Skip to content

README.md

Latest commit

 

History

History
executable file
·
68 lines (49 loc) · 2.73 KB

README.md

File metadata and controls

executable file
·
68 lines (49 loc) · 2.73 KB

Big Five Personality Test: Item & Scale Analysis

Psychometric analysis of the 50-item IPIP Big Five Factor Markers questionnaire in R, including item statistics, factor analysis, and reliability estimation.

Item Correlation Network Factor Loadings

Scale Intercorrelations Item Difficulty vs. Discrimination Reliability

Data

This project uses the Big Five Personality Test dataset from Kaggle (~1M responses to the 50-item IPIP Big Five questionnaire).

Setup

  1. Install R (>= 4.0) from r-project.org
  2. Download the data from Kaggle and place data-final.csv into the src_data/ folder
  3. Install packages by running setup.R: 
    source("setup.R")

Alternatively, download the data via the Kaggle CLI:

kaggle datasets download -d tunguz/big-five-personality-test -p src_data/ --unzip

Scripts

The scripts are meant to be run in order. Each script sources its dependencies automatically.

Script Content
01_data_preparation.R Load data, define item metadata, recode negatively keyed items, compute scale scores
02_item_analysis.R Item descriptives, item histograms, corrected item-total correlations, difficulty-discrimination plots
03_scale_analysis.R Dimension descriptives & histograms, scale intercorrelations, exploratory factor analysis (heatmap, network), parallel analysis, reliability

Big Five Dimensions

Code Dimension Items
EXT Extraversion EXT1 -- EXT10
EST Emotional Stability EST1 -- EST10
AGR Agreeableness AGR1 -- AGR10
CSN Conscientiousness CSN1 -- CSN10
OPN Openness OPN1 -- OPN10

Negatively keyed items are recoded in 01_data_preparation.R.

Required R Packages

tidyr, dplyr, moments, ggplot2, ggrepel, scales, Hmisc, psych, nFactors, qgraph

Install all at once via setup.R or manually:

install.packages(c("tidyr", "dplyr", "moments", "ggplot2", "ggrepel",
                    "scales", "Hmisc", "psych", "nFactors", "qgraph"))

License

The IPIP items are in the public domain (ipip.ori.org). The Kaggle dataset was collected by Open Psychometrics.