Interrupted interviews reduce respondents’ cooperation and can lead to systematic sample shifts, which calls into question the representativeness of the survey data obtained. Sensitive survey topics, such as the discussion of the Russian-Ukrainian conflict, may on more occasions encourage respondents to stop discussing them. Therefore it is especially important to investigate the matter of interrupted interviews in such surveys. The purpose of this article is to identify and describe the reasons for interview interruptions in surveys about military operations in Ukraine. Research methods: content analysis of audio recordings of interviews with respondents and traditional qualitative analysis of documents, namely audio recordings of interviews with respondents who stopped the survey prematurely. Quantitative research (content analysis) allows you to measure the prevalence of reasons for respondents walking out of interviews, and qualitative analysis of audio recordings of interviews provides material to describe the narratives behind these reasons for walking away from a survey. Selection consists of two stages. At the first stage the sample is random, stratified by federal districts. It represents the Russian population by gender, age and federal districts of residence of the respondents. During the first stage, the sample consisted of 2507 interviews, including 888 interviews that were shut down. At the second stage, the sample is solid, formed by the method of the main array: 538 out of 888 interrupted interviews were listened to. Here are the main findings of the study. The reasons for shutting down an interview are divided into avoidable and unavoidable. Each of these groups of causes is classified into specific types of causes of interruptions. The ratio of avoidable and unavoidable causes in the listened to sample of recordings was 68 to 470 or 13% to 87%, that is, the potential to reduce the number of interruptions is quite small. It is worth noting how the length of the questions and the questionnaire as a whole influenced the number of interview interruptions: the longer the questionnaire and its questions, the more interruptions occurred. Therefore, conciseness when it comes to wording questions (reducing their length, avoiding tabular questions in telephone surveys, reducing the number of options in nominal scales), reducing the length of the questionnaire to 10 minutes and working with the emotional burnout of interviewers will reduce the number of shut down interviews. The sensitivity of military-political topics also increases the number of interruptions, but it is not yet possible to find a decent solution to this problem. All that researchers can do so far is minimize the number of military-political questions in the questionnaire, formulate them in the most comfortable way for respondents (hypothetically, with the possibility of refusal to answer, etc.) and “dilute” the questionnaire with less sensitive topics.
