Factors of (un)Successful Interviews in Telephone Surveys: A System of Final and Intermediate Codes

This study is devoted to the analysis of such survey method as a telephone interview. The work is aimed at improving the response rate by developing an empirical classification of factors that interrupt telephone interviews and recommendations for reducing the number of unavailable observation units. To achieve this aim 2310 audio recordings of unsuccessful interviews were analyzed. During the listening and analyzing audio files, was developed an empirical typology of incomplete interviews, containing classifications of codes that were assigned to each case depending on the outcome of the conversation. The code grouping resulting from open coding contains 3 major categories, each of which is divided into smaller ones. Each case in the group was given an interpretation, based on which it was possible to write recommendations for each category, helping to prevent failures in the interview in the future. The results showed, firstly, that refusals from interviews occurring in the first minute of a conversation are the strongest threat to the external validity of a telephone survey. Secondly, the instructions for the interview are incomplete and not practical, since it does not provide for the specifics of possible interview outcomes. Moreover, violation of the instructions affects the quality of data collected, but blindly following the instructions without understanding the real communication situation can also cause harm and increase the level of non-responses.