Fitting the fractional polynomial model to non-gaussian longitudinal data

Ji Hoon Ryoo, Jeffrey D. Long, Greg W Welch, Arthur Reynolds, Susan S Swearer

Research output: Contribution to journalArticle

Abstract

As in cross sectional studies, longitudinal studies involve non-Gaussian data such as binomial, Poisson, gamma, and inverse-Gaussian distributions, and multivariate exponential families. A number of statistical tools have thus been developed to deal with non-Gaussian longitudinal data, including analytic techniques to estimate parameters in both fixed and random effects models. However, as yet growth modeling with non-Gaussian data is somewhat limited when considering the transformed expectation of the response via a linear predictor as a functional form of explanatory variables. In this study, we introduce a fractional polynomial model (FPM) that can be applied to model non-linear growth with non-Gaussian longitudinal data and demonstrate its use by fitting two empirical binary and count data models. The results clearly show the efficiency and flexibility of the FPM for such applications.

Original languageEnglish (US)
Article number1431
JournalFrontiers in Psychology
Volume8
Issue numberAUG
DOIs
StatePublished - Aug 22 2017

Fingerprint

Statistical Models
Nonlinear Dynamics
Normal Distribution
Growth
Longitudinal Studies
Cross-Sectional Studies

Keywords

  • Chicago longitudinal study
  • Fractional polynomial
  • Generalized additive model
  • Non-Gaussian longitudinal data
  • Reading of the mind

ASJC Scopus subject areas

  • Psychology(all)

Cite this

Fitting the fractional polynomial model to non-gaussian longitudinal data. / Ryoo, Ji Hoon; Long, Jeffrey D.; Welch, Greg W; Reynolds, Arthur; Swearer, Susan S.

In: Frontiers in Psychology, Vol. 8, No. AUG, 1431, 22.08.2017.

Research output: Contribution to journalArticle

@article{5adb16f778164571b456e6f3e4747a2a,
title = "Fitting the fractional polynomial model to non-gaussian longitudinal data",
abstract = "As in cross sectional studies, longitudinal studies involve non-Gaussian data such as binomial, Poisson, gamma, and inverse-Gaussian distributions, and multivariate exponential families. A number of statistical tools have thus been developed to deal with non-Gaussian longitudinal data, including analytic techniques to estimate parameters in both fixed and random effects models. However, as yet growth modeling with non-Gaussian data is somewhat limited when considering the transformed expectation of the response via a linear predictor as a functional form of explanatory variables. In this study, we introduce a fractional polynomial model (FPM) that can be applied to model non-linear growth with non-Gaussian longitudinal data and demonstrate its use by fitting two empirical binary and count data models. The results clearly show the efficiency and flexibility of the FPM for such applications.",
keywords = "Chicago longitudinal study, Fractional polynomial, Generalized additive model, Non-Gaussian longitudinal data, Reading of the mind",
author = "Ryoo, {Ji Hoon} and Long, {Jeffrey D.} and Welch, {Greg W} and Arthur Reynolds and Swearer, {Susan S}",
year = "2017",
month = "8",
day = "22",
doi = "10.3389/fpsyg.2017.01431",
language = "English (US)",
volume = "8",
journal = "Frontiers in Psychology",
issn = "1664-1078",
publisher = "Frontiers Research Foundation",
number = "AUG",

}

TY - JOUR

T1 - Fitting the fractional polynomial model to non-gaussian longitudinal data

AU - Ryoo, Ji Hoon

AU - Long, Jeffrey D.

AU - Welch, Greg W

AU - Reynolds, Arthur

AU - Swearer, Susan S

PY - 2017/8/22

Y1 - 2017/8/22

N2 - As in cross sectional studies, longitudinal studies involve non-Gaussian data such as binomial, Poisson, gamma, and inverse-Gaussian distributions, and multivariate exponential families. A number of statistical tools have thus been developed to deal with non-Gaussian longitudinal data, including analytic techniques to estimate parameters in both fixed and random effects models. However, as yet growth modeling with non-Gaussian data is somewhat limited when considering the transformed expectation of the response via a linear predictor as a functional form of explanatory variables. In this study, we introduce a fractional polynomial model (FPM) that can be applied to model non-linear growth with non-Gaussian longitudinal data and demonstrate its use by fitting two empirical binary and count data models. The results clearly show the efficiency and flexibility of the FPM for such applications.

AB - As in cross sectional studies, longitudinal studies involve non-Gaussian data such as binomial, Poisson, gamma, and inverse-Gaussian distributions, and multivariate exponential families. A number of statistical tools have thus been developed to deal with non-Gaussian longitudinal data, including analytic techniques to estimate parameters in both fixed and random effects models. However, as yet growth modeling with non-Gaussian data is somewhat limited when considering the transformed expectation of the response via a linear predictor as a functional form of explanatory variables. In this study, we introduce a fractional polynomial model (FPM) that can be applied to model non-linear growth with non-Gaussian longitudinal data and demonstrate its use by fitting two empirical binary and count data models. The results clearly show the efficiency and flexibility of the FPM for such applications.

KW - Chicago longitudinal study

KW - Fractional polynomial

KW - Generalized additive model

KW - Non-Gaussian longitudinal data

KW - Reading of the mind

UR - http://www.scopus.com/inward/record.url?scp=85028070474&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028070474&partnerID=8YFLogxK

U2 - 10.3389/fpsyg.2017.01431

DO - 10.3389/fpsyg.2017.01431

M3 - Article

C2 - 28878723

AN - SCOPUS:85028070474

VL - 8

JO - Frontiers in Psychology

JF - Frontiers in Psychology

SN - 1664-1078

IS - AUG

M1 - 1431

ER -