Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: J Behav Med. 2020 Dec 23;44(2):270–276. doi: 10.1007/s10865-020-00166-x

Social Predictors of Daily Relations Between College Women’s Physical Activity Intentions and Behavior

Leah M Schumacher 1, Coco Thomas 2, M Cole Ainsworth 3, Danielle Arigo 4
PMCID: PMC7965240  NIHMSID: NIHMS1657395  PMID: 33355885

Abstract

Women perform less physical activity (PA) than men, and this gap widens during college. This study examined college women’s daily PA intentions and behavior, and whether social support or social comparison orientation (SCO) moderated the PA intention-behavior relation. College women (N=80) completed measures of social support and SCO at baseline. For seven consecutive days, participants completed an electronic survey to assess PA intentions and wore an activity monitor to assess minutes of moderate-to-vigorous intensity physical activity (MVPA). Results indicated that intended and performed MVPA minutes were weakly related (p=0.17, sr=0.16). Social support did not moderate the intention-behavior relation, but SCO did (p=0.04, sr=0.21). Participants with stronger (vs. weaker) SCO, particularly a tendency to compare downward (i.e., to worse-off others), showed smaller discrepancies between intended and completed MVPA. College women frequently fail to achieve PA goals, but stronger tendencies to make (downward) social comparisons may minimize this gap and be a target for intervention.

Keywords: social comparison, social support, gender, exercise, adherence

Many individuals intend to be physically active but have difficulty translating these intentions into physical activity (PA) behavior (Rhodes & Bruijn, 2013). This PA intention-behavior gap may present a particular problem for women, who achieve PA recommendations less often than men throughout the lifespan (Troiano et al., 2008) and who may experience unique difficulty during college (Irwin, 2004). Understanding more about the PA intention-behavior gap among college women could help to address the PA gender disparity by better informing PA promotion programs for this population.

Previous research has identified individual-level characteristics that moderate the PA intention-behavior gap. Such information is useful for identifying and targeting individuals who may be especially prone to difficulty implementing PA intentions. Previously identified moderators include PA intention stability, perceived behavioral control, and conscientiousness (Rhodes & Dickau, 2013). Perceptions of the social environment, including perceived social support and social comparisons, have received little attention as potential moderators of the PA intention-behavior relation, but may be particularly relevant for college women. Perceived social support has been consistently linked with positive health behaviors among college women and may assist with enacting their PA intentions (Hale et al., 2005). For example, women with stronger social support resources may be better able to obtain emotional support (e.g., encouragement), instrumental support (e.g., exercise partners), and informational support (e.g., advice) for PA, thus facilitating improved follow-through on their PA intentions. Greater global social support may also facilitate PA indirectly through greater beliefs about capabilities and greater general coping resources (Scarapicchia et al., 2017).

Social comparisons (i.e., self-evaluations relative to others; Festinger, 1954) are common among young women and are known to influence PA intentions and behavior in this population (Rancourt et al., 2015). In particular, a stronger tendency to value social comparison information (called social comparison orientation; SCO) has been associated with PA both cross-sectionally and prospectively (Arigo & Butryn, 2019; Luszczynska et al., 2004). SCO might also moderate the PA intention-behavior relation. Individuals who more strongly value comparison information in general might feel particularly motivated for PA when comparing with peers who either succeed with PA (i.e., upward comparison), as these peers may serve as role models, or when comparing with those who struggle with PA (i.e., downward comparison), who provide information about how to avoid a similar outcome. Individuals’ tendencies to make comparisons in a certain direction (upward vs. downward) may also relate to their PA and health behaviors. For example, past work indicates that a stronger (vs. weaker) tendency for upward comparisons predicted less weight loss in an obesity intervention, perhaps because upward comparisons were de-motivating or triggered negative affect that interfered with behavior change (Arigo & Butryn, 2019). As little research has examined the roles of perceived social support and SCO in PA, particularly among college women, further research into these potential influences on PA and the intention-behavior gap is warranted.

This observational study examined daily PA intentions and objectively assessed moderate-to-vigorous intensity PA (MVPA) across seven days among insufficiently active college women, with an emphasis on prospective relations between social support/SCO and the discrepancy between PA intentions and behavior. In this study, we sought to assess naturalistic relations among these variables; no PA intervention was provided. We also sought to examine baseline measures of social support and SCO as predictors of the PA intention-behavior relation, as such information can inform targeting of interventions to individuals with certain characteristics. Aim 1 was to determine the strength of relations between reported PA intentions and MVPA achieved the next day. Because we were primarily interested in understanding factors related to MVPA performance and the successful translation of intentions for PA into behavior, our analyses and interpretation included days on which PA intentions > 0 minutes were set. We expected to observe weak-to-moderate associations, indicating frequent failure to implement PA intentions when intentions were set. Aims 2 and 3 were to assess whether social support and SCO, respectively, moderated the PA intention-behavior relation. We hypothesized that greater social support and greater SCO would both predict a stronger PA intention-behavior relation.

Method

Participants

Participants were recruited for this seven-day, observational study using print and electronic advertisements. Eligible individuals were women enrolled at a small, private university in northeastern Pennsylvania (USA). Inclusion criteria required women to have no prior experience using wristworn or smartphone-based PA sensors, to engage in MVPA less than 100 minutes per week (per self-report), and to be in or beyond their second year of college. These first two criteria were set to recruit low-active women in order to learn more about factors impacting PA among individuals who might benefit most from structured PA intervention. First-year students were excluded due to variable PA patterns after relocating to college, which may not accurately reflect their current PA levels (Van Dyck et al., 2015). The resulting sample (N=80) was predominantly Caucasian; students were in their sophomore to senior years in college with a mean age of 20 (SD = 1.07), and mean BMI of 23.1 (SD = 3.87). For additional sociodemographic information and a study flowchart, see the main outcomes paper (Arigo, Pasko, & Mogle, 2020).

Procedure

All procedures were approved by the relevant Institutional Review Board. Potential participants completed an online survey to determine eligibility, which included the measures of social support and SCO described below. Eligible participants were invited to a face-to-face appointment to learn more about the study, provide written informed consent, receive instructions for completing daily diaries, and receive training in the use and care of the Fitbit. Participants also had their height and weight measured and were scheduled for a follow-up appointment to take place at the end of the data collection period. Participants completed the nightly electronic survey using a computer, tablet, or smartphone. Surveys completed between the time of receipt (10:00 pm) and 6:00 am the next morning were considered valid. Participants wore the Fitbit during waking hours, taking it off to charge each night, when showering, and for any other water immersion (e.g., swimming). Participants were not given any instructions to modify their PA. Fitbit data were not accessible to participants during data collection, and summary files were downloaded at follow-up appointments. At these appointments, participants returned the Fitbit and received compensation (i.e., either $10 to Starbucks or course credit).

Baseline Measures

Demographics.

Survey items assessed age, gender, and racial and ethnic identification.

Height and weight.

Height was measured by research staff using a standard stadiometer and weight was measured using a Fitbit Aria digital scale. These values were used to calculate body mass index (BMI).

Social support.

Perceived social support was assessed with the Social Support Appraisals Scale (Vaux et al., 1986). This measure uses a four-point response scale ranging from 1 (strongly disagree) to 4 (strongly agree). Items include “My friends respect me” and “I am well liked.” This scale has shown strong reliability and validity (Vaux et al., 1986).

Social comparison orientation.

The Iowa-Netherlands Comparison Orientation Measure (Gibbons & Buunk, 1999) assessed participants’ general tendency to compare themselves to others. The response scale ranges from 1 (strongly disagree) to 5 (strongly agree) for items such as “I often compare myself with others with respect to what I have accomplished in life.” Eleven items assess overall SCO, six items assess specific interest in upward comparison information, and six items assess interest in downward comparison information. This scale has shown good psychometric properties in validation research (Gibbons & Buunk, 1999).

Daily Measures

PA intentions.

As part of a larger daily diary study on predictors of change in PA among low-active college women (Arigo et al., 2020), participants received an electronic survey via email at 10:00 each night. They reported their PA intentions by recording minutes and type of PA they planned to perform the following day (“Do you have plans to do cardiovascular exercise tomorrow [going to the gym or for a run]?”). Response options were No, no plans to exercise and Yes. If they answered “yes,” they were asked for how many minutes they intended to engage in PA. These responses were used to determine participants’ PA intention for the following day. These question formats and response options, which are similar to those used in some prior studies (e.g., Bond et al., 2013), were chosen in an effort to make questions as straightforward as possible and thus minimize participant burden with the daily diary protocol.

Objectively-assessed PA.

PA behavior was assessed using a Fitbit Flex, worn on the wrist for seven days. Daily minutes of MVPA were determined using Fitbit records. Specifically, all minutes of MVPA per day that were accrued in bouts of ≥10 minutes were summed from Fitbit records to provide an estimate of daily aerobic-intensity PA. The Fitbit was selected due its affordability, attractiveness (which we believed would enhance wear), and ease of wear. While some research suggests that the Fitbit overestimates minutes of MVPA, particularly for higher intensity activities (Feehan et al., 2018), findings from several studies have found it to have at least moderate validity for determining time spent in MVPA (Adam Noah et al., 2013; Dominick et al., 2016). Valid wear days were days with ≥1000 steps (Bassett Jr et al., 2010).

Data Analysis

A total of 80 participants with seven days of data collection was intended to produce 560 individual observations. Participant retention was 100%. Overall Fitbit wear compliance was 94% and survey completion (excluding those completed after 6:00 am the subsequent day) was 96%. Participants identified PA intentions on six days for which MVPA data were available the following day. Two participants experienced Fitbit malfunctions resulting in no MVPA data and a total of 29 individual observations were missing PA data due to nonwear. Based on simulations by Hox (Hox et al., 2017; Maas & Hox, 2005), the remaining 427 observations afforded adequate power to test for the hypothesized cross-level (between*within-person) interactions.

These interactions and the hypothesized within-person, daily relations between PA intentions and MVPA were tested with two-level multilevel models (days nested within participants) using SAS PROC MIXED (Version 9.4). We tested the relation between intention and behavior in two ways: (1) using intention (yes/no) on a given day as a dichotomous predictor of minutes of actual MVPA the following day (to determine whether participants achieved more minutes of MVPA on days for which they had [vs. had not] set a PA intention), and (2) using intended minutes on a given day for which intentions were set (intention = yes) as a continuous predictor of minutes of actual MVPA the following day (to determine whether the amount of intended minutes predicted actual minutes; e.g., Bond et al., 2013). The interaction effects of baseline social support and SCO were tested as moderators of intention as a continuous predictor (#2 above). Of note, we also tested these models using all days, including those for which intentions were not set (i.e., intended minutes = 0), which did not differ from the results or conclusions presented below. All models employed restricted maximum likelihood estimation to account for missing data and controlled for participant BMI and typical PA intentions. Effect sizes are expressed as semipartial correlation coefficients (sr).

Results

Of 560 individual days of data collection, participants set intentions to engage in MVPA on the following day on 194 days (36%). The average participant set PA intentions for the following day on 2.43 of seven days (SD = 1.97). On days with intentions set for the following day, average intended MVPA minutes was 40.57 (SE = 2.06). Less than 50% of the variability in intended MVPA was attributable to stable, between-person differences (ICC = 0.48), leaving significant within-person variability (and error; p < 0.0001). Across participants, average observed MVPA over seven days was 26.43 minutes per day (SE = 1.90). As 26% of the variability in observed MVPA was attributable to stable, between-person differences (ICC = 0.26), the majority was attributable to within-person change and error (residual p < 0.0001).

When examining intention to perform PA (yes/no) as a predictor of MVPA (controlling for participant BMI and typical intentions), participants engaged in 12.70 additional minutes of MVPA on days for which they did (vs. did not) set intentions (F = 16.83, p < 0.0001, sr = 0.46). However, when examining number of intended PA minutes as a predictor of MVPA among days with set intention (controlling for BMI and typical intentions), the relation between intended and actual minutes of MVPA on the following day was not significant (F = 1.93, p = 0.17, sr = 0.16).

Among days with set intentions, perceived social support did not moderate the relation between intended and actual MVPA minutes (ps > 0.70; see Table 1), though global SCO did (F = 3.43, p = 0.04, sr = 0.21). Specifically, participants with greater SCO showed stronger intention-behavior relations over seven days. Upward SCO was not associated with this relation (F = 3.25, p = 0.81, sr = 0.03), while downward SCO was (F = 3.25, p = 0.05, sr = 0.20). Participants who endorsed greater downward SCO showed less of a PA intention-behavior gap.

Table 1.

Descriptive statistics for key variables and multilevel models testing relations between PA intentions and MVPA minutes on the following day.

Descriptive Statistics B (SE)
Intended MVPA Minutes 40.57 (2.06)
Actual MVPA Minutes 26.43 (1.90)
Multilevel Model Outcome: Actual MVPA Minutes
B (SE)
 Intercept (BP) 36.49 (6.30)***
 BMI (BP) 0.77 (0.60)
 Intention (BP) −0.06 (0.14)
 Intention (WP) – Set vs. Not Seta 12.70 (3.10)***
 Intention (WP) – Number of Minutesb 0.31 (0.22)
Proposed Moderators
 Social Support (BP) x Intention (WP) b −0.004 (0.02)
 Social Comparison Orientation (General, BP) x Intentionb (WP) 0.05 (0.03)*
  Upward Orientation (BP) x Intention (WP)b −0.009 (0.04)
  Downward Orientation (BP) x Intention (WP)b 0.05 (0.03)*
Open in a new tab

Note: PA = physical activity; MVPA = moderate-to-vigorous physical activity; BP = between-person (stable), WP = within-person (varying by day)

*

p < 0.05

**

p < 0.001.

a

Model included all PA intention surveys

b

Model included only PA intention surveys were participants endorsed an intention to perform MVPA the next day

Discussion

This study provides new insight into discrepancies between college women’s PA intentions and MVPA, with a particular focus on how perceptions of the social environment predict the intention-behavior relation. Participants reported PA intentions on approximately one-third (35%) of the total days recorded. This relatively low frequency of PA intention setting is unsurprising given that eligibility required participants to be insufficiently active. Although participants performed approximately 13 more minutes of PA on days for which PA intentions were (vs. were not) set, as expected, we observed weak relations between the number of intended and achieved PA minutes across days when participants set PA intentions. These findings suggest that participants had considerable difficulty translating set PA intentions into behavior and align with prior work revealing discordance between PA intentions and behavior, including among college women (Rhodes & Bruijn, 2013). This pattern of findings is particularly notable given some past research suggesting Fitbits overestimate MVPA (Feehan et al., 2018); the true relation between PA intentions and behavior may thus be weaker than that detected in this study.

As hypothesized, participants with a stronger SCO showed stronger intention-behavior relations. These relations were not the same across both directions of SCO, however. While participants’ downward comparison orientation (i.e., comparisons to “worse off” others) was positively related to the strength of the PA intention-behavior relation, participants’ upward comparison orientation was not related to the PA intention-behavior association. These findings are consistent with a recent study showing that stronger SCO (in either direction) was associated with better maintenance of MVPA among adults participating in behavioral weight loss treatment (Arigo & Butryn, 2019), and extend this relation to a broader context. Comparison to “worse off” individuals may increase PA self-efficacy and facilitate PA behavior by demonstrating progress toward one’s goals (Carmona et al., 2008). These results are consistent with previous work indicating that anticipated regret moderates the PA intention-behavior relation (Rhodes & Dickau, 2013); downward comparisons may also facilitate PA by prompting individuals to want to avoid a similar undesirable future state (Wills, 1981). Individuals with stronger tendencies to make downward comparisons may thus be better positioned to notice and capitalize upon these comparison opportunities in ways that bolster self-efficacy, motivation, and PA. Like downward comparisons, upward comparisons have potential to increase PA self-efficacy. However, upward comparisons also remind oneself of the gap between the current and ideal self and may thus be less motivating than downward comparisons (Pila et al., 2016).

In contrast, perceptions of social support were not associated with the intention-behavior relation. This finding contrasts with some prior work that has revealed an association between social support and aerobic PA among young adults (Pengpid et al., 2015), but is consistent with other work showing that social comparison outperforms social support as a predictor of health outcomes among college women (Arigo & Cavanaugh, 2016; Zhang et al., 2016), and with results from a meta-analysis that found social support was not a strong predictor of PA among slightly younger populations (i.e., adolescent girls; Laird et al., 2016). It is possible that global social support has limited relevance to the PA intention-behavior relation, particularly when individuals may not actively try to increase PA (as in this observational study). One’s perceived level of social support may be more important when individuals are actively seeking to modify behavior (e.g., by providing resources for instrumental or emotional support). The general (vs. PA-specific) social support measure we used may have also influenced our results; while broad social support may be less strongly associated with PA, one’s perceived social support specific to PA might have greater implications for PA intention-behavior relations (e.g., Leslie et al., 1999). Thus, further research on the impact of social support on college women’s PA is needed.

Strengths of the current study include assessment of SCO and social support using validated measures; examination of PA intention-behavior relations specifically among inactive college women, who represent an important target population for PA interventions; and differentiation of between- and within-person variability in both PA intentions and behavior. We also assessed PA intentions via daily diaries and assessed MVPA with wristworn sensors, thereby reducing biases associated with retrospective self-reports over long periods. Limitations include a short assessment period, use of general (vs. PA-specific) social support and SCO measures, use of a PA intention measure that differs from some prior work (e.g., Conroy et al., 2013), and potential limited generalizability of our findings to other samples (e.g., due to specifically recruiting low-active college women).

As discussed above, it is also possible that the Fitbit overestimated MVPA minutes (Feehan et al., 2018); the true concordance between PA intentions and behavior may thus be weaker than reported here. As with other studies that monitor behavior or involve frequent self-report assessments, it is also possible that participants’ reported PA intentions or behavior was influenced by demand characteristics. However, there is little evidence that demand influenced intention reports, given that participants set intentions on less than 50% of days. Excluding days without set intentions reduced power for some models. However, we did not observe meaningful changes in findings when conducting sensitivity analyses including days without set intentions.

There are several implications of this work. First, college women’s social comparison tendencies could be an important target for understanding and strengthening PA intention-behavior relations, thus helping to address PA gender disparities. Our findings also add to the literature indicating that social comparison serves as a better predictor of PA engagement than social support (Arigo & Cavanaugh, 2016; Zhang et al., 2016). Future research should focus on clarifying the processes underlying the benefits of comparisons for PA promotion (e.g., if downward comparisons bolster self-efficacy), and assess the influence of comparisons on PA intentions and behavior in real time. Experimental examination of how social comparisons might be leveraged to increase PA (e.g., via use of leaderboards in PA programs) is also warranted. Future studies should seek to clarify in real-time the reasons why individuals do not follow through on their PA intentions (e.g., low motivation, anticipated boredom), whether social support or social comparisons are differentially influential in overcoming these barriers, and evaluate variability in the frequency of PA intention attainment across participants.

In conclusion, this study adds to the body of research documenting frequent PA intention-behavior discordance among college women and expands on this literature by suggesting that SCO—particularly downward orientation—moderates the intention-behavior relation. Further research is warranted to assess the impact of social comparisons on PA among insufficiently active college women, including studies that assess potential intervention techniques for capitalizing on these social comparison processes to promote PA. Additional research on the role of social support in PA among low active college women is also needed.

Acknowledgements

The authors would like to thank Kristen Pasko, Sabrina DiBisceglie, Zuhri Outland, Marissa DeStefano, and Nicole Plantier for their assistance with data collection.

Declarations

Funding: Data collection for this study was funded by internal resources at the last author’s institution. Preparation of this manuscript was supported by T32 HL076134 (Schumacher).

Footnotes

Conflicts of Interest: Drs. Schumacher and Arigo receive funding from the National Heart, Lung, and Blood Institute (Schumacher: T32 076134; Arigo: K23 136657). The authors declare that they have no other conflict of interest.

Ethics Approval: Approval was obtained from the Institutional Review Board of The University of Scranton and all study procedures adhered to the tenets of the Declaration of Helsinki.

Consent to Participate: Informed consent was obtained from all individual participants included in the study.

Contributor Information

Leah M Schumacher, Weight Control and Diabetes Research Center, The Miriam Hospital/Brown Alpert Medical School, Providence, RI

Coco Thomas, Philadelphia College of Osteopathic Medicine, Philadelphia, PA

M Cole Ainsworth, Department of Psychology, Rowan University, Glassboro, NJ

Danielle Arigo, Department of Psychology, Rowan University, Glassboro, NJ

References

  1. Adam Noah J, Spierer DK, Gu J, & Bronner S (2013). Comparison of steps and energy expenditure assessment in adults of Fitbit Tracker and Ultra to the Actical and indirect calorimetry. Journal of Medical Engineering and Technology, 37(7), 456–462. [DOI] [PubMed] [Google Scholar]
  2. Arigo D, & Butryn ML (2019). Prospective Relations between Social Comparison Orientation and Weight Loss Outcomes. Behavioral Medicine, 45(3), 249–254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arigo D, & Cavanaugh J (2016). Social perceptions predict change in women’s mental health and health-related quality of life during the first semester of college. Journal of Social and Clinical Psychology, 35(8), 643–663. [Google Scholar]
  4. Arigo D, Pasko K, & Mogle JA (2020). Daily relations between social perceptions and physical activity among college women. Psychology of Sport and Exercise, 47, 101528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bassett DR Jr, Wyatt HR, Thompson H, Peters JC, & Hill JO (2010). Pedometer-measured physical activity and health behaviors in United States adults. Medicine and Science in Sports and Exercise, 42(10), 1819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bond DS, Thomas JG, Ryder BA, Vithiananthan S, Pohl D, & Wing RR (2013). Ecological momentary assessment of the relationship between intention and physical activity behavior in bariatric surgery patients. International Journal of Behavioral Medicine, 20(1), 82–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Carmona C, Buunk AP, Dijkstra A, & Peiro JM (2008). The relationship between goal orientation, social comparison responses, self-efficacy, and performance. European Psychologist, 13(3), 188–196. [Google Scholar]
  8. Conroy DE, Elavsky S, Doerksen SE, & Maher JP (2013). A daily process analysis of intentions and physical activity in college students. Journal of Sport and Exercise Psychology, 35(5), 493–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dominick GM, Winfree KN, Pohlig RT, & Papas MA (2016). Physical activity assessment between consumer-and research-grade accelerometers: a comparative study in free-living conditions. JMIR mHealth and uHealth, 4(3), e110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Feehan LM, Geldman J, Sayre EC, Park C, Ezzat AM, Yoo JY, … & Li LC (2018). Accuracy of Fitbit devices: systematic review and narrative syntheses of quantitative data. JMIR mHealth and uHealth, 6(8), e10527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Festinger L (1954). A theory of social comparison processes. Human Relations, 7(2), 117–140. [Google Scholar]
  12. Gibbons FX, & Buunk BP (1999). Individual differences in social comparison: development of a scale of social comparison orientation. Journal of Personality and Social Psychology, 76(1), 129. [DOI] [PubMed] [Google Scholar]
  13. Hale CJ, Hannum JW, & Espelage DL (2005). Social support and physical health: The importance of belonging. Journal of American College Health, 53(6), 276–284. [DOI] [PubMed] [Google Scholar]
  14. Hox JJ, Moerbeek M, & Van de Schoot R (2010). Multilevel analysis: Techniques and applications. Routledge. [Google Scholar]
  15. Irwin JD (2004). Prevalence of university students’ sufficient physical activity: a systematic review. Perceptual and Motor Skills, 98(3), 927–943. [DOI] [PubMed] [Google Scholar]
  16. Laird Y, Fawkner S, Kelly P, McNamee L, & Niven A (2016). The role of social support on physical activity behaviour in adolescent girls: a systematic review and meta-analysis. International Journal of Behavioral Nutrition and Physical Activity, 13(1), 79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Leslie E, Owen N, Salmon J, Bauman A, Sallis JF, & Lo SK (1999). Insufficiently active Australian college students: perceived personal, social, and environmental influences. Preventive medicine, 28(1), 20–27. [DOI] [PubMed] [Google Scholar]
  18. Luszczynska A, Gibbons FX, Piko BF, & Tekozel M (2004). Self-regulatory cognitions, social comparison, and perceived peers’ behaviors as predictors of nutrition and physical activity: A comparison among adolescents in Hungary, Poland, Turkey, and USA. Psychology & Health, 19(5), 577–593. [Google Scholar]
  19. Maas CJ, & Hox JJ (2005). Sufficient sample sizes for multilevel modeling. Methodology, 1(3), 86–92. [Google Scholar]
  20. Pengpid S, Peltzer K, Kassean HK, Tsala JPT, Sychareun V, & Müller-Riemenschneider F (2015). Physical inactivity and associated factors among university students in 23 low-, middle-and high-income countries. International Journal of Public Health, 60(5), 539–549. [DOI] [PubMed] [Google Scholar]
  21. Pila E, Barlow MA, Wrosch C, & Sabiston CM (2016). Comparing the body to superior others: Associations with daily exercise and body evaluation in men and women. Psychology of Sport and Exercise, 27, 120–127. [Google Scholar]
  22. Rancourt D, Leahey TM, LaRose JG, & Crowther JH (2015). Effects of weight-focused social comparisons on diet and activity outcomes in overweight and obese young women. Obesity, 23(1), 85–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Rhodes RE, & Bruijn GJ (2013). How big is the physical activity intention–behaviour gap? A meta-analysis using the action control framework. British Journal of Health Psychology, 18(2), 296–309. [DOI] [PubMed] [Google Scholar]
  24. Rhodes RE, & Dickau L (2013). Moderators of the intention-behaviour relationship in the physical activity domain: a systematic review. British Journal of Sports Medicine, 47(4), 215–225. [DOI] [PubMed] [Google Scholar]
  25. Scarapicchia TMF, Amireault S, Faulkner G, & Sabiston CM (2017). Social support and physical activity participation among healthy adults: a systematic review of prospective studies. International Review of Sport and Exercise Psychology, 10(1), 50–83. [Google Scholar]
  26. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, & McDowell M (2008). Physical activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise, 40(1), 181–188. [DOI] [PubMed] [Google Scholar]
  27. Van Dyck D, De Bourdeaudhuij I, Deliens T, & Deforche B (2015). Can changes in psychosocial factors and residency explain the decrease in physical activity during the transition from high school to college or university? International Journal of Behavioral Medicine, 22(2), 178–186. [DOI] [PubMed] [Google Scholar]
  28. Vaux A, Phillips J, Holly L, Thomson B, Williams D, & Stewart D (1986). The social support appraisals (SS-A) scale: Studies of reliability and validity. American Journal of Community Psychology, 14(2), 195–218. [Google Scholar]
  29. Wills TA (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90(2), 245. [Google Scholar]
  30. Zhang J, Brackbill D, Yang S, Becker J, Herbert N, & Centola D (2016). Support or competition? How online social networks increase physical activity: A randomized controlled trial. Preventive Medicine Reports, 4, 453–458. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES