Open AccessSpecifics of logit and probit regression in education sciences - why wouldn't we use it?
Author(s) -
Siniša Opić
Publication year - 2020
Publication title -
cypriot journal of educational sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.22
H-Index - 5
eISSN - 1305-905X
pISSN - 1305-9076
DOI - 10.18844/cjes.v15i6.5305
Subject(s) - probit , logit , probit model , ordered probit , logistic regression , econometrics , regression analysis , statistics , multinomial probit , ordered logit , linear regression , mathematics
Regression is one of the dominant analysis methods used in the social sciences and educational sciences. There are different regression methods based on the type of research that is being conducted. The probit and logit regression models are regression methods which are being used recently by most researchers. However, their interpretations are not straightfoward and most researchers end up misinterepreting the results from the probit and logit regression models. This research therefore aims to examine the differences between the probit and logit models, in comparison with other linear regression models. Using a comparative research design, this study utilises resources from previous researchers, hence, the study took a form of a literature review. The results of this study is essential to educational and social sciences researchers who make use of the probit, logit and other regression methods. The research also explains why logit and probit should be used in place of other regression models.
Keywords: education sciences; Linear regression; Logit; Probit; Regression