Dimensionality of indicators in questionnaire for diagnostics of hand and foot preference - adult population
Abstract
The aim of the study was to verify the structural hypothesis whether the items of the questionnaire survey, assessing motor manifestations of upper and lower limb laterality in the adult population, consist of two separate latent variables or not. The selection of the questionnaire items was based on an expert assessment of content validity of items that were part of already created standardized foreign questionnaires.
The diagnostic tool originally consisted of 10 items (of tool, skill and non-skill character). The research was carried out using 440 individuals aged 17-19 years.
In order to analyze the data, structural equation modelling (SEM) was used, specifically the confirmatory approach for categorical data called item factor analysis (IFA). The results showed that the most appropriate model is a one-factor model with seven items.
This model had a very high diagnostic quality, expressed by fit indices RMSEA 0.028, CFI 0.99, TLI 0.99, and signification of the model on the level of p - 0.29. However, the resultant one-factor model suggests that particularly answers to items of skill character in which strong factor loadings were found may be influenced by the subjective assessment of the degree of lateral preference, resulting in local dependence, i.e. the answer to a particular item is affected by the answer to the previous item.