Recruiting Adolescent Survivors of Cancer in Childhood for a Mail Survey of Self-Reported Health Status: Explaining Response in Terms of Process. William Furlong 1 | Judith Grant 2 | Amy Cranston 3 | Neil Barr 4 | Ronald Barr 5 1 Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada. 2 Oncology, Aberdeen Royal Infirmary, Aberdeen, Scotland. 3 Pediatric Oncology, McMaster Children’s Hospital, Hamilton, ON, Canada. 4 Kinesiology, McMaster University, Hamilton, ON, Canada. 5 Pediatrics, McMaster University, Hamilton, ON, Canada. Keywords: Adolescents, data collection, mail survey, quality of life, questionnaires, response rates, survey method

PDF File

Abstract

Introduction

Mail and telephone surveys are often the only practical ways to collect questionnaire data from geographically dispersed populations.  Mail surveys are considered attractive because of low initial costs to researchers¹.  Mail survey data may be more valid than data from telephone surveys but telephone surveys by well-trained interviewers provide for better control of data collection and fewer missing responses2,3.  Response rate is an important consideration because a low response rate may not provide sufficient sample size for precise parameter estimates and power for statistical hypothesis testing, or produce biased samples undermining the generalizability of study results.
There is no generally accepted standard for minimum response rates.  However, published guidelines for acceptable minimum response rates vary from 50% to 70%4,5.