Weryfikacja zewnętrzna informacji uzyskanych w wywiadzie kwestionariuszowym

Abstract

Results of external verification of two kinds of information obtained by means of questionnaire interviews in methodological investigations conducted in Włocławek in 1973 are presented. Each kind of information was verified by using a different method. In Part I are presented the results of verification of 7 questionnaire questions, among which 6 related to opinions. Information on the same topics obtained through in-depth interviews was used as a verification criterion. An in-depth interview was used along with an interview about an interview which constituted part of the method permitting to precisely characterize and evaluate the quality of the answers given by respondents. Part II presents the results of verification by means of the data contained in the files of work establishments and relating to 11 information items obtained in questionnaire interviews. The items concerned: the year and place of birth, educational level, number of completed training courses, the year in which the respondent started his work, level of wages in June, number of days of absence at work (warranted and unwarranted), number of late arrivals to work as well as of penalties and rewards recorded in the files. Discrepancies between interview answers and the files which appeared most frequently (30% of respondents) related to the numbers of days of absence at work and the numbers of late arrivals to work. These discrepancies not always indicate a contradiction between the information obtained in interview and the reality as the file records cannot be treated as absolutely reliable. By comparing answers about wages with the corresponding data from the files it was ascertained that 57% of respondents tended to overstate their wages by over 100 zlotys. This tendency is contrary to the findings presented by K. Lutyńska in the study contained in the present volume. It is recommended to apply simultaneously the less and more reliable methods of verification in order to evaluate the quality of the former methods as they can be used much more frequently than the latter.