Improving Recall of Past Political Behaviour: Results from a question-wording Experiment in Austria, Germany and the UK

Research on political behaviour relies to a large extent on respondents’ self-reports on their attitudes and behaviours collected through standardized surveys. For example, studies on electoral behaviour generally use self-reports to capture whether respondents voted and whom they voted for. Yet respondents are often, intentionally or unintentionally, inconsistent in reporting their past (electoral) behaviour, creating an importing challenge to the validity of self-reports in surveys. Scholars have identified (1) unintentional consistency errors based on memory and/or attention failure as well as (2) intentional misreporting based on social desirability. This projects focuses on reducing unintentional inconsistency in reporting past electoral behaviour. Using a survey experiment in Austria, Germany and UK, we attempt to reduce unintentional misreporting by examining the influence of two cognitive processes. To correct for memory failure we modify the standard turnout and vote choice recall question to provide respondents with a recall cue designed to help them to search their memory. To address attention failure we alter the standard recall question by either encouraging respondents to use more time to recall information or by stressing the importance of the question. In sum, we test three potential quick ‘fixes’ to the way we usually ask people to recall their past turnout and vote choice in standardized surveys that can potentially serve scholars and practitioners in future.