Time-Aggregation Bias in Hazard-Rate Models with Covariates

Trond Petersen, Kenneth W Koput

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

This article elaborates the results found in Petersen (1991) by discussing how to minimize time-aggregation bias in hazard-rate models with measured covariates. It first considers a model with a single categorical covariate with h categories. It derives analytically the bias of the estimator that assumes the durations are exactly measured, when the durations are in fact rounded up to the nearest integer. Second, the article conducts both large-and small-sample Monte Carlo studies for several estimators of the covariate effects in the presence of time aggregation. There are three central findings. (1) It is shown that the likelihood that adjusts for the grouped nature of the duration measures recaptures the parameters very well. (2) The estimator that assumes that the durations are exactly measured, when they in fact are rounded up to the nearest integer, is biased in several ways. (3) The estimator that uses the midpoint adjustment suffers from the same weaknesses as the estimator that rounds up to the nearest integer, but to a lesser degree.

Original languageEnglish (US)
Pages (from-to)25-51
Number of pages27
JournalSociological Methods & Research
Volume21
Issue number1
DOIs
StatePublished - 1992
Externally publishedYes

Fingerprint

aggregation
trend
time

ASJC Scopus subject areas

  • Social Sciences (miscellaneous)
  • Sociology and Political Science

Cite this

Time-Aggregation Bias in Hazard-Rate Models with Covariates. / Petersen, Trond; Koput, Kenneth W.

In: Sociological Methods & Research, Vol. 21, No. 1, 1992, p. 25-51.

Research output: Contribution to journalArticle

@article{830858dc474a4036a57f65fef40e2aba,
title = "Time-Aggregation Bias in Hazard-Rate Models with Covariates",
abstract = "This article elaborates the results found in Petersen (1991) by discussing how to minimize time-aggregation bias in hazard-rate models with measured covariates. It first considers a model with a single categorical covariate with h categories. It derives analytically the bias of the estimator that assumes the durations are exactly measured, when the durations are in fact rounded up to the nearest integer. Second, the article conducts both large-and small-sample Monte Carlo studies for several estimators of the covariate effects in the presence of time aggregation. There are three central findings. (1) It is shown that the likelihood that adjusts for the grouped nature of the duration measures recaptures the parameters very well. (2) The estimator that assumes that the durations are exactly measured, when they in fact are rounded up to the nearest integer, is biased in several ways. (3) The estimator that uses the midpoint adjustment suffers from the same weaknesses as the estimator that rounds up to the nearest integer, but to a lesser degree.",
author = "Trond Petersen and Koput, {Kenneth W}",
year = "1992",
doi = "10.1177/0049124192021001002",
language = "English (US)",
volume = "21",
pages = "25--51",
journal = "Sociological Methods and Research",
issn = "0049-1241",
publisher = "SAGE Publications Inc.",
number = "1",

}

TY - JOUR

T1 - Time-Aggregation Bias in Hazard-Rate Models with Covariates

AU - Petersen, Trond

AU - Koput, Kenneth W

PY - 1992

Y1 - 1992

N2 - This article elaborates the results found in Petersen (1991) by discussing how to minimize time-aggregation bias in hazard-rate models with measured covariates. It first considers a model with a single categorical covariate with h categories. It derives analytically the bias of the estimator that assumes the durations are exactly measured, when the durations are in fact rounded up to the nearest integer. Second, the article conducts both large-and small-sample Monte Carlo studies for several estimators of the covariate effects in the presence of time aggregation. There are three central findings. (1) It is shown that the likelihood that adjusts for the grouped nature of the duration measures recaptures the parameters very well. (2) The estimator that assumes that the durations are exactly measured, when they in fact are rounded up to the nearest integer, is biased in several ways. (3) The estimator that uses the midpoint adjustment suffers from the same weaknesses as the estimator that rounds up to the nearest integer, but to a lesser degree.

AB - This article elaborates the results found in Petersen (1991) by discussing how to minimize time-aggregation bias in hazard-rate models with measured covariates. It first considers a model with a single categorical covariate with h categories. It derives analytically the bias of the estimator that assumes the durations are exactly measured, when the durations are in fact rounded up to the nearest integer. Second, the article conducts both large-and small-sample Monte Carlo studies for several estimators of the covariate effects in the presence of time aggregation. There are three central findings. (1) It is shown that the likelihood that adjusts for the grouped nature of the duration measures recaptures the parameters very well. (2) The estimator that assumes that the durations are exactly measured, when they in fact are rounded up to the nearest integer, is biased in several ways. (3) The estimator that uses the midpoint adjustment suffers from the same weaknesses as the estimator that rounds up to the nearest integer, but to a lesser degree.

UR - http://www.scopus.com/inward/record.url?scp=84965509416&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84965509416&partnerID=8YFLogxK

U2 - 10.1177/0049124192021001002

DO - 10.1177/0049124192021001002

M3 - Article

VL - 21

SP - 25

EP - 51

JO - Sociological Methods and Research

JF - Sociological Methods and Research

SN - 0049-1241

IS - 1

ER -