Referierte Aufsätze Web of Science
Maximilian Longmuir, Markus M. Grabka
In: Survey Research Methods 18 (2024), 3, S. 251-261
This paper investigates how the number of brackets and the choice of upper cutoffs in grouped data affect the metric approximation of income and wealth. The literature currently lacks a definition of what should be considered too few brackets or too-low cut-offs. Using German survey data, we show that more than six (eight) brackets and an upper cut-off at the 95th (97th) percentile are sufficient to provide an adequate approximation of the income (wealth) distribution.
Keywords: grouped data, wealth, income, survey design
DOI:
https://doi.org/10.18148/srm/2024.v18i3.8187