I am currently in Accra, Ghana for a cultural psychology project organised by the University of Amsterdam in cooperation with the University of Ghana at Legon. As part of the trip, I have been doing some data collection for the Ghana part of a large cross-cultural study on perceptions of power and leadership (run by Eftychia Stamkou of UvA). Here’s three things I learned from collecting data for a survey experiment:
- Participant recruitment is easy (at least sometimes): for our study, we recruited university students at the Legon campus of the University of Ghana. Much to my surprise, recruiting participants for an (unpaid) 20 minute survey was very easy: when we knocked on doors in the university dorms, most residents were happy to participate in our survey. (In contrast, a group recruiting teachers for an hour-long study had a much harder time to find willing participants).
- Asking about age can be sensitive: in discussing our questionnaire with our Ghanaian collaborators, we were told that questions about the participant’s age can pose some problems. For example, there is a perceived threshold marriage age – 25 years – and participants may feel uncomfortable with indicating that they are older than that (although we did not ask about marital status). We resorted to providing participants with the opportunity to indicate an age range instead; and participants appeared to be happy with that.
- Likert scales are not always intuitive: one of the most common survey response formats in psychological research are Likert scales, often times with five or seven points. Many of our participants seem to have used our seven-point scales as three-point scales however, with most or all answers on the options one, four, and seven. In a place where many participants may be unfamiliar with Likert scales, using a three-point scale (or binary measure with ‘don’t know’ option) may be preferable.
Overall, my experience collecting data here in Accra has been quite positive – it certainly went much faster than expected. I suspect that collecting more representative data, or in more rural locations, will be more difficult; but that’s for some future experience. So far, I’ve learned some very useful things about doing research here, and about the research done here – more on that in my next post.