.

27.

Lox

Jackson

Wasley

. (2000) Revisiting the measurement of exercise induced feeling states: The Physical Activity Affect Scale (PAAS). Measurement in Physical Education and Exercise Science 4: 79–95.

28.

Morrow

Bain

Frierson

. (2011) Long-term tracking of physical activity behaviors in women: The WIN Study. Medicine and Science in Sports and Exercise 43: 165–170.

29.

Nigg

Borrelli

Maddock

. (2008) A theory of physical activity maintenance. Applied Psychology: An International Review 57: 544–560.

30.

Papandonatos

Williams

Jennings

. (2011) Mediators of physical activity behavior change: Findings from a 12-month randomized controlled trial. Health Psychology 31: 512–520.

31.

Raudenbush

Bryk

(2002) Hierarchical Linear Models: Applications and Data Analysis Methods (2nd edn). Thousand Oaks, CA: SAGE.

32.

Resnick

Zimmerman

Orwig

. (2000) Outcome expectations for exercise scale: Utility and psychometrics. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences 55: S352–S356.

33.

Rothman

Baldwin

Hertel

. (2011) Self-regulation and behavior change: Disentangling behavioral initiation and behavioral maintenance. In: Vohs

Baumeister

(eds) Handbook of Self-Regulation: Research, Theory, and Applications (2nd edn). New York: Guilford Press, pp. 106–122.

34.

Rothman

Sheeran

Wood

(2009) Reflective and automatic processes in the initiation and maintenance of dietary change. Annals of Behavioral Medicine 38(Suppl. 1): 4–17.

35.

Schwarzer

(2008) Models of health behaviour change: Intention as mediator or stage as moderator? Psychology & Health 23: 259–263.

36.

Sears

Stanton

(2001) Expectancy-value constructs and expectancy violation as predictors of exercise adherence in previously sedentary women. Health Psychology 20: 325–333.

37.

Singer

Willett

(2003) Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York: Oxford University Press.

38.

US Department of Health and Human Services (USDHHS) (2008a) Be Active Your Way: A Guide for Adults. Washington, DC: USDHHS.

39.

US Department of Health and Human Services (USDHHS) (2008b) Physical Activity Guidelines Advisory Committee Report. Washington, DC: USDHHS.

40.

Van Zundert

Ferguson

Shiffman

. (2010) Dynamic effects of self-efficacy on smoking lapses and relapse among adolescents. Health Psychology 29: 246–254.

41.

Wilcox

Castro

King

(2006) Outcome expectations and physical activity participation in two samples of older women. Journal of Health Psychology 11: 65–77.

42.

Williams

(2010) Outcome expectancy and self-efficacy: Theoretical implications of an unresolved contradiction. Personality and Social Psychology Review 14: 417–425.

43.

Williams

Anderson

Winett

(2005) A review of the outcome expectancy construct in physical activity research. Annals of Behavioral Medicine 29: 70–79.