| Popis: |
The question-behaviour effect, how asking attitude, intention and prediction questions influences behaviour, has been widely examined since its discovery by Sherman (1980). This PhD thesis on the question-behaviour effect consists of three parts: 1) A meta-analysis of the research thus far, 2) A series of studies investigating possible underlying mechanisms of the question-behaviour effect, and 3) A series of studies investigating the moderating roles of self-affirmation and goal difficulty in question-behaviour effect interventions. The meta-analysis in the first part of this thesis is carried out to examine the effectiveness of the question-behaviour effect as influence technique. Studies were included if they used an experimental design with random allocation of participants, where the experimental condition consisted of asking attitude, intention and/or prediction questions, and the dependent variable was a behavioural measurement. This resulted in 55 comparisons in 35 papers, with a total of 49108 participants. Applying a random-effects model on the data resulted in a small effect (d = 0.26, 95%CI [0.18, 0.34]). Methodological causes and moderators related to the applicability and universality of the technique are discussed. The second part of this thesis consists of three experimental studies that investigated whether dissonance processes can explain question-behaviour effects. The studies investigated how predicting future behaviour can influence participants’ recalled past behaviour regarding both positive and negative behaviours. Study 1 investigated how asking participants about positive (daily exercise) or negative (not brushing your teeth before going to bed) behaviours affected recalled past behaviour. Study 2 focused on how recalling past behaviour in a precise (number of times) or vague manner (never – all the time) influences the effect of prediction questions on recalling a negative behaviour (procrastination). Study 3 investigated how combining a future behaviour prediction with a positive or negative prime could influence recalled past behaviour regarding buying bottled water. The results of Study 1 showed that asking participants about a negative behaviour (going to bed without brushing your teeth) reduced recalled past behaviour. Study 2 showed that offering specific rather than vague answering possibilities is required to find an effect and the results of Study 3 showed that a positive or negative prime can influence recalled past behaviour in that direction, while adding a future behaviour prediction question attenuates the effect. Theoretical explanations of the findings and ideas for further research are discussed and a fourth, correlational, study is described that suggests social proof might explain some of the question-behaviour effect findings. In the third part of this thesis, the moderating role of self-affirmation and goal difficulty in question-behaviour effect interventions was investigated. In studies 5-7, participants were asked a prediction question related to eating five-a-day (hard-goal) or eating fruit and vegetables in general (easy-goal). Some participants received a self- affirmation task as part of their questionnaire and all participants received a voucher they could use to collect a free bowl of fruit or vegetables after completing the questionnaire. Results showed that easy-goal prediction questions (eating fruit and vegetables) resulted in an increase in voucher use while hard-goal prediction questions (eating five-a- day) resulted in a decrease in voucher use. Adding self-affirmation to the intervention attenuated these question-behaviour effects. Implications of findings and general themes of the thesis are discussed in the general discussion chapter. |
