[Voluntary versus involuntary sample surveys in the microcensus]

Area of Science:

• Sociology
• Statistics
• Demography

Background:

• Voluntary sample surveys may introduce non-response bias.
• Involuntary surveys, like the German microcensus, sample the entire population.
• The validity of voluntary surveys depends on respondent data mirroring the total population.

Purpose of the Study:

• To compare voluntary and involuntary sample surveys in West Germany.
• To assess potential non-response bias in voluntary surveys using microcensus data.

Main Methods:

• Comparative analysis of data from the German microcensus (1985-1987) and voluntary test sample surveys.
• Statistical comparison of demographic and socio-economic variables between survey types.

Main Results:

• The null hypothesis that respondent populations do not differ from the total population was rejected.
• Significant discrepancies were found in region, community size, age, marital status, income, and social security.
• Non-response bias in voluntary surveys was confirmed.

Conclusions:

• Voluntary sample surveys may not yield statistically valid inferences for the total population.
• Involuntary surveys are crucial for accurate demographic and socioeconomic data collection.
• Non-response bias poses a significant challenge in survey research.