| Self-regulation may be defined as the process of inhibiting automatic responses that interfere with long-term goals. Previous research has shown that self-regulation, such as when restrained eaters resist impulses to consume tempting snacks, relies on a limited resource that becomes temporarily depleted with use (Baumeister, Bratslavsky, Muraven, & Tice, 1988). However, the underlying causal processes by which such impulses are resisted remain unclear.;Thought suppression was considered as a candidate self-regulatory mechanism. In light of the ironic process theory of mental control (Wegner, 1994) it was expected that resisting temptation would ironically be associated with increased accessibility of temptation-related thoughts, under concurrent cognitive load. Furthermore, intentionally focusing attention on the temptation would ironically reduce the accessibility of these thoughts. Finally, the accessibility of diet-goal related thoughts was considered. It was expected that despite prior research (Fishbach, Friedman, & Kruglanski, 2003), goal accessibility here would not be sufficient to decrease snack consumption.;Sixty one highly restrained eaters (50 female; Polivy, Herman, & Warsh, 1978) completed a lexical decision task (LDT) in the presence of appealing snacks (donut holes). Participants were either allowed to eat (self-regulation condition; SR), required to eat (indulgence condition; I) or allowed to eat and instructed to focus on the snacks (self-regulation with intentional focus; SRF). In a 3x3x2 mixed design, the LDT presented neutral, temptation and goal-related word stimuli, under varying levels of cognitive load, to each participant.;It was hypothesized that a three-way interaction would be identified, such that under high cognitive load, those in the self-regulation condition would demonstrate decreased reaction time to temptation-related words (i.e. increased accessibility of temptation thoughts). Furthermore, those in the self-regulation with intentional focus condition would demonstrate increased RT to these words. Decreased RT was predicted for goal-related words, in both SR and SRF conditions, regardless of cognitive load. Rates of snack consumption were predicted to be highest in I, followed by SR and lowest in SRF. Goal accessibility was predicted to have no relationship to snack consumption.;Examination of self-reported resistance to temptation revealed no significant between-group differences. Analysis of variance did not detect a significant main effect of condition or word type, or any significant interaction effects. Between-group differences in snack consumption were found, as predicted. SRF participants consumed fewer donut holes than SR (p = .06, d= -.64) and I participants (p < .01, d = -1.0), and SR participants consumed fewer (though not significantly) than I participants. Snack consumption was not related to goal accessibility. Post-hoc analyses explored participants' LDT responses based on subjective report of resistance to temptation, revealing the predicted three-way interaction (p= .03, etap2 = .12), but not in the expected group: participants reporting less resistance to temptation demonstrated increased accessibility of temptation words, under high cognitive load.;The ironic pattern associated with thought suppression did not emerge as predicted, very possibly because the between-groups manipulation of self-control did not succeed. The ironic pattern did emerge in examination of self-reported self-control. Surprisingly, it was present in those reporting low, and not high, resistance to temptation, and was associated with reduced snack consumption, lending support to growing literature regarding the relative importance of non-conscious rather than conscious self-regulatory processes. |
