• Survey Research
• /
• v.10 no.1
• /
• pp.57-72
• /
• 2009

In Korea, telephone surveys have been used in election forecasting since 1992. In some elections, predictions were excellent, but in some elections, the predictions based on telephone surveys were not good. So, exit polls have been used along with the telephone surveys in predicting election outcomes since 2001 by the major broadcasting networks. Though telephone surveys, in general, have been less accurate than exit polls in election forecasting from 2000 to 2003, they were more accurate in the 2004 General Election than the exit polls. All predictions on the winners by the telephone surveys turned out to be accurate. But such success has not persisted. In the 2008 General Election, the telephone surveys was less accurate than the exit polls and actually its accuracy fell clown to the level of the 2000 General Election. This paper tried to find out. the factors responsible for the fluctuation of the accuracy of telephone polls.

• Survey Research
• /
• v.5 no.1
• /
• pp.1-26
• /
• 2004

The aim of this study lies in answering the question "How to form weighting cells to enhance sample representativeness in telephone, Internet and mobile surveys\ulcorner". For this, we explored 2% raw data of Year 2000 Population and Housing Census of Korea looking for meaningful patterns for ownership of telephones, the usage of Internet and/or mobile phones. We found that telephone coverage rates vary significantly by household size; 84.6% for one member households, contrasting 98.5% for two-or-more member households. Thus, telephone survey samples need to be weighted differently in sub-groups by household size for proportional representation of target population. Searching socio-demographic factors influencing the use of Internet by C5.0 tree models, we found that education levels and the occupation (or housing type, the automobile ownership) are two most important factors in addition to gender and age. Thus, surveyor might form weighting cells by such factors at the stage of post-stratification or set quotas, a priori, proportional to size of the cells by such factors. For mobile surveys, we approached similarly and found that education levels and the occupation (or the automobile ownership, marriage status) are two additional factors that may be used in forming weighing cells or in setting quotas for cells.

• The Korean Journal of Applied Statistics
• /
• v.22 no.2
• /
• pp.375-385
• /
• 2009

For 2007 Korea presidential election, most polls by telephone surveys indicated Lee Myung-Bak led the second runner-up Jung Dong-Young by certain margin. The margin between two candidates can be estimated accurately by averaging individual poll results, provided there exists no systematic bias in telephone surveys. Most Korean telephone surveys via telephone directory are based on quota samples, with the region, the gender and the age-band as quota variables. Thus the surveys may result in certain systematic bias due to unbalanced factors inherent in quota sampling. The aim of this study is to answer the following questions by the analytic methods adopted in Huh et al. (2004): Question 1. Wasn't there systematic bias in estimates of support rates. Question 2. If yes, what was the source of the bias? To answer the questions, we collected eighteen surveys administered during the election campaign period and applied the iterated proportional weighting (the rim weighting) to the last eleven surveys to obtain the balance in five factors - region, gender, age, occupation and education level. We found that the support rate of Lee Myung-Bak was over-estimated consistently by 1.4%P and that of Jung Dong-Young was underestimated by 0.6%P, resulting in the over-estimation of the margin by 2.0%P. By investigating the Lee Myung-Bak bias with logistic regression models, we conclude that it originated from the under-representation of less educated class and/or the over-representation of house wives in telephone samples.

• Survey Research
• /
• v.12 no.2
• /
• pp.77-88
• /
• 2011

Korean telephone surveys have been based on telephone directory and thus criticized for considerable under-coverage. Now, Korean survey institutions progress to random digit dialing (RDD) very actively. But still most surveys are administered by quota sampling, prone to assign heavier weights to social classes with more hours staying indoor. As a practical remedy, time-balanced quota sampling scheme was proposed by Huh and Hwang (2006). This study compares two telephone surveys on TV audience environment in Korea: RDD with conventional quota sampling versus RDD with time-balanced quota sampling.

• Survey Research
• /
• v.9 no.3
• /
• pp.55-69
• /
• 2008

In most telephone surveys in Korea, telephone numbers are selected from the directories. Inevitably, such samples may lack representativeness due to poor coverage rate. To resolve the problem, Kang et al.(2008) implemented RDD(random digit dialing) method for nationwide sampling in Korea. The aim of this study is to compare an RDD sample with a traditional telephone quota sample that were collected independently by two survey institutes commissioned by the KBS-MBC consortium for the 2007 Presidential Election of Korea.

• Proceedings of the Korean Association for Survey Research Conference
• /
• 2002.11a
• /
• pp.241-253
• /
• 2002

Since the 1980's, telephone surveys have been widely used in the world due to a rapid spread of fixed telephones and the development of a variety of sampling techniques. Telecommunications technology has rapidly developed since the early 1990's, and the number of mobile phones grew rapidly in most countries. Also, it has dramatically increased over the last few years. With the growth of households that abandon fixed telephones and move to a mobile only status, fixed telephone household coverage rates are currently decreased in many countries. In this paper, we take a look at the changes and trends in fixed telephone households coverage rates in the European Union, Asia including South Korea, and North America, and deal with the coverage problems in telephone surveys to be faced even in Korea.

• Proceedings of the Korean Association for Survey Research Conference
• /
• 2001.04a
• /
• pp.1.2-35
• /
• 2001

On the election day of the 16th Korea national Assembly, nationwide TV stations such as KBS, SBS and MBC broadcasted election forecasting based on telephone surveys and exit polls. The result turned out to be significantly wrong, undermining general audience's confidence on the 'ientific'surveys. The purpose of this study consists of the followings. I) What form of telephone surveys and exit polls were adopted by survey institutions in real field\ulcorner ii) What were the problems\ulcorner iii) What can we do to make it straight\ulcorner Major findings of the study include the ignorance of call-back rules in telephone survey and the inadequate number of sampled election posts to make satisfactory forecasting. It is pointed out that it is necessary to amend election laws related to election opinion polls to make a substantial progress.

• Survey Research
• /
• v.2 no.1
• /
• pp.1-35
• /
• 2001

On the election day of the 16th Korea National Assembly. nationwide TV stations such as KBS. SBS and MBC broadcasted election forecasting based on telephone surveys and exit polls. The result turned out to be significantly wrong. undermining general audience's confidence on the “scientific” surveys. The purpose of this study consists of the followings. i) What form of telephone surveys and exit polls were adopted by survey institutions in real field\ulcorner ii) What were the problems\ulcorner iii) What can we do to make it straight\ulcorner Major findings of the study include the ignorance of call-back rules in telephone survey and the inadequate number of sampled election posts to make satisfactory forecasting. It is pointed out that it is necessary to amend election laws related to election opinion polls to make a substantial progress.

• Survey Research
• /
• v.9 no.1
• /
• pp.1-22
• /
• 2008

Telephone directories ille still being used as the sampling frame in almost all fixed-line telephone surveys in Korea, causing potentially serious coverage error. RDD (random digit dialing) sampling is an obvious alternative to solve the problem. The aim of this paper is twofold: 1) proposal of RDD methodology suitable to the telephone system of Korea and 2) the identification of socio-demographic and socio-psychological differences between listed-number and unlisted-number respondents. Major findings of RDD telephone survey conducted experimental]y are as follows. 1) Population coverage by telephone directories is 60% or less. 2) Unlisted-number households have statistically larger income compared to listed-number households. 3) Unlisted-number households have smaller family size compared to listed-number households. 4) Unlisted-number respondents are more sensitive about confidentiality, leaks, 5) Unlisted-number respondents are more liberal compared to unlisted-number respondents. These facts suggest that directory-based telephone surveys tend to be biased in socio-economic aspects.

• Asian Journal for Public Opinion Research
• /
• v.9 no.2
• /
• pp.188-211
• /
• 2021

Decreasing response rates to traditional survey methods, like face-to-face and telephone interviews, have led survey practitioners around the world to seek new ways of conducting surveys in recent years." The COVID-19 pandemic exacerbated this problem because it made conducting face-to-face interviews even more difficult than before. For example, it made conducting face-to-face surveys infeasible in 2020 in South Korea, and so the Korean Academic Multimode Open Survey (KAMOS) was unable to conduct a planned face-to-face survey to recruit new panel members. The entire 8,514-member panel, established via two-stage probability-based sampling from 2016 to 2019, was invited to take three online/telephone surveys in 2020. Of these panel members, 1,352 responded to at least one survey in 2020. To test to what extent the panel remained representative of the adult South Korean population, we compared the two groups of panel members: those who responded to at least one survey in 2020 and those who did not. After weighting both groups on the basis of age, sex, and geographical area, we analyzed their responses to some of the questions that were asked during multiple rounds of the face-to-face panel-recruiting interviews. Using Cohen's d for survey items that could be analyzed numerically and Cramér's V for categorical items, we were able to conclude that the respondents to the 2020 surveys were equivalent to the non-respondents in terms of both demographics and in the answers they originally gave to substantive questions on a variety of topics related to social science or public opinion research, including questions about quality of life, societal issue, and politics (Cohen's d items <0.2, 95% CI; Cramér's V items <0.1, 95% CI). This analysis may provide a model for others who wish to test the continued representativeness of their panel or who would like to use a different survey mode or change some other aspect of their methodology and test whether it is equivalent to their former methodology. Our success in building a panel that retained its representativeness may be useful to those in other countries where face-to-face surveys had previously been the norm but are becoming increasingly difficult to conduct.