TY  - JOUR
A1  - Schweizer, Karl
A1  - Gold, Andreas
A1  - Krampen, Dorothea
T1  - Effect of dichotomization on the latent structure of data
T2  - Frontiers in applied mathematics and statistics
N2  - We investigated whether dichotomous data showed the same latent structure as the interval-level data from which they originated. Given constancy of dimensionality and factor loadings reflecting the latent structure of data, the focus was on the variance of the latent variable of a confirmatory factor model. This variance was shown to summarize the information provided by the factor loadings. The results of a simulation study did not reveal exact correspondence of the variances of the latent variables derived from interval-level and dichotomous data but shrinkage. Since shrinkage occurred systematically, methods for recovering the original variance were fleshed out and evaluated.
KW  - dichotomous data
KW  - interval-level data
KW  - dichotomization
KW  - confirmatory factor analysis
KW  - shrinkage correction
KW  - latent structure
Y1  - 2021
UR  - http://publikationen.ub.uni-frankfurt.de/frontdoor/index/index/docId/63391
UR  - https://nbn-resolving.org/urn:nbn:de:hebis:30:3-633915
SN  - 2297-4687
VL  - 7
IS  - art. 723258
SP  - 1
EP  - 9
PB  - Frontiers Media
CY  - Lausanne
ER  -