Monday, March 30, 2015

Here comes the sun...time to launch a survey !

We all feel happier in the sun, well most of us anyway. And it’s no stretch of the imagination to imagine that we change our behavior when we are happy. We might go outside more, we may talk more, probably sunbathe more. It also seems, according to paper by Guéguen and Jacob [1],  that we are more likely to respond to requests for an interview when it is sunny. They tested the hypothesis that respondents would be more likely to comply with a request for a face to face survey when it was sunny rather than when it was cloudy. They controlled for outside temperature and interviewer gender, and found that they had more completed interviews when it was sunny. They did also note that male respondents were more likely to complete surveys when the interviewer was female, but there were no other interactions. Of course this was a personal intercept situation, not a web survey. It may be that the interviewers are happier being outside in sunny weather and this made their invitations to take a survey more attractive.  Either way, weather had an effect on respondents’ co-operation. It would be interesting to see the variability of responses to web surveys with relation to the weather. I can make a guess that being stuck indoors on a beautiful day may not help your recall of shampoo products used in the last few months.

The sun is a good example of an environmental influence on respondent behavior. There is also our genetic make up that can also influence how we behave. I found the paper by Hatemi and McDermott [2] on “The genetics of politics: discovery, challenges, and progress” utterly fascinating. Geneticists have developed analytical techniques to parse out what part of a behavior is genetically derived, environmentally derived or “unique” environmentally derived. It is all based on identical twins, meaning that they genetic material is exactly the same in two individuals. Using some fancy statistics they can get indications as to how much a behavior may be hereditary (genetic), derived from the general experience of the person or from their unique experience as an individual Hatemi and McDermott [2] collated studies on political attitudes from twin and kinship studies over a period of some 30 years. According to them “political knowledge and sophistication” is nearly 60% determined by genetics. On the other hand “political party affiliation” is less than 5% determined by genetics. “Participation and voter turnout” is over 40% determined by genetics. It’s seems our politics grow outside the womb.

The Hatemi and McDermott [2] reviewed studies all dealing with politically oriented characteristics. A more survey interview oriented study that used twins by Littvay, Popa and Fazekas [3] attempted to validate measures of survey response propensity. There is always the question that non-responders may not be the same as responders in characteristics that the survey wants to measure. Non-responders represent a possible bias, they can be fundamentally different from responders.  As part of their study of propensity variables Littvay et al [3] identified as part of a larger study a number of monozygotic (identical) and dizygotic (non-identical twins). The idea was to see if genetic variability was related to the validity of measures used for propensity scoring. An interesting fact is that twins, of any kind, have a tendency to respond to surveys more. Littvay et al [3] found a couple of interesting effects. First non-response in a panel or follow-up situation seems to be highly heritable, that is there is a strong genetic component to it. Secondly non-response to requests for information about close friends or the respondents’ social security number is mediated by environmental influences rather than genetic influences.

As usual, why respondents respond or don’t respond is complicated. Some people just don’t like answering survey questions, it’s a genetic thing. It does seem that if you want to ask a respondent about their friends though, pick a sunny day…..

1[1]  Nicolas Guéguen, Céline Jacob. 2014. “‘Here comes the sun’: Evidence of the Effect of the Weather Conditions on Compliance to a Survey Request “.  Survey Practice,  Vol 7, #5.

2[2] Peter K. Hatemi, Rose McDermott. 2012. “The genetics of politics: discovery, challenges, and progress”. Trends in Genetics, Vol. 28, Issue 10, p525–533.

3[3] Levente Littvay, Sebastian Adrian Popa, Zoltán Fazekas, 2013. “Validity of Survey Response Propensity Indicators: A Behavior Genetics Approach”.  Social Science Quarterly, Vol 94, Issue 2, p569-589.

No comments:

Post a Comment