The purpose of this study is to verify the effect of survey tools and data collection methods used in the National Criminal Victim Survey before and after its reformation and analyze the process in which each respondent forms his/her responses. Through the verification and analysis, this Study seeks to achieve methodological advancement of the Survey. Under such purpose, the researchers conducted a questionnaire survey using a split-half form adapted for the survey, as well as cognitive interviews. The survey was conducted for all members (14 or older) of 6,000 sample households in Seoul and Gyeonggi regions, applying three different types of methods to three groups each consisting of 2,000 households, respectively. Specifically, Survey Type A collected data through household visit survey using the crime victim questionnaire before reformation; Survey Type B collected data through household visit survey using the reformed crime victim questionnaire; and Survey Type C collected data through household visit online survey using the reformed crime victim questionnaire. By conducting surveys using different survey types and maintaining other aspects identical, the researchers were able to analyze the effectiveness of the questionnaires by comparing the results of Survey Type A and Survey Type B, and analyze the effectiveness of the data collection method by comparing the results of Survey Type B and Survey Type C. As for the cognitive interview, this interview method has gained fresh attention in relation to questionnaire design or uation. This method looks into the process in which the respondent forms his/her responses, which allows the researcher to reduce response errors. This Study used this method for interviews with 10 subjects. The key findings from the surveys and interviews are summarized below.

First, this Study analyzed the effectiveness of the questionnaires, to see if there is any difference in the ability to capturing the victims’ experiences between the previous questionnaire for crime victim survey before 2006 and the current questionnaire used as of 2012. The analysis showed that the previous survey tool before 2006 produced a victim rate 1.8 times higher than the current questionnaire. However, such rate alone does not definitely show us the result of effectiveness analysis. That is, we also need to look into the level of understanding on the questionnaires before and after reformation, as well as the possible response errors. The respondents were fully aware of the benefits of the reformed survey tool, with some of them reporting that the reformed questionnaire had actually reminded them of their experience as crime victims, while they did not experience stimulated recall with the previous questionnaire. However, in order to achieve the original purpose of the reformation, the reformed questionnaire has yet to account for a number of issues in terms of the illustrations given, the format, as well as the method of presenting the questions for recall stimulation. Despite such shortcomings, if we accept the legitimacy of the reformation, we need to improve the questionnaire further through continuous efforts which will include finding a way to achieving the original goal of the reformation, and reviewing the original questionnaire where it resulted in higher victim report rate. Since the National Crime Victim Survey involves national statistics and longitudinal research data, it needs to be conducted in a stable fashion. Therefore, rather than amending and improving the questionnaire for every Survey, it would be better to build the research database for a certain amount of time while systematically analyzing the effectiveness of the improvements, and improve the overall questionnaire at a certain point.

Second, as for the effectiveness of different data collection methods, household visit survey produced relatively higher victim report rate than online survey. However, respondents who experienced online survey showed less positive and more negative responses when asked about their reaction and perception of the survey compared with the respondents who did not：this suggests the need to introduce a means to ensure understanding and convenience of survey responses when using the online survey method.

Thus, this Study looked into the effect of the questionnaire and data collection method on the crime victim report rate by conducting a questionnaire survey, and also explored how the differences in the questionnaire and survey method result in differences in the reaction and attitude of the respondents towards the survey. And the Study re-confirmed the need to understand the process where the respondents form their responses, as well as the need to consider the respondents’ perspective when designing the questionnaire, and revise the structure and contents of the questionnaire accordingly. It should be pointed out, however, that this Study only involved household members residing in Seoul or Gyeonggi regions, or only 10 respondents in case of cognitive interviews. Despite these limitations, the analysis on the effectiveness of the questionnaire and the cognitive interview provided concrete proof for the expected effects of the revised questionnaire, verifying the benefits and limitations of the questionnaires before and after reformation. In addition, this Study made it possible to propose more meaningful ways to apply the new data collection method of online survey. Subsequent studies need to put more efforts in developing survey tools and data collection methods that ensure convenience and efficiency of crime victim report and response, while continuing with the researches that lay the foundation for methodological development of crime victim survey. The researches hole that this Study will be the first of many studies taking the same approach, so as to create stable time-series data based on accuracy, validity, usefulness and reliability through continued accumulation and review of the researches.