@article{oai:kansai-u.repo.nii.ac.jp:00018454,
 author = {林, 英夫 and 村田, 晴路},
 issue = {1},
 journal = {関西大学社会学部紀要},
 month = {Nov},
 note = {郵送調査の方法論的研究では，返信率を高めることに多大な努力が払われてきた。しかし，わが国では，調査データの質について，信頼性と妥当性の両面から検討した研究は少ない。本稿では1995年12月から翌年2月にかけて，山口県萩市で成人女性500名を対象に2か月の間隔をおいて実施した2回にわたる郵送調査の結果に基づき，応答の正確度および安定度ならびに返信時期を異にする応答の質的差異を検討した。その結果，以下の諸点が明らかにされた。返信率65％程度を得れば，人口統計的指標に関し，母集団ないし計画標本と達成標本の分布間に大きな歪みは生じない。また事実に関する応答は正確で安定しているが，購入経験を尋ねる質問の応答の安定度は，購入品目によってばらつきが見られる。早期返信者と後期返信者それぞれの基本的属性に差異は認められず，返信時期を左右する要因は，調査内容に対する調査対象者の関心の程度によるらしいことなどがわかった。, Mail survey researchers have been preoccupied with methodological research in order to increase response rates to mail questionnaires. However, in Japan there has been less emphasis among researchers in investigating means to improve reliability and validity of data obtained from mail questionnaires. Our primary objective in this study is to examine the stability or consistency and accuracy of responses, and response quality of early versus late respondents. The study was carried out in a rural city where the first questionnaire was mailed to a random sample of 500 adult women in December 1995 and the second identical questionnaire was mailed to the same sample in February 1996. The results suggest that if we can achieve about a 65% response rate, sociodemographic distributions are similar among the population, the designed sample and the obtained sample, and that responses about factual information are accurate and stable, but responses about purchasing experience vary according to the item involved. The examination of the composition of respondents revealed no basic differences in demographic characteristics between earlier and later respondents, with response speed seeming to depend upon the degree of interest of respondents towards mail questionnaires.},
 pages = {171--189},
 title = {郵送調査における応答誤差 : 応答の正確度および安定度ならびに返信時期による応答の差異},
 volume = {28},
 year = {1996}
}