Risk behaviour and descriptive normative beliefs among university students in 7 European countries - the longitudinal study SLiCE

Abstract

Background The risk behaviour of young people constitutes a widespread problem in Europe despite significant efforts in its prevention. Descriptive normative beliefs (DNB) have been successfully applied in the research and intervention of risky alcohol use. However, much less is known about DNB in other types of risk behaviour and longitudinal studies on this topic have been scarce. The aim of this study was to explore whether an association exists between the changes in DNB and different risk behaviours in a large sample of university students from 7 European countries within a longitudinal design. Methods The data were collected online from 1044 university students (74.2% females; Mage = 21.9; SD = 2.7) from 7 countries (23.8% Lithuania; 22.9% Slovakia; 19% Czech Republic; 18.3% Hungary; 7.8% Bulgaria; 4.4% Turkey; 3.9% Poland) at times T1 & T2 with a one year follow up. Students were asked regarding: alcohol use (AUDIT-C), smoking, marihuana use, sexual risk behaviour and answered items in DNB about typical students’ risk behaviour. The data were analysed by a series of linear regressions. The dependent variables were respective risk behaviours at T2 while independent variables were gender, country, risk behaviour at T1 and change of DNB (T2-T1). Results Linear models were statistically significant and explained 53.8%, 62.8%, 33.9% and 33.3% of the variance in alcohol use (AU), smoking (S), marihuana use (MU) and sexual risk behaviour (SRB), respectively. It was found that: (1) country [β = .055, p = .021], AU at T1 [β = .731, p ˂ .001] and the change of DNB [β = .114, p ˂ .001] contributed to the AU at T2; (2) country [β=.115, p ˂ .001], S at T1 [β = .768, p ˂ .001] and the change of DNB [β = .102, p = .020] contributed to the S at T2; (3) MU at T1 [β = .579, p ˂ .001] and the change of DNB [β = .097, p ˂ .001] contributed to the MU at T2; (4) country [β = .189, p ˂ .001], gender [β = .081, p [...].