Effects of Interviewer Attitudes and Behaviors on Refusal in Household Surveys

Abstract

Interviewers play a crucial role in gaining cooperation from a sample unit. This paper aims to identify the interviewer characteristics that influence survey cooperation. Of principal interest to survey practitioners are interviewer attributes associated with higher cooperation rates, particularly among sample members whose characteristics are traditionally associated with a lower probability of response. Our data source is unusually rich, in that it contains extensive information on interviewers including their attitudes and behaviors which is linked to detailed information on both responding and nonresponding sample units. An important value of the data is that they permit examining a host of as yet unanswered questions about whether some interviewer attributes stimulate cooperation among some respondents but not others. In short, we investigate if some sample units react favorably to certain interviewer characteristics. A multilevel cross-classified logistic model with random interviewer effects is used to account for clustering of households within interviewers due to unmeasured interviewer attributes, and for the cross-classification of interviewers within areas. The model allows for statistical interactions between interviewer and household characteristics.We find that interviewer confidence and attitudes towards persuading reluctant respondents play an important role in explaining between-interviewer variation in refusal rates. We also find evidence of interaction effects between the interviewer and householder, for example with respect to gender and educational level, supporting the notion of similarity of interviewers and respondents generating higher cooperation. The results are discussed with respect to potential implications for survey practice and design.

Representation of Vulnerability and the Elderly. A Total Survey Error Perspective on the VLV Survey

“…However, to reduce the number of negative answers as much as possible, we chose to have some interviewers specialize; they were requested to make the first phone contact, not only for themselves but for the entire team. The crucial nature of this first interaction has been studied for some 15 years (see Snijkers et al 1999;Durrant et al 2010) and now appears to be an explanation for national variations, as well as being likely to affect the comparability of results (Blom et al 2011). In our case, certain interviewers clearly proved to be more effective than others in obtaining the cooperation of potential participants.…”

“…The latter category is mainly related to the questionnaire's design, clearness of the definitions, terms used, format, etc. Category c) considers the environmental conditions of the interview (e.g., place, presence of other persons), as well as the collection methodology used, the standardization of the interview conditions, the training and supervision of the interviewers, and also the monitoring and observation of these variables by the interviewers and researchers (Durrant et al 2010;Groves and Lyberg 2010;Marchese 2011;Blom et al 2011).…”

Do low survey response rates bias results? Evidence from Japan

“…An organization's procedures, including training given to field workers, could influence response rates. Similarly, field workers vary in their abilities (including training, experience, and basic personality traits), which can affect the likelihood of a potential respondent cooperating (Dijkstra and Smit 2002;Durrant et al 2010;Maynard, Freese, and Schaeffer 2010;West and Olson 2010). Factors include their outward appearance (mannerisms, age, size, facial expressions, race/ethnicity, and clothing), their cleverness in locating places, and their persuasive abilities.…”

Using Prior Wave Information and Paradata: Can They Help to Predict Response Outcomes and Call Sequence Length in a Longitudinal Study?

“…Interviewer observations were reported to have an influence on response propensities by Durrant et al (2010) and Durrant and Steele (2009). West (2013) also reported that interviewer observations were predictive of nonresponse in a cross-sectional survey.…”

“…Table 1 presents the distributions of the three response variables used in the analysis. The literature has identified a range of variables associated with household nonresponse, including sociodemographic characteristics of respondents, physical impediments of the property, social and environmental attributes of households, interviewer observations about households as well as call timing and frequency of calls (Bethlehem et al 2011;Groves and Couper 1996;Groves and Couper 1998;Durrant et al 2010;Durrant and Steele 2009;West 2013;Groves 2006). Groves and Couper (1998) concluded that physical impediments to accessing the household, at-home patterns of the household (larger household sizes, presence of elderly persons or young children), and the timing (day of the week and time of the day) and number of interviewer visits to the unit as well as residence of household (urban or rural) were significantly associated with household contactability.…”