Prospective prediction of health-related behaviours with the Theory of Planned Behaviour: a meta-analysis

Using meta-analytic path analysis to test theoretical predictions in health behavior: An illustration based on meta-analyses of the theory of planned behavior

“…Specifically, we conducted a path analysis specifying theory predictions using the meta-analytically derived correlations behavioral control on intentions (β < .075) in alcohol behaviors is a cause for concern for theory validity (c.f., Seaton et al, 2010). The inclusion of past behavior increased the variance explained in the key dependent variables, intentions and behavior, and also attenuated many of the effects in the theory, most prominently the intention-behavior relationship, consistent with previous findings (Hagger et al, 2002;McEachan et al, 2012). The finding that past behavior reduced the effects of intentions on behavior indicates that intentions are only mildly effective in accounting for stability and change in health behaviors, consistent with Sniehotta et al's (2014) contention that the theory of planned behavior is a static rather than dynamic theory.…”

“…The syntheses provide more accurate estimations of the true effect among the constructs across the literature and evaluate the extent of variability of the effect after taking into account the corrected artefacts. There has been a proliferation in meta-analytic syntheses of the evidence for these theories including the theories of reasoned action and planned behavior (e.g., McEachan et al, 2012;Rich et al, 2015), self-determination theory (e.g., Chatzisarantis et al, 2003;Ng et al, 2012), social cognitive theory (e.g., Young et al, 2015), and the prototype willingness model (e.g., Todd et al, 2016;van Lettow et al, 2016) in the context of health behavior.…”

“…In fact, the inclusion of past behavior in the prediction of prospectivelymeasured behavior in social cognitive theories models the stability of the behavior over time. If the theory constructs fail to account for behavioral stability and to predict unique variance behavior over to whether it is considered effective as an account of health behavior, was reduced from a substantial to a relatively trivial effect size with the inclusion of past behavior (Hagger et al, 2002;McEachan et al, 2012). Failure to account for control variables like past behavior in path analytic meta-analyses of theories applied in health contexts places considerable limits on researchers' capacity to arrive at a definitive decision on the effectiveness of the model to explain behavioral outcomes.…”

The impact of changing attitudes, norms, and self-efficacy on health-related intentions and behavior: A meta-analysis.

“…The Impact of Changing Attitudes, Norms, and Self-Efficacy on Health-Related Intentions and Behavior: A Meta-Analysis Meta-analyses of correlational studies indicate that beliefs concerning (a) the appeal and consequences of behavior (attitudes), (b) other people's approval and performance of the behavior (social norms), and (c) one's ability to execute the relevant responses (self-efficacy) are reliable predictors of health behaviors (e.g., Conner & Sparks, 2005;Godin & Kok, 1996;McEachan et al, 2011). However, evidence from correlational studies that a variable predicts behavior does not necessarily indicate that interventions that change the same variable will cause changes in behavior (Sheeran, Harris, & Epton, 2014).…”

Predicting and explaining behavioral intention and hand sanitizer use among US Army soldiers

“…This is important so that positive numbers reflect a positive attitude to the target behavior. In a previous meta-analysis, the bipolar scoring of both belief items was superior to unipolar scoring (McEachan, Conner, Taylor, & Lawton, 2011). The interpretation of scores is easier when the midpoint of the scale is zero, and it is clear whether the final score represents an influence for (positive) or against (negative) enacting the behavior.…”