Increasing response rates in telephone surveys: a randomized trial

Area of Science:

• Public Health Research Methods
• Survey Methodology
• Health Services Research

Background:

• Response rates and selection bias are critical in population-based research, influenced by sampling frames and contact methods.
• Evaluating the impact of pre-contact strategies and telephone survey validity is essential for reliable health data collection.

Purpose of the Study:

• To determine if a mailed letter preceding telephone contact increases response rates.
• To assess the concurrent validity of telephone surveys for collecting health research and service data.

Main Methods:

• Randomly selected households (n=2000) from electronic white pages were divided into two groups: one received a mailed letter before telephone contact, the other did not.
• Within households, a respondent aged 18+ was randomly selected using a Kish grid for telephone interviews; interviewers were blinded to the letter group allocation.

Main Results:

• The overall response rate was 68%. Households receiving a mailed letter had a 76% response rate compared to 60% for those who did not.
• Random respondent selection via Kish grid slightly decreased response rates (<10%).
• Internal validity was high (Kish grid applied correctly in 93% of cases). External validity showed no significant differences in population estimates compared to face-to-face surveys.

Conclusions:

• Preceding telephone contact with a mailed letter significantly and cost-effectively increases survey response rates.
• Telephone surveys are a valid and useful method for collecting data in health research and service evaluations.