SKIP SEQUENCING: A DECISION PROBLEM IN QUESTIONNAIRE DESIGN.

Area of Science:

• Decision Analysis
• Survey Methodology
• Statistical Inference

Background:

• Questionnaire design involves complex trade-offs between data quality, cost, and respondent burden.
• Skip sequencing is a common technique to reduce survey length and cost, but its impact on data quality is not fully understood.
• Traditional approaches may not adequately address the inferential challenges posed by item nonresponse and response error.

Purpose of the Study:

• To formalize questionnaire design as a decision problem, specifically focusing on skip sequencing strategies.
• To propose the use of an explicit loss function for quantifying the cost-informativeness trade-off in survey planning.
• To provide a framework for choosing optimal survey design strategies under conditions of item nonresponse and response error.

Main Methods:

• Formulating the choice between three survey design options: full inquiry, skip sequencing, and omission.
• Developing a decision-theoretic approach using a loss function to guide design choices.
• Analyzing the impact of item nonresponse and response error on the informativeness and cost of different design options.

Main Results:

• Demonstrates how a survey planner can systematically choose between asking all, using skip sequencing, or omitting a question based on a defined loss function.
• Quantifies the trade-offs between the cost and informativeness of each design option.
• Highlights how inferential problems like nonresponse and response error influence the optimal design choice.

Conclusions:

• An explicit loss function provides a robust method for optimizing skip sequencing in questionnaire design.
• The proposed framework enables informed decisions that balance survey costs with desired data informativeness, even with data quality challenges.
• This approach enhances survey planning by formally integrating cost-benefit analyses with statistical considerations.