“…These controls included age (in cubic form); gender; ethnicity, education; labour force status; interview mode; the number of sources used for information about news and current affairs; type of newspapers used for information about news and current affairs; political party supports/most aligned to; self-assessed general health; whether the respondent suffers from long-term health problems; personality traits-openness, neuroticism, extraversion, conscientiousness and agreeableness-measured using the short 15-item Big-Five inventory (BFI-15); the logarithm of deflated household income (adjusted by the OECDmodified equivalence scale); marital status; the number of dependent children in the household; the square root of household size; housing tenure; whether the respondent is the household financial decision maker; whether the respondent lives in an urban location as well as a set of region of residence dummy variables. These factors have been shown to be strong predictors of voting behaviour (Becker et al, 2017;Arnorsson & Zoega, 2018;Alabrese et al, 2019;Simpson & Startin, 2023).…”