README.md

In brunomontezano/HCLpsychometrics: Performs the Psychometric Analyzes of the HCL-32 Instrument in a Database of a Cohort Study in Southern Brazil

HCL-32 Psychometric Properties

The objective of HCLpsychometrics is to provide functions to perform the analyzes used in a study on the factor structure of the HCL-32 instrument in a population sample. The package has functions for executing routines for confirmatory factor analysis and Cronbach’s alpha coefficient estimation.

Note

This repository is not intended to be used by others. It is a package aggregating several tools used to replicate the specific analysis of an article. The dataset named hcl inside the package contains an example tibble (10% of the original sample) to check the proper functioning and give an idea on how the package works.

Installation

The current version can be installed from GitHub with:

# install.packages("remotes")
remotes::install_github("brunomontezano/HCLpsychometrics")

Examples

Confirmatory Factor Analysis

library(HCLpsychometrics)

# First, the models to be tested are created
created_models <- create_models()

# Then, we can fit these models on given dataset
# that contains yX as variable names, X being numbers from 1 to 32
fitted_models <- fit_models(
  data = hcl,
  models = created_models
)

# Finally, I could for example, summarize the parameters
# (In this case, I just printed the first 3 rows to save space)
summarize_parameters(
  fits = fitted_models
  ) %>% purrr::map(head, 3)
#> $Bech
#>   Factor Item     B    SE     Z p.value  Beta
#> 1 Active  y28 1.488 0.258 5.777       0 0.984
#> 2 Active   y4 1.318 0.235 5.616       0 0.872
#> 3   Risk  y23 1.000 0.000    NA      NA 0.851
#> 
#> $Forty
#>   Factor Item     B    SE     Z p.value  Beta
#> 1 Active  y28 2.143 0.556 3.852       0 0.947
#> 2 Active  y19 1.962 0.513 3.828       0 0.867
#> 3 Active  y17 1.928 0.496 3.890       0 0.852
#> 
#> $`HCL-28`
#>   Factor Item     B    SE     Z p.value  Beta
#> 1 Active  y28 1.564 0.289 5.415       0 0.985
#> 2 Active  y19 1.432 0.269 5.330       0 0.902
#> 3 Active   y3 1.416 0.269 5.261       0 0.891

As can be seen from the output of the functions, they work in order to fit three models for CFA: a model by Bech et al. (2011), another by Forty et al. (2010) and a third model called HCL-28, developed by the authors of the paper.

Note that the summarize_fit and summarize_parameters functions’ outputs are returned as an R list, facilitating the individual check of the results of each model through the elements in this list.

Cronbach’s Alpha

# The alpha_hcl28 function can be used to calculate Cronbach's
# alpha based on the HCL-28 model in the input dataset
HCLpsychometrics::alpha_hcl28(hcl)
#> $`HCL-28`
#> [1] 0.8821557
#> 
#> $`Active factor`
#> [1] 0.9055125
#> 
#> $`Risk-taking factor`
#> [1] 0.6307153

As you can see from the output, the alpha_hcl28 function generates Cronbach’s alpha for the structure (HCL-28) as a whole, and separately by factor.

Acknowledgement

I would like to thank designer Guilherme Bueno for creating the repository logo.

Logo icon adapted from Freepik.

Contact

Feel free to contact me here on GitHub or ResearchGate.

brunomontezano/HCLpsychometrics documentation built on Dec. 19, 2021, 11:50 a.m.