Bias in a binary risk behaviour model subject to inconsistent reports and dropout in a South African high school cohort study

Perpetual Chikobvu, Carl J. Lombard, Alan J. Flisher, Gary King, Loraine Townsend, Martie Muller

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

We describe a methodology for analysing self-reported risk behaviour transitional patterns in a binary outcome variable, subject to misclassification and a large loss to follow-up. The motivation stems from the analysis of self-reported transitional patterns in responses to the question 'have you ever smoked a whole cigarette?' in a cohort of South African school children. The partially complete records analysis (PCRA) introduced, estimates the transitional probability as: the ratio of the joint probability of the response at two time points based on the complete records for this time sequence over the marginal probabilities of the response based on the complete records at the first time point, and assumes a non-informative missing pattern. A comparison was made using un-weighted complete records and inverse probability weighted logistic regression. The estimates of the probabilities of reporting ever having smoked a cigarette obtained from the three methods were similar for a particular transition. The PCRA method lacked precision compared with the inverse probability weighted logistic regression. A simulation study indicated an association between bias and reporting error in all three methods. The PCRA method can be considered as a method for the estimation of transition probabilities in a cohort study where there is consistency in the self-reported risk behaviour pattern and the sample size is large at baseline. The inverse probability weighting approach is more precise and is suitable for this setting in order to determine risk factors for the incidence of self-reported substance used in a cohort with a high dropout rate.

Original languageEnglish (US)
Pages (from-to)494-509
Number of pages16
JournalStatistics in Medicine
Volume28
Issue number3
DOIs
StatePublished - Feb 10 2009

Fingerprint

Cohort Study
Drop out
Risk-Taking
Inconsistent
Cohort Studies
Binary
Logistic Regression
Inverse Probability Weighting
Tobacco Products
Model
Binary Outcomes
Misclassification
Logistic Models
Risk Factors
Transition Probability
Estimate
Africa
Baseline
Incidence
Sample Size

All Science Journal Classification (ASJC) codes

  • Epidemiology
  • Statistics and Probability

Cite this

Chikobvu, Perpetual ; Lombard, Carl J. ; Flisher, Alan J. ; King, Gary ; Townsend, Loraine ; Muller, Martie. / Bias in a binary risk behaviour model subject to inconsistent reports and dropout in a South African high school cohort study. In: Statistics in Medicine. 2009 ; Vol. 28, No. 3. pp. 494-509.
@article{6255ead4061547bbadf61e5963914b6e,
title = "Bias in a binary risk behaviour model subject to inconsistent reports and dropout in a South African high school cohort study",
abstract = "We describe a methodology for analysing self-reported risk behaviour transitional patterns in a binary outcome variable, subject to misclassification and a large loss to follow-up. The motivation stems from the analysis of self-reported transitional patterns in responses to the question 'have you ever smoked a whole cigarette?' in a cohort of South African school children. The partially complete records analysis (PCRA) introduced, estimates the transitional probability as: the ratio of the joint probability of the response at two time points based on the complete records for this time sequence over the marginal probabilities of the response based on the complete records at the first time point, and assumes a non-informative missing pattern. A comparison was made using un-weighted complete records and inverse probability weighted logistic regression. The estimates of the probabilities of reporting ever having smoked a cigarette obtained from the three methods were similar for a particular transition. The PCRA method lacked precision compared with the inverse probability weighted logistic regression. A simulation study indicated an association between bias and reporting error in all three methods. The PCRA method can be considered as a method for the estimation of transition probabilities in a cohort study where there is consistency in the self-reported risk behaviour pattern and the sample size is large at baseline. The inverse probability weighting approach is more precise and is suitable for this setting in order to determine risk factors for the incidence of self-reported substance used in a cohort with a high dropout rate.",
author = "Perpetual Chikobvu and Lombard, {Carl J.} and Flisher, {Alan J.} and Gary King and Loraine Townsend and Martie Muller",
year = "2009",
month = "2",
day = "10",
doi = "10.1002/sim.3482",
language = "English (US)",
volume = "28",
pages = "494--509",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "3",

}

Bias in a binary risk behaviour model subject to inconsistent reports and dropout in a South African high school cohort study. / Chikobvu, Perpetual; Lombard, Carl J.; Flisher, Alan J.; King, Gary; Townsend, Loraine; Muller, Martie.

In: Statistics in Medicine, Vol. 28, No. 3, 10.02.2009, p. 494-509.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Bias in a binary risk behaviour model subject to inconsistent reports and dropout in a South African high school cohort study

AU - Chikobvu, Perpetual

AU - Lombard, Carl J.

AU - Flisher, Alan J.

AU - King, Gary

AU - Townsend, Loraine

AU - Muller, Martie

PY - 2009/2/10

Y1 - 2009/2/10

N2 - We describe a methodology for analysing self-reported risk behaviour transitional patterns in a binary outcome variable, subject to misclassification and a large loss to follow-up. The motivation stems from the analysis of self-reported transitional patterns in responses to the question 'have you ever smoked a whole cigarette?' in a cohort of South African school children. The partially complete records analysis (PCRA) introduced, estimates the transitional probability as: the ratio of the joint probability of the response at two time points based on the complete records for this time sequence over the marginal probabilities of the response based on the complete records at the first time point, and assumes a non-informative missing pattern. A comparison was made using un-weighted complete records and inverse probability weighted logistic regression. The estimates of the probabilities of reporting ever having smoked a cigarette obtained from the three methods were similar for a particular transition. The PCRA method lacked precision compared with the inverse probability weighted logistic regression. A simulation study indicated an association between bias and reporting error in all three methods. The PCRA method can be considered as a method for the estimation of transition probabilities in a cohort study where there is consistency in the self-reported risk behaviour pattern and the sample size is large at baseline. The inverse probability weighting approach is more precise and is suitable for this setting in order to determine risk factors for the incidence of self-reported substance used in a cohort with a high dropout rate.

AB - We describe a methodology for analysing self-reported risk behaviour transitional patterns in a binary outcome variable, subject to misclassification and a large loss to follow-up. The motivation stems from the analysis of self-reported transitional patterns in responses to the question 'have you ever smoked a whole cigarette?' in a cohort of South African school children. The partially complete records analysis (PCRA) introduced, estimates the transitional probability as: the ratio of the joint probability of the response at two time points based on the complete records for this time sequence over the marginal probabilities of the response based on the complete records at the first time point, and assumes a non-informative missing pattern. A comparison was made using un-weighted complete records and inverse probability weighted logistic regression. The estimates of the probabilities of reporting ever having smoked a cigarette obtained from the three methods were similar for a particular transition. The PCRA method lacked precision compared with the inverse probability weighted logistic regression. A simulation study indicated an association between bias and reporting error in all three methods. The PCRA method can be considered as a method for the estimation of transition probabilities in a cohort study where there is consistency in the self-reported risk behaviour pattern and the sample size is large at baseline. The inverse probability weighting approach is more precise and is suitable for this setting in order to determine risk factors for the incidence of self-reported substance used in a cohort with a high dropout rate.

UR - http://www.scopus.com/inward/record.url?scp=63049099100&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=63049099100&partnerID=8YFLogxK

U2 - 10.1002/sim.3482

DO - 10.1002/sim.3482

M3 - Article

VL - 28

SP - 494

EP - 509

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 3

ER -