Selection and Mode Effects in Risk Preference Elicitation Experiments

H.M. von Gaudecker, A.H.O. van Soest, E. Wengstrom

Research output: Working paperDiscussion paperOther research output

269 Downloads (Pure)

Abstract

We combine data from a risk preference elicitation experiment conducted on a representative sample via the Internet with laboratory data on student subjects for the same experiment in order to investigate effects of implementation mode and of subject pool selection. We find that the frequency of errors in the lab experiment is drastically below that of the representative sample in the Internet experiment, and average risk aversion is lower as well. Considering the student-like subsample of the Internet subjects and comparing a traditional lab design with an Internet-like design in the lab gives us two ways to decompose these differences into differences due to subject pool selection and differences due to implementation mode. Both lead to the conclusion that the differerences are due to selection and not to implementation mode. An analysis of the various steps leading to participation or non-participation in the Internet survey leads to the conclusion that these processes are selective in selecting subjects who make fewer errors, but do not lead to biased conclusions on risk preferences. These findings point at the usefulness of the Internet survey as an alternative to a student pool in the laboratory if the ambition is to use the experiments to draw inference on a broad population.
Original languageEnglish
Place of PublicationTilburg
PublisherEconometrics
Number of pages28
Volume2008-11
Publication statusPublished - 2008

Publication series

NameCentER Discussion Paper
Volume2008-11

Fingerprint

Preference elicitation
Risk preferences
World Wide Web
Experiment
Internet survey
Lab experiment
Participation
Usefulness
Inference
Risk aversion

Keywords

  • Risk aversion
  • Internet surveys
  • Laboratory experiments

Cite this

von Gaudecker, H. M., van Soest, A. H. O., & Wengstrom, E. (2008). Selection and Mode Effects in Risk Preference Elicitation Experiments. (CentER Discussion Paper; Vol. 2008-11). Tilburg: Econometrics.
von Gaudecker, H.M. ; van Soest, A.H.O. ; Wengstrom, E. / Selection and Mode Effects in Risk Preference Elicitation Experiments. Tilburg : Econometrics, 2008. (CentER Discussion Paper).
@techreport{35714555f6544522bcb1b35ec93c5f4a,
title = "Selection and Mode Effects in Risk Preference Elicitation Experiments",
abstract = "We combine data from a risk preference elicitation experiment conducted on a representative sample via the Internet with laboratory data on student subjects for the same experiment in order to investigate effects of implementation mode and of subject pool selection. We find that the frequency of errors in the lab experiment is drastically below that of the representative sample in the Internet experiment, and average risk aversion is lower as well. Considering the student-like subsample of the Internet subjects and comparing a traditional lab design with an Internet-like design in the lab gives us two ways to decompose these differences into differences due to subject pool selection and differences due to implementation mode. Both lead to the conclusion that the differerences are due to selection and not to implementation mode. An analysis of the various steps leading to participation or non-participation in the Internet survey leads to the conclusion that these processes are selective in selecting subjects who make fewer errors, but do not lead to biased conclusions on risk preferences. These findings point at the usefulness of the Internet survey as an alternative to a student pool in the laboratory if the ambition is to use the experiments to draw inference on a broad population.",
keywords = "Risk aversion, Internet surveys, Laboratory experiments",
author = "{von Gaudecker}, H.M. and {van Soest}, A.H.O. and E. Wengstrom",
note = "Pagination: 28",
year = "2008",
language = "English",
volume = "2008-11",
series = "CentER Discussion Paper",
publisher = "Econometrics",
type = "WorkingPaper",
institution = "Econometrics",

}

von Gaudecker, HM, van Soest, AHO & Wengstrom, E 2008 'Selection and Mode Effects in Risk Preference Elicitation Experiments' CentER Discussion Paper, vol. 2008-11, Econometrics, Tilburg.

Selection and Mode Effects in Risk Preference Elicitation Experiments. / von Gaudecker, H.M.; van Soest, A.H.O.; Wengstrom, E.

Tilburg : Econometrics, 2008. (CentER Discussion Paper; Vol. 2008-11).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Selection and Mode Effects in Risk Preference Elicitation Experiments

AU - von Gaudecker, H.M.

AU - van Soest, A.H.O.

AU - Wengstrom, E.

N1 - Pagination: 28

PY - 2008

Y1 - 2008

N2 - We combine data from a risk preference elicitation experiment conducted on a representative sample via the Internet with laboratory data on student subjects for the same experiment in order to investigate effects of implementation mode and of subject pool selection. We find that the frequency of errors in the lab experiment is drastically below that of the representative sample in the Internet experiment, and average risk aversion is lower as well. Considering the student-like subsample of the Internet subjects and comparing a traditional lab design with an Internet-like design in the lab gives us two ways to decompose these differences into differences due to subject pool selection and differences due to implementation mode. Both lead to the conclusion that the differerences are due to selection and not to implementation mode. An analysis of the various steps leading to participation or non-participation in the Internet survey leads to the conclusion that these processes are selective in selecting subjects who make fewer errors, but do not lead to biased conclusions on risk preferences. These findings point at the usefulness of the Internet survey as an alternative to a student pool in the laboratory if the ambition is to use the experiments to draw inference on a broad population.

AB - We combine data from a risk preference elicitation experiment conducted on a representative sample via the Internet with laboratory data on student subjects for the same experiment in order to investigate effects of implementation mode and of subject pool selection. We find that the frequency of errors in the lab experiment is drastically below that of the representative sample in the Internet experiment, and average risk aversion is lower as well. Considering the student-like subsample of the Internet subjects and comparing a traditional lab design with an Internet-like design in the lab gives us two ways to decompose these differences into differences due to subject pool selection and differences due to implementation mode. Both lead to the conclusion that the differerences are due to selection and not to implementation mode. An analysis of the various steps leading to participation or non-participation in the Internet survey leads to the conclusion that these processes are selective in selecting subjects who make fewer errors, but do not lead to biased conclusions on risk preferences. These findings point at the usefulness of the Internet survey as an alternative to a student pool in the laboratory if the ambition is to use the experiments to draw inference on a broad population.

KW - Risk aversion

KW - Internet surveys

KW - Laboratory experiments

M3 - Discussion paper

VL - 2008-11

T3 - CentER Discussion Paper

BT - Selection and Mode Effects in Risk Preference Elicitation Experiments

PB - Econometrics

CY - Tilburg

ER -

von Gaudecker HM, van Soest AHO, Wengstrom E. Selection and Mode Effects in Risk Preference Elicitation Experiments. Tilburg: Econometrics. 2008. (CentER Discussion Paper).