Search
Close this search box.

Law and Social Change Survey Database

2019 Law and Social Change Survey R5 (2)

【Legal Cutulre and Social Change Survey】

 

Methodology

This survey will be carried out by means of face-to-face interview. The interviewer will visit the address provided in the sampling list. With the aid of a tablet computer as well as a computer-assisted personal interviewing system (hereafter referred to as “CAPI”), the interviewer will then randomly select an interviewee from the household. The interview will be conducted in a question-and-answer manner, through which relevant information will be collected. The interviewer is also required to record the date, time, and result of each household visit in the CAPI system. The interviewer should also record any unexpected situation that may have arisen during the interview.

Research Design

  • Survey Population 
    The survey population comprises regular Taiwanese citizens aged 18 and above (i.e. born before December 31, 2000) while the sampling frame is defined by the household registration address provided by the Ministry of the Interior. Within the selected households, regular occupants who meet the age criterion above will then undergo intra-household random selection in order to choose a single interviewee. However, the said intra-household random selection will not be carried out in case the selected address turns out to be that of a military unit, hospital, nursing home, school, vocational training center, dormitory, or prison.
  • Sampling Design
    The The Both the pilot and official surveys make use of stratified sampling and an inflated sample size. The stratified sampling method is adopted in order to increase the accuracy of any estimations made, while an inflated sample size will ensure that the target number of interviews are made, especially when potential respondents may refuse an interview, or may not be successfully contacted.

    •  Stratified Sampling
      The stratified sampling method applied in this research is based on the concepts of demographic structure and economic variables in human ecology. The 358 administrative divisions (including townships, cities, and districts) in Taiwan were divided into seven strata, which were further divided into 19 substrata corresponding to the six regions of Taiwan. This method will ensure that selected households are representative of all Taiwanese regions and eliminate the need to verify, after every selection, whether the selected sample fulfills the requirement of being representative of a particular region, hence simplifying the sampling process.
    •  Inflation of Sample Size
      Reference was made to the completion rate of a recent large-scale project carried out by the Center for Survey Research, RCHSS, Academia Sinica when determining the inflation index (that is, the number of times the sample size should be multiplied) of the survey’s sample, namely the number of townships, villages, and addresses to be selected for survey. The inflation index of the pilot survey was found to fall between 3.8 and 6.1, while the inflation index of the official survey was found to fall between 1.6 and 5.6. inflation indices of townships, cities, districts, and villages to be surveyed in the pilot and official surveys

 

Survey Method

  • Pilot Survey
    The pilot survey, which aimed to complete 120 successful interviews, adopted the stratified multi-stage systematic sampling method coupled with the use of intra-household respondent selection by interviewers after entering the household. In order to reduce surveying costs, the pilot survey was not conducted in Penghu, Hualien, and Taitung counties. In addition, only the first five strata were considered for pilot survey, while the pilot survey locations (or administrative divisions) were determined after considering the interviewers’ circumstances. As for sample allocation, the population proportion of each stratum was calculated based on the demographic data provided by the Household Registration Department of the Ministry of the Interior at the end of December 2018, through which the number of administrative divisions (including townships, cities, districts, and villages) to be selected for the five strata was determined. A total of 123 successful interviews were expected to be conducted during the pilot survey.
    The first stage of sampling involved selecting the townships, cities, and districts to be surveyed, while the villages to be surveyed were selected in the second stage. The addresses of households to be surveyed were selected in the third stage of sampling, while the fourth and final stage of sampling saw interviewers select a survey respondent from the selected household by making use of the intra-household respondent selection system developed by Hung Yung-Tai (2001) to increase the probability of the selection of younger respondents. As stated above, the inflation index of the pilot survey falls between 3.8 and 6.1, translating to the selection of 577 survey respondents.
  • Official Survey
    The official survey adopted the stratified multi-Stage probability proportional to size (PPS) sampling method coupled with intra-household sampling. That is, probability proportional to size sampling was applied in the first three sampling stages of each stratum, after which the sample size was inflated and intra-household sampling was carried out. The first stage of sampling involved selecting the townships, cities, and districts to be surveyed, while the villages to be surveyed were selected in the second stage. The addresses of households to be surveyed were selected in the third stage of sampling, while the fourth and final stage of sampling saw interviewers select a survey respondent from the selected household. However, after the geographical uniqueness of administrative units in Hualien and Taitung counties was considered, it was decided that the first sampling stage for these two counties would be tweaked to select villages (instead of townships, cities, or districts) to be surveyed, in order to ensure sufficient representation of the people of Hualien and Taitung. The next sampling stage for Hualien and Taitung saw the selection of addresses of households to be surveyed, followed by the selection of survey respondents. A total of 2,000 successful interviews were expected to be conducted during the official survey.
    As for sample allocation, the population proportion of each stratum was calculated based on the demographic data provided by the Household Registration Department of the Ministry of the Interior at the end of December 2018, through which the number of administrative divisions (including townships, cities, districts, and villages) to be selected for the five strata was determined. As stated above, the inflation index of the official survey falls between 1.6 and 5.6, translating to the selection of 7,300 survey respondents. The table below describes the sample design, as well as how the sample was inflated and allocated.

 

Survey Result

  • Pilot Survey
    The pilot survey sampled 577 households, of which 157 were successfully surveyed.
  • Official Survey
    The official survey had a sample size of 7,300 households, of which 2,052 were successfully surveyed, exceeding the expectation of 2,000 households. According to the American Association for Public Opinion Research (AAPOR) calculator, the survey has an unweighted response rate (RR1) of 31.32% and a refusal rate (REF1) of 11.63%.

 

𝐓𝐎【2019 臺灣法律與社會變遷調查第5期 (2)】

Questionnaires

很抱歉,目前沒有相關的資料。

很抱歉,目前沒有相關的資料。