Objective: To investigate the role of time perspective and desire for control in selfreported substance use and to test for a moderating effect of desire for control in the relation between time perspective and substance use. Procedure: A random sample of 240 persons, aged 15 years and over, selected in various public spaces in an urban region in central France. Main outcome measures: Time perspective was measured using subscales of the Zimbardo Time Perspective Inventory (ZTPI, Zimbardo & Boyd, 1999), Desire for control was measured using a translated version of the Desire for Control Scale (DCS, Burger & Cooper, 1979), and substance use was selfreported. Results: After controlling for age and gender, significant links were found between time perspective and substance use. Desire for control did not appear to be directly related to substance use. The interaction effect between TP and desire for control appeared to be related to substance use. There was evidence that the relation between TP and substance use is buffered by low desire for control. Conclusion: This study converges with previous studies demonstrating the relation between TP and substance use, but provides evidence of the moderating role played by desire for control. Desire for control thus appears as worthy of interest in the analysis of selfregulatory process, and further research on the links between TP and various aspects of control is required. In order to be more effective, the design of future studies and interventions based on time-related issues should consider how desire for control plays a part in establishing vulnerability profiles.

Two experiments examined the relationships between the knowledge that another person has won in a gamble, the illusion of control and risk taking. Participants played a computer-simulated French roulette game individually. Before playing, some participants learnt that another person won a large amount of money. Results from a first experiment (n = 24) validated a causal model where the knowledge of another person's win increased the illusion of control, measured with betting times, expectancy and self-reports on scales, which in turn encourages risk taking. In the second experiment (n = 36), some participants were told the previous player acknowledged the win to be fortuitous. The suppression of the belief that the previous winner had himself exerted control over the outcome resulted in lower rates of risk-taking behaviors. This suggests that it was not the knowledge of another person's win in itself that increased risk taking, but rather, the belief that the other person had some control over the gamble's outcome. Theoretical implications for the study of social mechanisms involved in gambling behavior are discussed.

This short paper aims to untangle the effect of loss-framing versus gain and non-gain; explaining when, how and why it influences individuals' intentions to engage in cholesterol screening. We argue that framing-effects are (1) significant only when individuals perceive the issue to be highly relevant and (2) are mediated by perceived negative consequences (resulting from undergoing the test) and response-efficacy. In a 2(issue-relevance: high vs low) × 3(framing: gain vs non-gain vs loss) experiment, 229 participants received a message and answered to a questionnaire measuring personal relevance, perceived negative consequences, response-efficacy, intention. Results validated a mediation model, explaining that lossframing is more persuasive than non-gain, which is more persuasive than gain-framing, partly because of their effect on individuals' perceptions of response-efficacy.