Author/Authors
selvi, hüseyin mersin üniversitesi - tıp fakültesi - tıp eğitimi ad, turkey , derici yildirim, didem mersin üniversitesi - tıp fakültesi - biyoistatistik ve tıbbi bilişim ad, turkey
Title Of Article
INVESTIGATION OF LATENT CLASS ANALYSIS IN DIFFERENT SCORING SITUATIONS
شماره ركورد
45845
Abstract
The aim of this study is to examine the Latent Class Analysis in different scoring situations. For this purpose, a test consists of multiple choice items which can be scored in different types such as dichotomous, expert judgement and weighted experimentally, respectively. A total of 595 students who are studying at Mersin University in the Faculty of Education were included in the study. Latent Class Analysis was used to determine the best fitted model of data. The findings of the study showed that the same number of classes was calculated in latent class analysis for dichotomous and expert judgment scoring. The minimum number of classes was obtained for experimentaly scoring with the least parameter estimation. For dichotomous and expert judgment scoring, the number of classes was higher than experimental scoring. In the latent class analysis, the most appropriate model is the best fitted model with the least number of latent classes and least parameter estimation. Therefore, it can be stated that the experimental scoring method model is the most appropriate model for latent class analysis.
From Page
754
NaturalLanguageKeyword
Latent class analysis , dichotomous scoring , expert judgment scoring , experimentally weighted scoring
JournalTitle
Abant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi
To Page
766
JournalTitle
Abant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi
Link To Document