Response Propensity

Determinants of Response Propensity

Response propensity is essentially a characteristic of the sampled person. Response propensities vary across persons according to their sociodemographic characteristics and various psychological predispositions. For example, persons in large households are easier to be contacted than those who live alone, whereas persons who are socially isolated are less likely to grant an interview than are people who are socially integrated. Response propensity can be regarded as a characteristic of the sampled person only, if defined as the probability that the sampled person will become a respondent in a random survey or, alternatively, as the probability that the sampled person will become a respondent in a survey with “average characteristics.” Considering a person's “baseline” response propensity within the characteristics of a certain survey then determines the probability that the person will be a respondent to that particular survey. So, in fact, response propensities are affected both by characteristics of the sampled person and by survey characteristics (see Figure 1).

Figure 1 Determinants of response propensity

Contactability and Cooperation

In the process of becoming a respondent, two important steps can be distinguished. The first step is that the respondent has to be contacted. The second step is that after being contacted, the respondent must be willing and able to cooperate with the surveyor's request. This observation suggests a straightforward decomposition of response propensity into two parts: contactability and cooperation.

The contactability Pcontact, ij is the probability that individual i is contacted for survey j at any given moment in time. Typical characteristics of the sampled person that affect the contactability include factors such as (a) how often the person is at home, (b) whether there are physical impediments to contacting the person, and (c) to what extent the person is prepared to answer any incoming survey requests. Survey characteristics that affect contactability include (d) the number of contacts made by the interviewer, (e) the length of the survey period, and (f) the timing of the contacts. [Page 757]For example, contact attempts in the evenings and on weekends are usually more successful than contact attempts at other times, thereby raising the response propensity.

...