Homogenizing Responses to Different Survey Questions on the Same Topic: Proposal of a Scale Homogenization Method...

Area of Science:

• Social Sciences
• Psychology
• Survey Methodology

Background:

• Survey data are frequently used for cross-national and longitudinal comparisons.
• Variations in response scales (e.g., number of options, labeling) hinder data comparability.
• Happiness research data is often incomparable due to diverse response scales.

Purpose of the Study:

• To provide an overview of Scale Homogenization methods for survey data.
• To propose and evaluate a new Reference Distribution Method for improving data comparability.
• To address the issue of incomparable happiness research data.

Main Methods:

• Review of early Scale Homogenization methods: Linear Stretch and Semantic Judgement of Fixed Word Value.
• Detailed discussion of the Semantic Judgement of Word Value in Context Method.
• Proposal and application of a new Reference Distribution Method.

Main Results:

• The Reference Distribution Method produces comparable time series across different questions.
• This new method effectively corrects for discontinuities in survey data.
• Application to Dutch happiness data (1989-2009) demonstrates its utility.

Conclusions:

• Scale Homogenization is crucial for effective survey data comparisons.
• The Reference Distribution Method offers a robust solution for creating comparable time series.
• This approach enhances the reliability and comparability of happiness research data.