Using Point Forecasts to Anchor Probabilistic Survey Scales

How to cite

Glas, Alexander. 2024. "Using Point Forecasts to Anchor Probabilistic Survey Scales." AEA RCT Registry. AEARCTR-0013118. https://doi.org/10.1257/rct.13118-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Alexander Glas*

* Primary investigator

Registry ID
AEARCTR-0013118
DOI
https://doi.org/10.1257/rct.13118-1.0
First registered
Last updated
Status
Completed
Trial dates
2023-02-15 – 2023-03-01
Countries
Germany ()
Registry record
http://www.socialscienceregistry.org/trials/13118

Abstract

We conduct an experiment where a random subset of the participants in the Bundesbank's household panel receive personalized response scales, centered at each respondent's point forecast. Personalized scales mitigate the impact of central tendency bias and eliminate the need to adjust the scale when inflation falls outside the range for which the scale was designed.

Experimental design

Our treatment design was included in Wave 38 of the monthly BOP-HH. The survey includes three questions on inflation expectations: The answer to the first (qualitative) question determines whether the respondent is asked about the rate of inflation or deflation in the second question. The second (quantitative) question asks for a point forecast of the future inflation/deflation rate. The treatment intervention occurs after this question. The reported point forecast, which may include up to one decimal place, determines the center of the response scale for respondents in the Personalized treatment. The respondents in Baseline receive the standard scale which is centered at zero inflation. We then compare the point forecasts and probabilistic expectations of both groups.

Design details

Randomization method
Randomization is done in office by a computer
Randomization unit
Individual-level randomization
Clusters
One cross-section of households.
Observations
Approximately 3,500 household heads.
Arms
2,100 (Personalized) versus 1,400 (baseline).
Primary outcomes
Point forecasts and probabilistic expectations for the inflation rate in Germany.

Cited by

No citation data available for this registration yet.

Keywords