Survey response in probabilistic questions and its impact on inference

Research output: Contribution to journalArticleScientificpeer-review

Abstract

We develop a panel data model of expectations of a continuous outcome variable elicited on a percentage-chance scale. The model explains the location and dispersion of the subjective distributions by socio-economic covariates and unobserved factors. Moreover, it accounts explicitly for non-response, non-informative focal answers, and recall and rounding errors. We apply the model to the expected retirement income replacement rate of Dutch wage workers. We find that incorporating these features of the answering process increases the size and significance of relationships with covariates. The estimates indicate substantial rounding but few focal answers. Respondents tend to stick to a certain answering strategy: non-response, rounding and especially non-informative focal answers are characterized by substantial unobserved heterogeneity across individuals.
Original languageEnglish
Pages (from-to)65-84
JournalJournal of Economic Behavior & Organization
Volume96
Issue numberDecember
DOIs
Publication statusPublished - Dec 2013

Fingerprint

Non-response
Inference
Covariates
Socio-economics
Unobserved heterogeneity
Wages
Workers
Retirement income
Replacement rate
Factors

Keywords

  • expectations
  • survey response
  • rounding
  • recall error
  • unobserved heterogeneity
  • aging

Cite this

@article{ef40a5e09b9349f5a2d5300fee4f2df9,
title = "Survey response in probabilistic questions and its impact on inference",
abstract = "We develop a panel data model of expectations of a continuous outcome variable elicited on a percentage-chance scale. The model explains the location and dispersion of the subjective distributions by socio-economic covariates and unobserved factors. Moreover, it accounts explicitly for non-response, non-informative focal answers, and recall and rounding errors. We apply the model to the expected retirement income replacement rate of Dutch wage workers. We find that incorporating these features of the answering process increases the size and significance of relationships with covariates. The estimates indicate substantial rounding but few focal answers. Respondents tend to stick to a certain answering strategy: non-response, rounding and especially non-informative focal answers are characterized by substantial unobserved heterogeneity across individuals.",
keywords = "expectations, survey response, rounding, recall error, unobserved heterogeneity, aging",
author = "{de Bresser}, J.R. and {van Soest}, A.H.O.",
year = "2013",
month = "12",
doi = "10.1016/j.jebo.2013.09.012",
language = "English",
volume = "96",
pages = "65--84",
journal = "Journal of Economic Behavior & Organization",
issn = "0167-2681",
publisher = "Elsevier Science BV",
number = "December",

}

Survey response in probabilistic questions and its impact on inference. / de Bresser, J.R.; van Soest, A.H.O.

In: Journal of Economic Behavior & Organization, Vol. 96, No. December, 12.2013, p. 65-84.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Survey response in probabilistic questions and its impact on inference

AU - de Bresser, J.R.

AU - van Soest, A.H.O.

PY - 2013/12

Y1 - 2013/12

N2 - We develop a panel data model of expectations of a continuous outcome variable elicited on a percentage-chance scale. The model explains the location and dispersion of the subjective distributions by socio-economic covariates and unobserved factors. Moreover, it accounts explicitly for non-response, non-informative focal answers, and recall and rounding errors. We apply the model to the expected retirement income replacement rate of Dutch wage workers. We find that incorporating these features of the answering process increases the size and significance of relationships with covariates. The estimates indicate substantial rounding but few focal answers. Respondents tend to stick to a certain answering strategy: non-response, rounding and especially non-informative focal answers are characterized by substantial unobserved heterogeneity across individuals.

AB - We develop a panel data model of expectations of a continuous outcome variable elicited on a percentage-chance scale. The model explains the location and dispersion of the subjective distributions by socio-economic covariates and unobserved factors. Moreover, it accounts explicitly for non-response, non-informative focal answers, and recall and rounding errors. We apply the model to the expected retirement income replacement rate of Dutch wage workers. We find that incorporating these features of the answering process increases the size and significance of relationships with covariates. The estimates indicate substantial rounding but few focal answers. Respondents tend to stick to a certain answering strategy: non-response, rounding and especially non-informative focal answers are characterized by substantial unobserved heterogeneity across individuals.

KW - expectations

KW - survey response

KW - rounding

KW - recall error

KW - unobserved heterogeneity

KW - aging

U2 - 10.1016/j.jebo.2013.09.012

DO - 10.1016/j.jebo.2013.09.012

M3 - Article

VL - 96

SP - 65

EP - 84

JO - Journal of Economic Behavior & Organization

JF - Journal of Economic Behavior & Organization

SN - 0167-2681

IS - December

ER -