Participation and contribution in crowdsourced surveys.

Area of Science:

• Social Sciences
• Computer Science
• Data Science

Background:

• Crowdsourced surveys are increasingly used for data collection.
• Understanding participant behavior is crucial for optimizing survey design.
• Incentive structures in crowdsourcing can influence participation and contribution quality.

Purpose of the Study:

• To identify trends in participation quantity and contribution quality in crowdsourced surveys.
• To analyze the relationships between user engagement and the predictive quality of posed questions.
• To investigate participant behavior in collective problem-solving tasks without explicit incentives.

Main Methods:

• Analysis of participation data from three distinct crowdsourced surveys.
• Examination of question submission rates and types.
• Evaluation of response accuracy within defined parameters.
• Correlation analysis between participation metrics and contribution quality.

Main Results:

• Participant question submission rates followed a heavy-tailed distribution across surveys.
• A significant proportion of users submitted non-obvious, predictive questions.
• Over 99.8% of responses adhered to predefined valid ranges, indicating high honesty.
• No significant direct correlation found between participation quantity and contribution quality.

Conclusions:

• Crowdsourced survey participants exhibit predictable collective behaviors.
• An optimal window exists for posing questions, following initial response submissions.
• Future crowdsourced surveys can be improved by strategically prompting question submission.
• Honesty in crowdsourced surveys is high, even without direct incentives.