Predicting gender differences as latent variables

summed scores, and individual item responses: a methods case study.

Ricardo Pietrobon, Marcus Taylor, Ulrich Guller, Laurence D. Higgins, Danny O. Jacobs, Timothy Carey

Research output: Contribution to journalArticle

Abstract

Modeling latent variables such as physical disability is challenging since its measurement is performed through proxies. This poses significant methodological challenges. The objective of this article is to present three different methods to predict latent variables based on classical summed scores, individual item responses, and latent variable models. This is a review of the literature and data analysis using "layers of information". Data was collected from the North Carolina Back Pain Project, using a modified version of the Roland Questionnaire. The three models are compared in relation to their goals and underlying concepts, previous clinical applications, data requirements, statistical theory, and practical applications. Initial linear regression models demonstrated a difference in disability between genders of 1.32 points (95% CI 0.65, 2.00) on a scale from 0-23. Subsequent item analysis found contradictory results across items, with no clear pattern. Finally, IRT models demonstrated three items were demonstrated to present differential item functioning. After these items were removed, the difference between genders was reduced to 0.78 points (95% CI, -0.99, 1.23). These results were shown to be robust with re-sampling methods. Purported differences in the levels of a latent variable should be tested using different models to verify whether these differences are real or simply distorted by model assumptions.

Original languageEnglish (US)
Article number59
JournalHealth and Quality of Life Outcomes
Volume2
StatePublished - 2004
Externally publishedYes

Fingerprint

Linear Models
Proxy
Back Pain
Surveys and Questionnaires

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health

Cite this

Pietrobon, R., Taylor, M., Guller, U., Higgins, L. D., Jacobs, D. O., & Carey, T. (2004). Predicting gender differences as latent variables: summed scores, and individual item responses: a methods case study. Health and Quality of Life Outcomes, 2, [59].

Predicting gender differences as latent variables : summed scores, and individual item responses: a methods case study. / Pietrobon, Ricardo; Taylor, Marcus; Guller, Ulrich; Higgins, Laurence D.; Jacobs, Danny O.; Carey, Timothy.

In: Health and Quality of Life Outcomes, Vol. 2, 59, 2004.

Research output: Contribution to journalArticle

Pietrobon, Ricardo ; Taylor, Marcus ; Guller, Ulrich ; Higgins, Laurence D. ; Jacobs, Danny O. ; Carey, Timothy. / Predicting gender differences as latent variables : summed scores, and individual item responses: a methods case study. In: Health and Quality of Life Outcomes. 2004 ; Vol. 2.
@article{44231e5b379048c8953f7e1d11cdea55,
title = "Predicting gender differences as latent variables: summed scores, and individual item responses: a methods case study.",
abstract = "Modeling latent variables such as physical disability is challenging since its measurement is performed through proxies. This poses significant methodological challenges. The objective of this article is to present three different methods to predict latent variables based on classical summed scores, individual item responses, and latent variable models. This is a review of the literature and data analysis using {"}layers of information{"}. Data was collected from the North Carolina Back Pain Project, using a modified version of the Roland Questionnaire. The three models are compared in relation to their goals and underlying concepts, previous clinical applications, data requirements, statistical theory, and practical applications. Initial linear regression models demonstrated a difference in disability between genders of 1.32 points (95{\%} CI 0.65, 2.00) on a scale from 0-23. Subsequent item analysis found contradictory results across items, with no clear pattern. Finally, IRT models demonstrated three items were demonstrated to present differential item functioning. After these items were removed, the difference between genders was reduced to 0.78 points (95{\%} CI, -0.99, 1.23). These results were shown to be robust with re-sampling methods. Purported differences in the levels of a latent variable should be tested using different models to verify whether these differences are real or simply distorted by model assumptions.",
author = "Ricardo Pietrobon and Marcus Taylor and Ulrich Guller and Higgins, {Laurence D.} and Jacobs, {Danny O.} and Timothy Carey",
year = "2004",
language = "English (US)",
volume = "2",
journal = "Health and Quality of Life Outcomes",
issn = "1477-7525",
publisher = "BioMed Central",

}

TY - JOUR

T1 - Predicting gender differences as latent variables

T2 - summed scores, and individual item responses: a methods case study.

AU - Pietrobon, Ricardo

AU - Taylor, Marcus

AU - Guller, Ulrich

AU - Higgins, Laurence D.

AU - Jacobs, Danny O.

AU - Carey, Timothy

PY - 2004

Y1 - 2004

N2 - Modeling latent variables such as physical disability is challenging since its measurement is performed through proxies. This poses significant methodological challenges. The objective of this article is to present three different methods to predict latent variables based on classical summed scores, individual item responses, and latent variable models. This is a review of the literature and data analysis using "layers of information". Data was collected from the North Carolina Back Pain Project, using a modified version of the Roland Questionnaire. The three models are compared in relation to their goals and underlying concepts, previous clinical applications, data requirements, statistical theory, and practical applications. Initial linear regression models demonstrated a difference in disability between genders of 1.32 points (95% CI 0.65, 2.00) on a scale from 0-23. Subsequent item analysis found contradictory results across items, with no clear pattern. Finally, IRT models demonstrated three items were demonstrated to present differential item functioning. After these items were removed, the difference between genders was reduced to 0.78 points (95% CI, -0.99, 1.23). These results were shown to be robust with re-sampling methods. Purported differences in the levels of a latent variable should be tested using different models to verify whether these differences are real or simply distorted by model assumptions.

AB - Modeling latent variables such as physical disability is challenging since its measurement is performed through proxies. This poses significant methodological challenges. The objective of this article is to present three different methods to predict latent variables based on classical summed scores, individual item responses, and latent variable models. This is a review of the literature and data analysis using "layers of information". Data was collected from the North Carolina Back Pain Project, using a modified version of the Roland Questionnaire. The three models are compared in relation to their goals and underlying concepts, previous clinical applications, data requirements, statistical theory, and practical applications. Initial linear regression models demonstrated a difference in disability between genders of 1.32 points (95% CI 0.65, 2.00) on a scale from 0-23. Subsequent item analysis found contradictory results across items, with no clear pattern. Finally, IRT models demonstrated three items were demonstrated to present differential item functioning. After these items were removed, the difference between genders was reduced to 0.78 points (95% CI, -0.99, 1.23). These results were shown to be robust with re-sampling methods. Purported differences in the levels of a latent variable should be tested using different models to verify whether these differences are real or simply distorted by model assumptions.

UR - http://www.scopus.com/inward/record.url?scp=84872239313&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872239313&partnerID=8YFLogxK

M3 - Article

VL - 2

JO - Health and Quality of Life Outcomes

JF - Health and Quality of Life Outcomes

SN - 1477-7525

M1 - 59

ER -