Response rates and non-response bias in a health-related mailed survey

S. Nakai, Shuji Hashimoto, Y. Murakami, M. Hayashi, K. Manabe, H. Noda

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

A health-related mailed survey was conducted to investigate the effect of follow-up mailings to response rates. In addition, the answer distributions among early and late respondents were compared to check non-response bias. Approximately 3,000 persons aged 40-64 were randomized into two groups; a questionnaire with four pages (twenty-two questions) was assigned to the first group, and a questionnaire with eight pages (thirty-five questions) to the second group. Both questionnaires contained questions of current health status, health-related practice, smoking status, etc. Follow-up mailings were sent twice to non-respondents. Response rates were increased from 38% to 62% by the first follow-up mail, and to 71% by the second follow-up mail. Although the length of the questionnaire did not affect response rates, response rates among the older subjects was higher than the younger subjects. Positive response to current smoking status was 10% lower among early respondents than late respondents, collected after the second follow-up mailing, and response to regular participation in physical/medical checkup was 15% higher among early respondents, whereas there were few differences for answers to other questions. Odds ratios between current health status and several health-related questions may not be biased by late response. However, increased response rates are needed to prevent non-response bias, because there were some differences in responses from follow-up mailings.

Original languageEnglish
Pages (from-to)184-191
Number of pages8
Journal[Nippon kōshū eisei zasshi] Japanese journal of public health
Volume44
Issue number3
Publication statusPublished - 01-01-1997
Externally publishedYes

Fingerprint

Health
Postal Service
Health Status
Smoking
Surveys and Questionnaires
Odds Ratio

All Science Journal Classification (ASJC) codes

  • Medicine(all)

Cite this

Nakai, S. ; Hashimoto, Shuji ; Murakami, Y. ; Hayashi, M. ; Manabe, K. ; Noda, H. / Response rates and non-response bias in a health-related mailed survey. In: [Nippon kōshū eisei zasshi] Japanese journal of public health. 1997 ; Vol. 44, No. 3. pp. 184-191.
@article{18b31e2245c3401e875d3ae26d186e36,
title = "Response rates and non-response bias in a health-related mailed survey",
abstract = "A health-related mailed survey was conducted to investigate the effect of follow-up mailings to response rates. In addition, the answer distributions among early and late respondents were compared to check non-response bias. Approximately 3,000 persons aged 40-64 were randomized into two groups; a questionnaire with four pages (twenty-two questions) was assigned to the first group, and a questionnaire with eight pages (thirty-five questions) to the second group. Both questionnaires contained questions of current health status, health-related practice, smoking status, etc. Follow-up mailings were sent twice to non-respondents. Response rates were increased from 38{\%} to 62{\%} by the first follow-up mail, and to 71{\%} by the second follow-up mail. Although the length of the questionnaire did not affect response rates, response rates among the older subjects was higher than the younger subjects. Positive response to current smoking status was 10{\%} lower among early respondents than late respondents, collected after the second follow-up mailing, and response to regular participation in physical/medical checkup was 15{\%} higher among early respondents, whereas there were few differences for answers to other questions. Odds ratios between current health status and several health-related questions may not be biased by late response. However, increased response rates are needed to prevent non-response bias, because there were some differences in responses from follow-up mailings.",
author = "S. Nakai and Shuji Hashimoto and Y. Murakami and M. Hayashi and K. Manabe and H. Noda",
year = "1997",
month = "1",
day = "1",
language = "English",
volume = "44",
pages = "184--191",
journal = "[Nippon koshu eisei zasshi] Japanese journal of public health",
issn = "0546-1766",
publisher = "Nihon Eisei Gammai/Japanese Society of Public Health",
number = "3",

}

Response rates and non-response bias in a health-related mailed survey. / Nakai, S.; Hashimoto, Shuji; Murakami, Y.; Hayashi, M.; Manabe, K.; Noda, H.

In: [Nippon kōshū eisei zasshi] Japanese journal of public health, Vol. 44, No. 3, 01.01.1997, p. 184-191.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Response rates and non-response bias in a health-related mailed survey

AU - Nakai, S.

AU - Hashimoto, Shuji

AU - Murakami, Y.

AU - Hayashi, M.

AU - Manabe, K.

AU - Noda, H.

PY - 1997/1/1

Y1 - 1997/1/1

N2 - A health-related mailed survey was conducted to investigate the effect of follow-up mailings to response rates. In addition, the answer distributions among early and late respondents were compared to check non-response bias. Approximately 3,000 persons aged 40-64 were randomized into two groups; a questionnaire with four pages (twenty-two questions) was assigned to the first group, and a questionnaire with eight pages (thirty-five questions) to the second group. Both questionnaires contained questions of current health status, health-related practice, smoking status, etc. Follow-up mailings were sent twice to non-respondents. Response rates were increased from 38% to 62% by the first follow-up mail, and to 71% by the second follow-up mail. Although the length of the questionnaire did not affect response rates, response rates among the older subjects was higher than the younger subjects. Positive response to current smoking status was 10% lower among early respondents than late respondents, collected after the second follow-up mailing, and response to regular participation in physical/medical checkup was 15% higher among early respondents, whereas there were few differences for answers to other questions. Odds ratios between current health status and several health-related questions may not be biased by late response. However, increased response rates are needed to prevent non-response bias, because there were some differences in responses from follow-up mailings.

AB - A health-related mailed survey was conducted to investigate the effect of follow-up mailings to response rates. In addition, the answer distributions among early and late respondents were compared to check non-response bias. Approximately 3,000 persons aged 40-64 were randomized into two groups; a questionnaire with four pages (twenty-two questions) was assigned to the first group, and a questionnaire with eight pages (thirty-five questions) to the second group. Both questionnaires contained questions of current health status, health-related practice, smoking status, etc. Follow-up mailings were sent twice to non-respondents. Response rates were increased from 38% to 62% by the first follow-up mail, and to 71% by the second follow-up mail. Although the length of the questionnaire did not affect response rates, response rates among the older subjects was higher than the younger subjects. Positive response to current smoking status was 10% lower among early respondents than late respondents, collected after the second follow-up mailing, and response to regular participation in physical/medical checkup was 15% higher among early respondents, whereas there were few differences for answers to other questions. Odds ratios between current health status and several health-related questions may not be biased by late response. However, increased response rates are needed to prevent non-response bias, because there were some differences in responses from follow-up mailings.

UR - http://www.scopus.com/inward/record.url?scp=0031090933&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031090933&partnerID=8YFLogxK

M3 - Article

C2 - 9175409

AN - SCOPUS:0031090933

VL - 44

SP - 184

EP - 191

JO - [Nippon koshu eisei zasshi] Japanese journal of public health

JF - [Nippon koshu eisei zasshi] Japanese journal of public health

SN - 0546-1766

IS - 3

ER -