Exploratory factor analysis attempts to identify the underlying factors that explain the pattern of correlations within a set of observed variables. The analysis is almost always performed with Pearson’s correlations even when the data are ordinal, but this is not appropriate since they are not quantitative data. The use of Likert scales is increasingly common in the field of social research, so it is necessary to determine which methodology is the most suitable for analysing the data obtained as non quantitative measures. In this context, also by means of simulation studies, we aim to illustrate the advantages of using Spearman’s grade correlation coefficient on a transformation operated by the copula function in order to perform exploratory factor analysis of ordinal variables. Moreover, by using the copula, we consider the general dependence structure, providing a more robust reproduction of the measurement model.
L’analisi fattoriale esplorativa vuole identificare i fattori latenti che spiegano un insieme di variabili osservate. L’analisi quasi sempre utilizza la correlazione di Pearson, anche quando i dati sono di natura ordinale, ma questo non e appropriato in quanto questi dati non sono quantitativi. L’uso di scale Likert ´ e´ sempre piu comune nel campo della ricerca sociale, risulta quindi necessario de- ´ terminare quale metodo risulta essere piu idoneo per l’analisi di tali dati tenendo ´ presente che spesso vengono analizzati utilizzando tecniche idonee solo per misure quantitative. In questo contesto, e mediante studi di simulazione, si illustrano i vantaggi nell’utilizzo dello Spearman grade correltion ottenuto mediante l’utilizzo dalla funzione copula anziche della correlazione di Pearson. Con l’utilizzo della ´ copula, si considera cos´ı la struttra di dipendenza generale, fornendo cos´ı una misurazione piu accurata
Exploratory factor analysis of ordinal variables: a copula approach = Analisi fattoriale esplorativa di variabili ordinali: un approccio via copula
Nai Ruscone, Marta
2017-01-01
Abstract
Exploratory factor analysis attempts to identify the underlying factors that explain the pattern of correlations within a set of observed variables. The analysis is almost always performed with Pearson’s correlations even when the data are ordinal, but this is not appropriate since they are not quantitative data. The use of Likert scales is increasingly common in the field of social research, so it is necessary to determine which methodology is the most suitable for analysing the data obtained as non quantitative measures. In this context, also by means of simulation studies, we aim to illustrate the advantages of using Spearman’s grade correlation coefficient on a transformation operated by the copula function in order to perform exploratory factor analysis of ordinal variables. Moreover, by using the copula, we consider the general dependence structure, providing a more robust reproduction of the measurement model.File | Dimensione | Formato | |
---|---|---|---|
5128.pdf
accesso aperto
Tipologia:
Documento in versione editoriale
Dimensione
465.39 kB
Formato
Adobe PDF
|
465.39 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.