Our tests are developed by a team of experienced developers led by Ilona Jerabek, PhD. Our technical team, directed by artificial intelligence (AI) researcher Vrat Jerabek, continuously conducts experiments and simulations to improve the AI systems that handle complicated scoring algorithms.
All of the tests offered on ARCH Profile 2wardsustainability are rigorously developed and based on a thorough statistical analysis. Each test is well researched and designed according to the APA (American Psychological Association) standards for educational and psychological testing.
Overview of Statistical Analyses
In our validation studies, we run a number of statistical analyses that help us determine the reliability and validity of the assessments. Typically, we analyse the distributions of scores, the internal consistency of the scales (Cronbach's alphas), correlations between scales, and the performance of individual items. We also conduct factor analysis (both exploratory and confirmatory) to evaluate the structure of the assessment, and compare groups of subjects on a number of variables (performance ratings, educational level, position within organisation etc.) using t-tests and ANOVAs. In addition, our intelligence tests are subjected to Rasch analysis. Results of these analyses are reported in statistical reports available to our clients.
Compliance with APA Testing Standards
Extensive statistical analyses are a very important part of the scientific validation process at Psychtests. We gather data from tens of thousands of test-takers on Psychtests, Queendom and other partner sites. In addition, we investigate the predictive validity of our personality and career aptitude tests in partnership with several companies, including Hickory Farms and Human Perspective International. Based on our findings, we continue to update the tests to provide the most accurate results possible.