A longitudinal behavior genetic model for ordered categorical variables

A model for the analysis of longitudinal twin data consisting of ordered categorical variables was proposed. Proportional changes in response categories across time were modeled within growth curve by means of mean and variance changes in underlying continuous variables. Variances of the growth factors were decomposed into genetic and environmental components. In simulation analyses, parameters were successfully estimated in all conditions although estimates of standard errors were biased and statistical powers to detect non-zero parameters were not sufficient for some parameters. A potential solution for these irregularities was discussed. Despite these limitations, the relative contributions of genetic and environmental components on growth factors were well estimated. Possible refinement of simulation analysis and expansion of the proposed model for multiple indicators at each measurement occasion were discussed.

Name thesis:degree_name

Ph. D.

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Psychological sciences (MU)

Grantor dc:publisher

University of Missouri--Columbia

Year dc:date.issued

2011

Author dc:creator

• Cho, Seung Bin

Advisor dc:contributor.advisor

• Wood, Phillip K. (Phillip Karl)

• behavior genetics
• structual equation
• growth curve model
• categorical variables

Statement dc:rights

• OpenAccess.

Language dc:language.iso

eng, English

DOI

https://doi.org/10.32469/10355/14198

OAI identifier oai:identifier

oai:mospace.umsystem.edu:10355/14198