I am planing to do the validation study of SGRQ based on Factor analysis. SGRQ includes items with both of polytomous and dichotomous options with continuous variable. Eg : likert scale, yes or no questions ? How to perform the item selection?
Thank you
You can use software programs (e.g., Mplus) that can properly fit factor analysis models (including EFA) to binary and ordinal items based on tetrachoric/polychoric correlations using appropriate estimation methods for categorical data.
I suggest to convert each item with different number of response categories to follow normal distribution in the score range 1 to 100 and then perform your statistical analysis invcluding EFA.
For the method of of transforming item scores please see my paper "Improved Scoring of Satisfaction with Life Scale", RevistaMultidisciplinar.com [RMd], April, 2023, 5 (2),105 – 126 https://doi.org/10.23882/rmd.23142
Choose a statistical software package that supports EFA with mixed data types. For example, you can use SPSS, R, or Mplus.
Factor analysis is a method to analyze the variance between several dependent variables based on their description in terms of a small number of latent variables (factors). In other words, factor analysis aims to simplify complex data by describing it in terms of a smaller number of variables.
In factor analysis, it is among the multivariate methods in which all variables are interdependent and the concept of modeling based on independent and dependent variables does not apply to it. In other words, in factor analysis, we seek to answer the following question: Is it possible to reduce the dimension of primary variables to a smaller set of variables, with the least amount of information loss?
Assign numerals for you dichotomous (ex. 1-Yes, 0-No) and for your polytomous (ex. Civil Status: 1- single, 2-Married, 3-separated, etc..).
You may use JAMOVI, software. It contains factors analysis - EFA and CFA. the output is already in APA format. It is a free software and does not expire.