Scientific understanding of physical activity has been constrained by a temporal and spatial leash produced by research designs that fail to capture the ebb and flow of behavior as it unfolds in daily life. In this issue of Exercise and Sport Sciences Reviews, Dunton (3) makes a timely and persuasive argument for using ecological momentary assessment (EMA) methods to loosen this leash and test new hypotheses about the synchronicity, sequentiality, and instability of physical activity.
Synchronicity hypotheses require us to consider the time scale of physical activity and its correlates. For example, physical activity intentions at a single occasion have frequently been linked with physical activity. EMA studies have shown that these intentions fluctuate over daily, weekly, and monthly time scales, perhaps in response to changes in time demands, health, and the seasons (1,2,5). When these dynamics have been modeled, associations between intentions and physical activity consistently emerge as a within-person, and not a between-person, process. People are more active after a momentary report of stronger intentions than usual, but people with stronger intentions (on average) do not necessarily engage in more physical activity.
Locating this association as a within-person phenomenon alters how we conceptualize the intention-behavior gap. A substantial group of unsuccessful intenders or inclined abstainers (~36% of the population) report intending to engage in physical activity but fail to transform that intention into action (4). These EMA-based findings suggest that inclined abstention is a momentary characteristic of a person rather than a characteristic of the person generally. Strategies for supporting self-regulation and overcoming time-varying vulnerabilities to this phenomenon would be valuable.
Sequentiality hypotheses address a second requirement for causal inferences: temporal precedence of causal variables. Physical activity may be a consequence (e.g., of motivation) or an antecedent (e.g., of health outcomes). In the studies described previously, physical activity intentions were assessed before and with respect to future physical activity. It is possible that these intentions are influenced by past physical activity as well, indicating a desire to recover after an activity. EMA methods open the door to testing this possibility and to ruling out other plausible sources of covariation (e.g., confounding by the social calendar can be addressed by controlling for the day of week).
Taken together, evidence for synchronicity and sequentiality provided by EMA methods strengthen causal inferences relative to what is possible with cross-sectional or prospective methods. Of course, this evidence is not sufficient. The missing requirement for causal inference, constant conjunction, requires that one rule out all other explanations for the association. Experimental designs are indispensable, but randomized controlled trials are expensive and time consuming. The synchronicity and sequentiality of intervention targets and outcomes should be established before investing resources in a trial.
Instability was the third hypothesis noted by Dunton (3), and this hypothesis has received the least attention in the literature. Yet, changes in context-contingent instability are expected during habit formation, so this hypothesis is critical for understanding adherence and the maintenance of behavior change. For example, habit measurement has proven to be controversial because introspection may not be sufficient to identify the cues that trigger behaviors automatically. EMA methods, combined with advances in sensor technology, create new opportunities for capturing and modeling cue-contingent behavioral patterns (i.e., habits).
EMA methods have opened a new frontier for physical activity research. Thoughtful applications of these methods can improve behavioral interventions and enrich our understanding of elusive processes that underlie the maintenance of behavior change.
David E. Conroy
Department of Kinesiology
The Pennsylvania State University
University Park, PA
and Department of Preventive Medicine
Northwestern University, Chicago, IL
1. Conroy DE, Elavsky S, Hyde AL, Doerksen SE. The dynamic nature of physical activity intentions: a within-person perspective on intention-behavior coupling. J. Sport Exerc. Psychol
. 2011; 33(6):807–27.
2. Conroy DE, Elavsky S, Maher JP, Doerksen SE. A daily process analysis of intentions and physical activity in college students. J. Sport Exerc. Psychol
. 2013; 35(5):493–502.
3. Dunton GF. Ecological momentary assessment methods in physical activity research. Exerc. Sport Sci. Rev
. 2017; 45(1):48–54.
4. Rhodes RE, de Bruijn GJ. How big is the physical activity intention-behaviour gap? A meta-analysis using the action control framework. Br. J. Health Psychol
. 2013; 18(2):296–309.
5. Scholz U, Nagy G, Schüz B, Ziegelmann JP. The role of motivational and volitional factors for self-regulated running training: associations on the between- and within-person level. Br. J. Soc. Psychol
. 2008; 47(Pt 3):421–39.