Relationship Between Past Survey Burden and Response Probability to a New Survey in a Probability-Based Online Panel

Area of Science:

• Survey Methodology
• Social Sciences
• Data Analysis

Background:

• Survey burden, including questionnaire length and frequency, can impact participant response rates.
• Understanding these effects is crucial for maintaining high-quality data in probability-based Internet panels.

Purpose of the Study:

• To examine how survey burden (questionnaire length and survey frequency) affects response probability in an Internet panel.
• To identify specific subgroups disproportionately affected by survey burden.

Main Methods:

• Idiographic analysis using a latent Markov chain model.
• Monte Carlo methods to estimate individual response probabilities.
• Pooling individual estimates to analyze overall relationships and subgroup effects.

Main Results:

• A positive relationship was found between questionnaire length and response probability overall.
• No significant relationship was observed between survey frequency and response probability.
• Longer questionnaires decreased response rates among racial/ethnic minorities and introverted participants.
• Frequent surveys decreased response rates among participants with large households.

Conclusions:

• Questionnaire length is a key factor influencing response probability, particularly for specific demographic and psychographic subgroups.
• Panel management strategies should consider targeted interventions for vulnerable subgroups to mitigate negative impacts of survey burden.
• Further research may explore nuanced effects of survey frequency on diverse populations.