|
Title
|
|
|
|
Methodology Latent Profile Analysis (LPA) used in the paper "Exploring learner profiles among low-educated adults in second-chance education: individual differences in quantity and quality of learning motivation and learning strategies"
| |
|
Author
|
|
|
|
| |
|
Abstract
|
|
|
|
| The folder 'Methodology' is an R-project, containing all materials underlying the methodology of the paper "Exploring learner profiles among low-educated adults in second-chance education: individual differences in quantity and quality of learning motivation and learning strategies" (10.1007/s10212-024-00834-5). A Latent Profile Analysis (LPA) was carried out in R, using Saaty's (1990) Analytic Hierarchy Process (AHP), in which different model fit criteria (AIC, AWE, BIC, CLC, and KIC) are mathematically combined into a composite relative importance vector (C-RIV). The C-RIV with the highest value represents the model with the most optimal number of profiles (Akogul & Erisoglu, 2017). Following the suggestions of Spurk et al. (2020), we checked three crucial assumptions for LPA: - inspection of missing data - inspection of outliers - inspection of normality of the distributions Content of the folder: - Data - Quarto-file containing a walkthrough and R-code underlying all results presented in the paper - Additional functions used in the walkthrough - Additional Excel-sheets (manual calculation C-RIV, presentation of local and global maximum estimations of log likelihood parameter (use of different starting values)) - Methodological literature supporting the methodological approach used. |
| |
|
Language
|
|
|
|
English
| |
|
Related publication(s)
|
|
|
|
| |
|
Publication
|
|
|
|
2023
| |
|
DOI
|
|
|
|
10.17605/OSF.IO/PCGVA
| |
|
Volume/pages
|
|
|
|
| |
|
Full text (Publisher's DOI)
|
|
|
|
| |
|