No, sensitive data is never stored. Once you upload your sheet and select the columns to use, only the necessary data is imported. All other data, including unused columns, is automatically deleted to ensure your information remains secure.
The reports include insights and visualizations, but specific recommendations are left to the discretion of the client and/or organization.
The number of bins in the histogram is set using the larger value between two statistical rules: Sturges' Rule and the Freedman-Diaconis Rule. These rules help determine the best number of bins to display the data effectively.
Once the number of bins is decided, the score values are divided into equal ranges (or "chunks") across those bins. For example, if there are 9 bins, the data is split into 9 equal groups based on the minimum and maximum values. The numbers shown (e.g., 89) represent the starting point of a specific bin—meaning any score within that range falls into that category.
Processing time depends on the size and complexity of the dataset, but optimizations are in place to ensure efficiency. Generation time should take no more than 90 seconds.
There is no limit, however, the larger the dataset, the longer it will take for the report to generate.
The individual components do not include a pass/fail or Selected/Not Selected threshold because, in a compensatory approach, each component is not designed to stand alone with its own decision point. In this model, overall performance across all components determines selection outcomes. To evaluate Adverse Impact, please refer to the Overall Report. The individual component or assessment reports are useful for examining mean differences (e.g., through ANOVA analyses), but they are not intended for independent pass/fail decisions.
A referent group is the baseline group used for comparison when evaluating selection rates, particularly under the 80% rule for adverse impact.
While often called the "unprotected class" for simplicity, this is not technically accurate — all races, ethnicities, and genders are legally protected under U.S. law. The term simply reflects that the referent group is usually the one least likely to experience discrimination and is used as the benchmark for fairness evaluations.
The "Traditional Adverse Impact Selection Ratio" is selected by default because historically, most lawsuits have noted the assessment and selection procedures favored the traditional group (White, males, under 40). If you wish to modify this selection, you can adjust it by clicking on the report wizard.
Adverse impact and mean differences are reviewed at each stage to ensure compliance with standard selection procedures and to identify any potential biases early in the process, ensuring fairness and equity throughout.
No, it is used by HR teams, compliance officers, recruiters, and legal teams for workforce analytics and selection fairness evaluation.
Statistical significance shows whether a result is likely real or just due to chance. If the p-value is below 0.05, it's generally considered significant, meaning there's a low probability (less than 5%) that the outcome happened randomly. However, significance doesn’t always mean the result is meaningful or causal.
Industrial-Organizational (I/O) Psychology is the study of human behavior in the workplace. It focuses on improving employee performance, well-being, and organizational effectiveness through research and applied psychology.
"Total Evaluations" shows the total number of reports you have generated, including any reports or datasets you may have deleted. It keeps a running tally to give you a clear idea of your overall usage.