Optimizing Call Patterns for Landline and Cell Phone Surveys.

Area of Science:

• Survey Methodology
• Public Health Research
• Data Science

Background:

• Cell phone surveys are increasingly prevalent, posing challenges for cost-effective data collection and high response rates.
• Traditional landline survey calling strategies may not be optimal for mobile phone surveys, necessitating tailored approaches.
• Previous research indicates differences in optimal calling times and contact rates between landline and cell phone samples.

Purpose of the Study:

• To investigate and compare calling outcome differences between landline and cell phone surveys.
• To identify key factors influencing respondent contact and interview completion for different sample types.
• To provide data-driven recommendations for enhancing the efficiency of cell phone surveys.

Main Methods:

• Utilized paradata from the 2010 and 2011 National Flu Surveys.
• Employed logistic regression models to predict respondent contact and interview completion.
• Analyzed the impact of calling day, time, call history, and time intervals on survey outcomes.

Main Results:

• Significant differences in calling outcomes were observed between landline and cell phone surveys.
• Logistic regression models identified specific calling parameters affecting contact and completion rates.
• Contact and completion likelihoods varied based on day of week, time of day, and call attempt history.

Conclusions:

• Calling strategies must be adapted for cell phone surveys to maximize cost-effectiveness and response rates.
• Understanding paradata is essential for optimizing survey fieldwork and improving data collection efficiency.
• Findings offer practical insights for researchers conducting mobile phone surveys to increase respondent engagement.