People in the UK are renowned for their high levels of binge drinking; indeed the UK (and Ireland) has the highest binge drinking rates in Europe. Why is this? One possible reason is the “drinking motives” that people have.
Multivariate ANOVA (MANOVA)
People in the UK are renowned for their high levels of binge drinking; indeed the UK (and Ireland) has the highest binge drinking rates in Europe. Why is this? One possible reason is the “drinking motives” that people have. Drinking motives represent motivational constructs that drive us to drink alcohol, for example some people drink as it facilitates social interaction, this would be a social motives. There are believed to be four major drinking motives.
1. Coping- drinking to cope with negative moods e.g. coping with anxiety and depression (internal, negative reinforcement)
2. Conformity- drinking so you do not feel left out of the group (external negative reinforcement)
3. Enhancement- because you enjoy the subjective effects/”buzz” that alcohol gives you (internal positive reinforcement)
4. Social- because it makes socialising more fun (external positive reinforcement)
A postgraduate student wishes to identify if there are any cross-cultural differences in drinking motives so creates an online questionnaire measuring drinking motives.
Participants also give information on where they were born, the researcher categorises people as
1. UK
2. Other EU
3. Rest of the world (other)
The student wishes to test if the IV- place of birth with 3 levels (UK/EU/other) has an effect on any of the four drinking motives (Coping/Conformity /Enhancement/Social). When investigating any significant effects the LSD test was used
The hypotheses are that (1) that UK drinkers will have stronger enhancement and social motives than the other groups. (2) Rest of the world drinkers will show higher conformity motives than the other groups. (3) EU drinkers will show higher coping motives than other groups
Write up:
3a. What does Box’s test measure and is this a problem for this data? (2 marks)
3b. Write up the multivariate test statistic in continuous prose using APA formatting. (2Marks)
3c. What is does Levene’s test measure and should we be concerned about any of the Levene’s tests? (2Marks)
3d. Write up the ANOVAs for the individual DVs in continuous prose using APA formatting, and report the post hoc tests where appropriate. (14 marks)
Other possible Q’s
3e. What could you do if your fail to find an acceptable Levene’s test? (2marks)
3f Do you support the hypotheses? Explain your answers? (6 marks)
A wide range of factors is known to influence adolescent alcohol use behaviours as reviewed by Hawkins, Catalano, and Miller (1992). Although somewhat outdated as a review, the categorization of the risk factors is still useful.
The authors distinguish between on the one hand individual and interpersonal factors and on the other hand contextual factors.
Like almost all human behaviour, alcohol use – though fundamentally the behaviour of individuals – occurs in the context of societies with their legal and cultural norms and other factors restricting or enabling behaviour. We start with individual risk factors (genetics, cognitive processes, personality), followed by interpersonal risk factors (e.g., influences of family and peers), and end with some environmental factors (e.g., neighbourhood influences).
Factors at different levels of description can influence each other; for example, a stressful family environment has an impact on the development of the stress-reactivity of an individual and this is moderated by genetic factors, which can lead to “developmental cascades” (Masten & Cicchetti, 2010).
We acknowledge that this is not a full systematic review in all of these wide domains, which was beyond the scope of this project. We illustrate current thinking about risk and protective factors across different levels of description and their interplay.
We also note that many risk and protective factors have not been studied across many different cultures, and in research on some risk factors there is a gender bias (e.g., acute alcohol effects, which has been primarily studied in men, mostly for pragmatic reasons such as risk of pregnancy, lower tolerance in women, etc.).
2• Risk and protective factors can be described at many different levels of description, ranging from genes and (sub-) cellular processes to societal characteristics. It is important to distinguish risk- and protective factors with respect to their causal status and with respect to their malleability. We focused on risk factors with some evidence of causality and/or which may be malleable through interventions.
3• Peers and parents are important factors across cultures in influencing underage drinking and related problems, with many indirect pathways (e.g., parental SES, neighbourhood-effects, etc.). Recent evidence has demonstrated a causal role for alcohol-specific rule setting on delaying age of onset both in North Europe and in the U.S.
4• Personality is an important risk factor, with different facets of personality influencing risk for underage drinking and related problems in different ways.
Externalizing characteristics and related traits (impulsivity, sensation seeking, weak self-control) have been related to early onset and escalation. Internalizing problems appear to be less strongly related to age of onset, but do appear to play an important role in the escalation of teen drinking, with an important role for motivations to drink in order to cope with problems.
5• Alcohol-related cognitive processes are important predictors of alcohol use and problems, with some operating more on explicit reasoning processes (e.g., motives to drink) and others operating more automatically (e.g., attention captured by alcohol-related stimuli).
6• Some risk factors (parenting, personality, cognitive processes) can be targeted in interventions, as are discussed in Chapter 3.
7• Two important issues regarding risk and protective factors are their causal status, and whether or not they are amenable to intervention. Regarding causal status, there are two ways to investigate this: first, longitudinal studies can be done where the risk factor is assessed and mediation of the risk factor regarding the outcome is tested (Baron & Kenny, 1986; MacKinnon & Fairchild, 2009; MacKinnon & Lockwood, 2003; Maric, Wiers, & Prins, 2012)
. The most convincing way to demonstrate causality, however, is to conduct an experiment and to manipulate the variable of interest in an experiment with random assignment and to demonstrate the effect on the outcome. In this context, that strategy is often a form of an intervention, although most prevention programmes are broad and rarely target a specific variable.
As an example from underage drinking, it has been demonstrated that strict rule setting regarding alcohol use by parents is longitudinally related to a delayed age of onset of drinking in children (van der Vorst, Engels, Dekovic, Meeus, & Vermulst, 2007). From this observation it seems likely that lack of strict alcohol rules (no drinking allowed before the legal drinking age) is a causal factor in the prediction of underage drinking.
However, there are other potential explanations: so called third-variables that could explain this relationship. For example, parents scoring high on intelligence or low on impulsivity could set more strict alcohol-specific rules than parents scoring low on intelligence or high on impulsivity. Given that these characteristics are influenced by heritability, offspring of highly intelligent (low impulsive) parents will score relatively high on intelligence (low on impulsivity), which could predict late age of onset.
In that case the correlation between parental rule-setting and late age of onset would be a spurious one. The only way to rule out the possibility of third variables is to conduct an experiment, and that has been done for this example (Koning et al., 2009).
Adolescents were randomly assigned to conditions combining universal prevention aimed at the adolescents themselves, with prevention aimed at the parents or to universal prevention alone.
Only adolescents receiving the universal prevention, and whose parents had also received the parent intervention, delayed their onset of drinking.
In this case, third variables can be ruled out, and the interpretation of this effect was strengthened by the fact that changes in parental attitudes and perceived rules mediated the intervention effect (Koning, van den Eijnden, Engels, Verdurmen, & Vollebergh, 2011).