Statistics III
Full course description
The goal of this course is twofold. On the one hand, it supplements Statistics II; that is the analysis of two-way designs with a dichotomous instead of quantitative dependent variable. On the other hand, the emphasis lies on the analysis of tests and questionnaires. In this way, this course provides students a solid statistical preparation for the course ‘Psychodiagnostics’.
In this course, students will study three techniques spanning several weeks: logistic regression, reliability analysis and factor analysis.
Logistic regression is the equivalent of ANOVA and regression analysis covered in ‘Statistics II’ if the dependent variable is dichotomous instead of continuous, such as recovery from disease or passing an exam. Logistic regression allows us to adjust the effects of multiple independent variables for each other (confounding) and to study interactions. In this way, it also expands upon the contingency table analysis from ‘Statistics I’ to multiple independent variables.
Reliability analysis is a classical psychometric method for analyzing tests and questionnaires. Oftentimes, persons' answers to multiple-choice questions (items) are scored dichotomously and summed to give a total score for e.g. intelligence or attitude. In doing so, one assumes that these items measure the same thing. Reliability analysis can verify whether each item fits into the scale and how reliable the total score is. In the course students receive a training in classical psychometrics and an introduction into modern psychometrics (the Rasch model), validity, and agreement between evaluators.
Factor analysis is a method used to reduce a multitude of variables to a small number of underlying factors. In the past, factor analysis was used to reduce the scores on various tests to a small number of dimensions, such as verbal and spatial intelligence, or extraversion and neuroticism. Nowadays, factor analysis is more often used to group items of one questionnaire into sub-scales. Factor analysis is thus related to psychometrics. In the course students receive a training in exploratory factor analysis with SPSS.
The corresponding practical for this course is: SPSS III
The final assessment for this course is a numerical grade between 0,0 and 10,0.
Course objectives
Students are able:
- to explain relevant concepts central to this module, including confounding and interaction, classical psychometrics, reliability, modern psychometrics, item response theory, Rasch model, validity, agreement;
- to explain and apply specific statistical techniques, such as three-way contingency table analysis, logistic regression, reliability analysis (including item analysis) and exploratory factor analysis, and they can interpret relevant output of these techniques;
- to specify the assumptions of statistical techniques that were discussed in this module and are able to apply this knowledge when analyzing data.
Prerequisites
Admission requirement: On reference date March 15 of the relevant year Statistics I has to be completed.