The next speaker in this session at the 5th Data Donation Symposium at the Weizenbaum-Institut in Berlin is Theo Araujo, whose interest is in inferential biases in data donation samples. Willingness to donate data to such studies is very unevenly distributed, and this also means that we lack data on non-participants; but do such biases on donation patterns actually result in biases in the eventual analysis?
This study worked with an existing true probability sample in the Netherlands, and investigated personality traits, ideological beliefs, and digital skills as the main sources of potential bias in the willingness to donate data. The focus in examining the transition from invitation to donation was on the behavioural variables of social media use, search engine use, news readership, and e-commerce behaviour.
Of the 5,000 invited participants, only about 500 donated data from at least one platform. Agreeableness increases consent, while conscientiousness decreases; agreeableness also increased the number of platforms from which data were donated. Trust in science and left-wing beliefs also increased consent and donations. Digital skills and age also increased, while privacy concerns tended to decrease consent and donation.
The full sample of participants and the sample of data donors did not differ substantially across many of the possible characteristics, however; participants are more agreeable and open, and less extroverted; more left-wing and more trusting in science; more educated and with better digital skills; but these differences are not strong enough to suggest that data donation studies conducted with this smaller group of participants will be unrepresentative of the broader representative panel.
