During February 14-22, 2023, the Kyiv International Institute of Sociology (KIIS) conducted its own all-Ukrainian public opinion survey "Omnibus". By the method of computer-assisted telephone interviews (CATI) based on a random sample of mobile phone numbers (with random generation of phone numbers and subsequent statistical weighting), 2,002 respondents living in all regions of Ukraine (except the Autonomous Republic of Crimea) were interviewed. The survey was conducted with adult (aged 18 and older) citizens of Ukraine who, at the time of the survey, lived on the territory of Ukraine (within the boundaries controlled by the authorities of Ukraine until February 24, 2022). The sample did not include residents of territories that were not temporarily controlled by the authorities of Ukraine until February 24, 2022 (AR of Crimea, the city of Sevastopol, certain districts of the Donetsk and Luhansk oblasts), and the survey was not conducted with citizens who left the country after February 24, 2022.
Formally, under normal circumstances, the statistical error of such a sample (with a probability of 0.95 and taking into account the design effect of 1.1) did not exceed 2.4% for indicators close to 50%, 2.1% for indicators close to 25%, 1.5% - for indicators close to 10%, 1.1% - for indicators close to 5%.
Under conditions of war, in addition to the specified formal error, a certain systematic deviation is added. In particular, if back in May, among all the respondents we interviewed, 2.5-4% lived in the territories occupied after February 24 (and this corresponded to the percentage of those who live there, because the generation of telephone numbers was random), now, due to the occupiers turning off the telephone connection, we managed to interview only 2 respondents (from 2002) who currently live in occupied settlements. It is important to note that although the views of the respondents who lived in the occupation were somewhat different, the general tendencies were quite similar. That is, the impossibility of interviewing such respondents does not significantly affect the quality of the results. There are other factors that can affect the quality of results in "wartime" conditions (see Annex 3). In general, we believe that the obtained results are still highly representative and allow a fairly reliable analysis of public moods of the population.