The PI Behavioral Assessment was designed using a normative sample of 8,000 employees across several industries. While it was designed as a normative test, the synthesis of the results is translated into a “within-person” representation. This unique approach determines the native drives and needs that are most critical for someone to function at their highest work- performance levels.
Recent meta-analytic evidence shows examples where the within-person approach (e.g., Factor combinations) allows for better prediction of job performance because it allows for enhanced analysis and interpretation of factor combinations. For example, if an individual’s Factor A (dominance) is higher than their Factor D (formality), this will better predict sales performance than only considering Factor A or Factor D alone. This combination analysis is otherwise not available in typical normative tests.
The Predictive Index predicts four primary personality constructs (Factors A, B, C, and D) and secondary constructs (Factors E — Decision Making and M — Response Level). These constructs are “normal” (non-clinical) characteristics that describe, explain and predict day-to-day workplace behaviors.
The Predictive Index behavioral assessment provides two patterns of an individual’s workplace behavior:
The Self pattern is a reflection of an individual’s native drives and needs and how they will naturally react in an environment or in situations.
The Self-Concept pattern reflects how an individual may be changing their native behaviors to adapt to the expectations of their work environment.
Comparing the Self to the Self-Concept pattern can reveal tension or synergy between an individual and their current workplace. This can be used to identify the root of observed behaviors and aid in managing individuals through the hire-to-retire lifecycle.
Test-retest reliability has been evaluated in multiple studies with results showing strong stability for the Predicitive Index in intervals of two weeks to four years and adequate reliability from retest intervals of five to eight years as would be expected. Internal consistency reliability has been evaluated recently in three studies with an average internal consistency reliability of .85 and a range of .82 to .87. These estimates far exceed the lower boundary of acceptability which is .70.
Multiple construct validation studies have been conducted through the years comparing the Predictive Index Factors to relevant scales in well-established normal personality assessments like the 16PF and NEO PI-R. These studies have shown the strong convergent validity necessary to claim that the Predicitve Index is construct valid. For example, the correlation between Dominance Factor and the 16PF’s Independence factor was .47 (p .01) and the correlation between the Predictive Index Extroversion Factor and the NEO PI-R Extroversion scale is .63.
The Predictive Index has been investigated in nearly 500 criterion-related validity studies since September of 1976 across almost all jobs and countries. This body of evidence supports the fact that the Predictive Index is indeed consistently related to important workplace outcomes such as tenure, turnover, sales and customer satisfaction. Two recent meta-analyses show that the PI predicts overall job- performance, tenure, sales performance and counterproductive behavior.
The Predictive Index is bias free. In every study in the last two decades, it has been shown that Predictive Index scores do not differ on the basis of age, gender or ethnicity. In addition, there is no evidence to indicate that the inclusion of the Predictive Index in a company’s personnel selection system – either in a compensatory or “multiple-hurdle” selection model – results in adverse impact against any protected class when examining adverse impact via the four-fifths rule, Adverse Impact (AI) Rule or the “two Standard Deviations” rule.
The Predictive Index is a brief (6.5 minutes on average), untimed, free choice (as opposed to forced choice) assessment that presents individuals with a list of adjectives and asks them to identify which items fit them and their current role. The Predictive Index is available in multiple formats (paper, online, mobile) and is available in more than 75 languages, including English Braille.
Constant improvement is the cornerstone of scientific research and development at PI. Since 1955, The Predictive Index Behavioral Assessment has undergone several iterations. PI also maintains a Science Advisory Board staffed with university professors, I/O psychologists, and other subject matter experts in psychometrics. These advisors review our science methodology and provide suggestions on how we can continue to build on the historic foundation of the assessments we provide.