DynaMetric Assessments
Web Application
This description of the PDP ProScan Survey includes excerpts of relevant research regarding its reliability, validity, and nondiscriminatory nature as well as information about how PDP complies with EEO regulations and legal requirements. We encourage you to review PDP Global’s full Research Reference 5.0 (2024) for more information.
Developed in 1976, the ProScan Survey instrument is composed of 60 highly researched self-descriptive adjectives for the purpose of measuring important behavioral traits in people from three perspectives: the true person (Basic/Natural Self), the pressures the person is experiencing (Priority Environments), and the projected persona (Predictor/Outward Self). The five traits that are measured are Dominance, Extroversion, Pace, Conformity, and Logic.
Individuals with high dominance are described as direct, decisive, innovative, and competitive. They get results and exert or challenge authority. Those with low dominance are supportive, moderate, and at peace with others.
Individuals with high extroversion are described as enthusiastic, persuasive, and influential; they seek opportunities, build teams, and delegate technical tasks. Those with low extroversion are contemplative, private, and unpretentious.
Individuals with high pace are described as steady, consistent, persistent, dependable, and cautious about change. Those with low pace are spontaneous, versatile, and action-oriented with a sense of urgency.
Individuals with high conformity value structure, rules, accuracy, and loyalty. They will follow and maintain established systems and procedures perceived to be right. Those with low conformity want independence, personal freedom, and minimal external controls.
Individuals with high logic are analytical and process decisions using facts. Those with low logic are intuitive and process decisions based on life experiences.
(For a more detailed discussion of these traits, see Research Reference 5.0, pages 2.9–10.)
Two factors determine the strength of an instrument—reliability, and validity.
Reliability refers to how dependably or consistently an instrument measures a characteristic. A common way to measure reliability is Cronbach’s alpha (α), which generates a reliability coefficient score that ranges from 0 to 1. The closer the score is to 1, the more reliable the instrument is (i.e., the more likely it is to yield the same results each time it is repeated). A score of .70 or above is acceptable. In initial research, PDP’s ProScan Survey received an excellent reliability score of α = .94.
Validity determines if an assessment measures what it claims to measure. Initial research of the ProScan Survey showed strong evidence of validity, with coefficients of concurrent validity ranging from .39 to .87, demonstrating moderate to very strong correlations. (For more on reliability and validity, see Research Reference 5.0, page 1.9.)
In 1977 and 1978, higher-level analytic techniques were used to develop the ProScan instrument on the foundation of four primary behavioral traits and a single secondary trait. In this process, a large sample (n = 1,024) of people was used to standardize the intensity of the traits within individuals as well as across the population. For example, an individual’s strongest trait could be determined in relation to not only his or her own overall results but also in relation to the population’s results. This study gave PDP confidence in the instrument, based on the excellent reliability score of α = .94 and strong evidence of validity, with coefficients of concurrent validity ranging from .39 to .87, demonstrating moderate to very strong correlations. An additional study in 1988 reaffirmed the strong validity of the ProScan and confirmed the validity of the JobScan—PDP’s module for attracting and hiring the best hire. These consistent results gave PDP confidence that both instruments were also capable of serving as predictors of job success.
As part of our continual improvement efforts, PDP Global completed a long-term study in July 2021 of 39,049 unique respondents to determine if even greater accuracy could be achieved with the ProScan Survey, with the reality of language evolution and change over time. Exploratory factor analysis (EFA) was used to test the effectiveness of existing as well as possible replacement survey words. This complex and rigorous process is ideal for determining whether scores from an instrument are valid. Eleven EFAs were conducted on sample sizes (n) ranging from 1,200 to 3,000. The words that performed the best were then tested in a confirmatory factor analysis (CFA). CFA is a large-sample procedure that is even more rigorous than EFA.
The purpose of the CFA study was to confirm that the proposed replacement words had superior psychometric properties to the words used in the current ProScan Survey. All three CFAs were performed on large samples of unique respondents (n = 1,579, 1,623, and 3,749 respectively). Reliability, using Cronbach’s alpha, was considered excellent for all three samples, with all coefficients far above the acceptable threshold of α =.70 (CFA 1 α = .88, CFA 2 α = .87, and CFA 3 α = .88). Validity revealed an excellent fit for each individual trait and for a 4-factor (Dominance, Extroversion, Pace, and Conformity) and 1-factor (Logic). The second CFA (CFA 2) was performed for replication purposes. To achieve an excellent fit, two slight modifications were made to allow two words with weak crossloading to contribute to two different traits. The third CFA with a sample size more than double the first CFA confirmed reproducibility and provided additional evidence that the replacement model was reproducible in a larger general population.
The long-term study provided an updated version of the ProScan Survey 500 Series that yields reliable and valid scores that are, most importantly, also reproducible. (For more, see Psychometric and Statistical Evolution of the ProScan Survey and ProScan Confirmatory Factor Analysis, or Research Reference 5.0, pages 1.4-16 and 3.13-20.)
Yes. Instrument development, if done rigorously, is an iterative process that requires psychometric assessment over time. Continually assessing reliability and validity of instrument scores facilitates rigor and instrumentation improvement as society also evolves. Recently, a Rasch analysis was conducted to test the psychometric properties of the 5 traits ProScan Survey Part 2, including the testing of new words. Data were analyzed using the Rasch measurement model, which is a special case of Item Response Theory, and provides a contemporary approach to assess reliability and validity. While similar to a Classical Test Theory (CTT) approach in that both CTT and IRT methods assess psychometric properties including reliability and validity, IRT offers several advantages over classical assessment methods. Item Response Theory (IRT) offers a more nuanced analysis of survey items by assessing how each item performs across different levels of ability, allowing for more accurate measurement of individual abilities. Unlike Classical Test Theory (CTT), IRT accounts for varying item difficulties and differing levels of the underlying trait being measured, providing a more detailed understanding of how items differentiate between survey-takers. Additionally, IRT’s models are more generalizable and less dependent on specific sample characteristics, making it easier to apply results across diverse populations.
Individuals who completed a ProScan Survey in the English language were included in this study (n = 10,349) (Table 1). Initially, all latent constructs were tested for reliability and validity using a single factor model for each trait. All mean-square infit and outfit values were within the acceptable range of .5 to 1.5, with scores ranging from .70 to 1.40, which demonstrates excellent evidence of validity. Item and person reliability values were all above the acceptable threshold of .90 and .70, with item reliability values for each trait between .99 to 1.0 and person reliability values between .72 and .79.
Table 1. Sample Size (n) for each Round of Data Collection
Round 1—2,091
Round 2—2,255
Round 3—3,240
Round 4—2,763
Overall, each trait exhibited excellent psychometric properties after rigorous analyses, providing confidence when using the ProScan Survey. Instrument development is inherently a cyclical process, demanding multiple iterations of data collection, statistical analysis, and refinement. However, this critical process is often disregarded due to the significant investment of time and resources it requires. By investing valuable resources and technology into the continual assessment and refinement of the ProScan survey, the tool remains a reliable and up-to-date modern instrument, allowing users to confidently rely on its accuracy and effectiveness.
For more, see PDP Global's Modern IRT Study or Research Reference 5.0, pages 3.1-3.11.
Yes. Multiple analyses were performed to investigate the predictive validity and disparate impact of the PDP ProScan and JobScan on a sample of 218 short-haul truck drivers at Fleetline, Inc. Predictive validity is evident when test scores are significantly related to one or more criteria of job success. Disparate impact is absent when the relationship of test scores to performance indices is not a function of a demographic variable (e.g., age). The results indicated that the ProScan behavioral traits were valid predictors of three criteria: average miles driven per day, tenure with Fleetline, and organizational turnover. (For more, see Research Reference 5.0, pages 3.21-47)
Yes. The ProScan was tested for bias in scoring results based on race, gender, and occupation. The overall results revealed that the ProScan does not produce biased scores or discriminate based on race, gender, or occupation because the primary ProScan behavioral traits remained invariant across these demographics. (For more on disparate impact, see Research Reference 5.0, pages 4.3-17.)
Yes. When collecting data electronically, and especially in large amounts, it is common to encounter issues in the quality and integrity of the data from some respondents. With any type of data collection, it is customary to encounter respondents who may rush through the instrument’s questions without paying much attention to their responses (speeders), become distracted while completing questions (distracted respondents), or simply respond to questions with unreasonable responses that may take the form of systematic or identifiable patterns (straightliners/patterns of response integrity). Although these issues in data collection are commonplace, they are often overlooked due to numerous factors including the complex nature in which to identify some of these issues, as well as the cost and the large amount of time spent to comprehensively identify all of the potential issues.
As part of the continual process to invest in the evolution and quality of the ProScan instrument, responses provided from 29,082 responses were assessed based on multiple criteria to identify any integrity issues with the data. Overall, the ProScan was evaluated for response integrity and quality control based on 50 criteria. While this type of thorough quality control is not the status quo in data collection, it is important to note that these integrity checks are being performed on the ProScan for each and every response.
For more, see PDP Global's Data Quality Control and Integrity Methodological Process or Research Reference 5.0, pages 5.1-5.4.
The ProScan and Job Dynamics Analysis surveys are written at a reading level appropriate for an average eighth-grader. The Flesch Reading Ease test gives a score from 0 to 100; the higher the score, the easier the material is to read. The ProScan has a Flesch Reading Ease score of 73.4, indicating that it is at or slightly below an eighth-grade reading level. The ProScan utilizes a 5-point Likert scale, providing respondents with a range to rate themselves, instead of utilizing a forced-choice method.
Yes. PDP continually promotes Equal Employment Opportunity rights through its commitment to both quality statistical research and the implementation of fair and non-discriminatory practices in the workplace. As early as 1990, PDP psychologist Dr. Bonnie Bass met with Dr. Raymond Kearns of the Office of General Counsel in Research and Analytics Services at the Equal Employment Opportunity Commission (EEOC) in Washington, DC, to ensure that PDP was in line with the best practices of the EEOC. Since that time, PDP has continued to provide instruments and certifications to clients to assist them in complying with EEOC guidelines. (For more on the EEOC and PDP, see Research Reference 5.0, pages 9.1-10)
Several legal and professional opinions have been documented. Here are two examples:
In 1993, Dr. Gardner, a professor of psychology at the University of Colorado Denver, praised PDP’s efforts:
“I am pleased with the high standards of research methodology you set, not only initially, but also in ongoing development. It is heartening to find a business that values accuracy and that takes additional measures to insure [sic]validity and reliability with their products. In my opinion, PDP has taken appropriate steps to ascertain the validity of their tests. They have consistently used sound experimental procedures to make these determinations and are continuing these activities on an ongoing basis.” (Research Reference 5.0, page 2.21)
In 2020, David N. Schachter, attorney at law, stated: “As the company’s counsel, based on all of the above, I am very comfortable advising you and any affiliates requiring such, that PDP, both in terms of internal practices and contents of its suite of SaaS (software as a service) products, is fully compliant with, and in fact exceeds, federal and state requirements and regulations regarding nondiscriminatory and nonbiased applications.” (See Legal Opinion for more information.)
For further information on PDP Global’s research and validity, please contact your PDP Representative.
© 2001, Rev. September 2024. PDP, Inc. USA
