INTRODUCTIONThe Work Personality Index 2nd Edition Revised (WPI2-R) is essentially a short-form of the Work Personality Index 2nd Edition (WPI2), which has been available for more than 20 years. While the current form of the assessment is not obsolete or outdated, the WPI2-R improves the quality of the assessment, as well as the experience for both test takers and administrators.Statistical analysis of the original 198 items and scale results indicated that the number of items could be reduced to 113 while maintaining the reliability and validity of the original assessment. The psychometric properties of the WPI2-R, as well as how the revised version of the Work Personality Index compares to the original version can be found throughout this document.
HOW DOES THE UPDATED ASSESSMENT COMPARE TO THE ORIGINAL VERSION?When updating any assessment, it is important to establish that the new version of the assessment is measuring the same constructs as the previous version. The extent to which one can generalize the established validity and interpretive guidelines from the WPI2 to the WPI2-R hinges on the relationship between the two forms. (For more information on these constructs and guidelines, please refer to the section titled “Conceptualization of the Work Personality Index model” below.)The following considerations were examined to establish that the updated version of the assessment is measuring the same constructs, and is comparable to, the WPI2.
In order to determine which items could be removed from the original WPI2, we gathered a research sample consisting of 113, 755 individuals. In order to run the analyses outlined below, we used the Raw Scores for the 113 item “Revised” version of the assessment. Keep in mind that each scale contains a different number of items, therefore the minimum and maximum Raw Score will vary from scale to scale.
For the analysis of the psychometric properties, we applied the 113-item model to the original normative group (n = 8360). Similar to the results above, the minimum and maximum score for each scale will be different in the analyses to follow (due to having a varying number of items for each scale). However, the scores found in the report use a standardized Sten score. Any interpretation found in the report are based off of these Sten scores, creating equality between the scores and allowing them to be compared.
During the collection phase for the research sample, respondents were able to voluntarily provide demographic information. The collection of demographic information was optional based on the modern criteria for data collection and data privacy, as well as the larger concerns regarding these areas. Anyone who provided demographic information did so of their own choice. There was no incentive for doing so.
As a result, 8995 individuals provided demographic information. This data was used to conduct Adverse Impact analyses to determine if the assessment was treating any groups unfairly.
DEMOGRAPHICSThe following demographic information was asked of each individual.
ANALYSES OF ADVERSE IMPACTUsing the collected data, we looked to see if there are any meaningful differences between groups. Out of the collected demographics, we examined gender, ethnicity, educational level and job level for group differences. Some notable exceptions:
Business professionals continue to search for methods that enable the effective use of human resources. In order to make the best use of their personnel, organizations have focused on identifying the characteristics of workers that predict exceptional performance. It is common knowledge that people have skills and capabilities that allow them to be successful in certain types of work. A wide variety of tools and techniques have been developed to identify these skills.
Currently, assessment experts recognize that a person’s success at work involves much more than his or her cognitive abilities. The ability requirements of an occupation represent only one type of attribute that influences the potential for success. Differences revolving around individuals’ preferences and typical work behavior also influence how successful they may be. The idea that people differ from one another is quite obvious, and it is conceivable that individuals with certain traits are more effective in some jobs than others. Research over the years has indicated that workers’ personality traits are related to job performance in a wide variety of occupations. Personality traits, which are the distinctive patterns of behaviour that characterize an individual, have been found to play a key role in an individual’s effectiveness at work.
Personality can be thought of as a combination of an individual’s traits and tendencies. These traits are relatively enduring dispositions that distinguish one person from another. Personality traits describe an individual’s tendency to act, think, or feel in a certain manner. These traits tend to be stable and do not change much over the course of a person’s lifetime. Similar to cognitive abilities, aspects of personality can be clearly and efficiently assessed through psychometric tests and questionnaires. Personality tests measure unique characteristics that cannot be identified by ability and aptitude tests, and as a result, they provide complimentary and unique information that can be extremely useful for selection, career, development and team-building applications.
The Work Personality Index assessment is a questionnaire that identifies personality traits that directly relate to work performance. By helping identify individuals’ personality traits, the WPI can help select candidates, guide career development, and improve team functioning. This primary focus of the WPI helps professionals make the most efficient and effective match between people and work roles.
The WPI focuses on the traits that are important in the work environment. It measures personality traits for the normal adult population and does not examine clinical or mental health related issues. As a result, low scores on the scales are not indicative of pathology, but rather, different preferences and motivations for working. These different preferences and motivations influence the type of work in which people are successful at and what they enjoy doing.
The WPI assesses 21 primary scales that measure distinct aspects of work personality which allow professionals to make logical and informed connections regarding an individual’s preferences and their work behavior. These 21 scales are categorized into five groups that provide a global view of work personality.
With the Work Personality Index tool you can assess aspects of personal work style for a number of different applications. The most common uses of the WPI include personnel selection, leadership development, personal development, and team building. Personnel selection is one of the primary applications of the WPI since it helps match a person’s characteristics with those required for successful performance on the job. For leadership and personal development the WPI helps individuals identify their preferences and motivations and how these relate to different occupations and work environments. In team building the WPI can help resolve conflict and improve team communication.
USING THE WPI IN PERSONNEL SELECTIONThe Work Personality Index assessment is useful in personnel selection because it was designed to help match candidates’ characteristics with job requirements. The WPI, when used in conjunction with other techniques and tools, can increase the effectiveness, accuracy and objectivity of an organization’s selection procedure. By conducting a job analysis to identify the personality traits of successful employees, and using the WPI to identify the presence or absence of these traits in candidates, professionals can increase the efficiency and accuracy of their personnel selection process. For example, when hiring salespeople, the candidates who tend to be successful score high on theAmbition, Persistence, and Outgoing scales. By using the WPI to screen for applicants who have these characteristics, you can enhance the quality of those hired and improve turnover reduction.
Career development revolves around helping people gain self-awareness of their skills, preferences and values and then identifying occupations that provide a good fit. The WPI can help individuals gain an understanding of their work preferences and link them to possibly satisfying occupations. This information is extremely useful for people who desire to identify potential careers and make choices regarding further education and skill development. Results from the WPI are also helpful for examining job satisfaction in a potential career area. For example, individuals who score low on Energy and high on Attention to Detail usually enjoy work that follows a steady pace, uses methodical procedures, and is well organized. When these same people are placed in a fast-paced, loosely structured environment, they tend to experience dissatisfaction, and potentially poorer performance outcomes.
USING THE WPI IN TEAM BUILDINGThe WPI can also be effective in team building applications. The key aspect of effective team building is to help individuals understand themselves and the other members of their team. This process allows the team to make the most of each person’s strengths and gives each team member an understanding of why and how others adopt different approaches to their work. By helping team members appreciate the differences amongst themselves, they can make constructive use of their diverse skills. For example, with teams working in a structured environment on projects that follow a strict schedule, it is important for someone on the team to adopt a leadership role. In settings where a team needs to brainstorm ideas and develop alternative solutions, it is important that team members be innovative and democratic.
USING THE WPI IN LEADERSHIP DEVELOPMENTThe WPI allows organizations to assist leaders identify their strengths and common challenges in areas important to leadership development including how they work with others, how they set goals, approach their work, solve problems, deal with stress, and manage change. Using the WPI in leadership development will help an individual increase their effectiveness and identify pivotal issues – including leveraging their strengths and overcoming key weaknesses that, when addressed, will significantly enhance their performance and effectiveness.
Personality is the pattern of behaviour that characterizes an individual and his or her reactions to the environment. Most experts believe that personality is composed of a combination of traits that distinguish one person from another. Traits are enduring tendencies to act, think or feel in a certain way in any given circumstance.
Every individual’s combination and strength of different traits leads him/her to act a certain way in a wide variety of circumstances. For example, extraverted people tend to be outgoing in most settings. When it comes to measuring extraversion and other traits that make up personality, we know that personality questionnaires can provide a reliable and valid assessment of these traits. Our current understanding of personality supports the claim that personality traits can predict effective performance in many occupations. Therefore, it is easy to see why organizations would like to measure personality. If they can identify an individual’s traits, which predict how the person will act and react to the job, the organization can select the candidates who have the traits associated with for successful performance increasing the probability of employee success. There are many models of personality that measure a large number of personality traits. Some of these models are built around two or three traits, while others may have upwards of thirty. There is considerable debate around the true number of personality traits that should be measured. Rather than trying to measure all aspects of human personality, the WPI examines only those traits that empirical research has shown to be related to successful work performance.
The Work Personality Index model is built upon the personality traits identified in the Occupational Information Network (O*NET) developed by the U.S. Department of Labor. This model is not based upon a theoretical view of human personality, but is a combination and ordering of personality traits that have been shown to predict job performance. The model was formulated by examining two main sources. Firstly, predictive or concurrent validity studies provide important evidence of the traits that can be measured effectively and that predict job performance. Examining these studies provide a number of personality traits that consistently relate to ongoing effectiveness on the job.
Second, existing taxonomies that are already used in personal development and personnel selection were reviewed. To examine these taxonomies, the personality measures that operationalize them were analyzed, and their research critiqued. These personality measures included: the California Psychological Inventory, by Harrison Gough, the Hogan Personality Inventory, by R. Hogan and J. Hogan, and the NEO PI-R, by Paul Costa and Robert McCrae. Reviewing these personality assessments led to the identification of other personality traits that are closely tied to work preferences and motivations.
By examining these two main sources, 17 primary scales were identified, which the O*NET researchers grouped into 7 global scales. Since the development and release of the original version of the WPI, further research has revealed the usefulness of additional 4 scales. These include Planning, Multi-Tasking, Social Confidence and Persuasion. This version of the WPI has 21 primary scales.
Factor analytic studies indicate that the 21 scales can be categorized into 5 global scales. Figure 2.1 illustrates how these scales are organized. The 5 groups are labeled Energy and Drive, Work Style, Working With Others, Problem Solving Style, and Dealing with Pressure and Stress. These constructs have been re-labeled since the first version of the WPI: Achievement Orientation is now Energy and Drive; Conscientiousness is now Work Style; Social Orientation is now Working with Others; Practical Intelligence is now Problem Solving Style; and Adjustment is now Dealing with Pressure and Stress. These groups closely mirror the global traits identified in the Five-Factor Model of Personality (Costa and McCrae, 1992). The 21 primary scales represent a finer grained assessment of the 5 constructs. For example, the Work Style construct contains the following Primary scales: Attention to Detail, Dependability, Persistence, Planning, and Rule-Following. Descriptions of the 5 global constructs and the primary traits are shown below.
ENERGY AND DRIVEEnergy and Drive involves working hard and wanting to get ahead, persisting in the face of obstacles, and striving for career success. This construct has been an important component of personality theory for many years. In the Five-Factor Model, Energy and Drive is captured by the Conscientiousness factor. However, the WPI separates achievement striving from the dependable and disciplined behaviours that are grouped in the Conscientiousness factor of the Five- Factor Model. This construct is commonly called Achievement Striving, Assertiveness, and Ambition.The Energy and Drive composite contains the following primary scales: Ambition, Initiative, Flexibility, Energy, Leadership, Multi-Tasking, Persuasion, and Social Confidence.
WORK STYLEWork Style involves being planful, careful, dependable and disciplined in one’s role. Research has shown that Work Style is consistently related to work performance in a wide variety of occupations.The five primary scales that reflect the Work Style composite are: Persistence, Attention to Detail, Rule-Following, Dependability and Planning.
WORKING WITH OTHERSWorking with Others is represented by sensitivity to the needs of others, a willingness to work cooperatively rather than independently, and a preference for working with others and establishing personal relationships. This composite closely resembles the Extraversion factor of the Five-Factor Model.The elements of Working with Others are found in the following primary scales: Teamwork, Concern for Others, Outgoing, and Democratic.
PROBLEM SOLVING STYLEProblem Solving Style involves characteristics such as insight, imagination, originality, being open to new ideas, and maintaining a thoughtful approach to work. This construct is commonly found in many personality taxonomies and has been labeled Openness to Experience, Openness, Culture, Intellect, and Intellectance in previous measures.The Problem Solving Style composite found in the WPI is composed of two primary scales: Innovation and Analytical Thinking.
DEALING WITH PRESSURE AND STRESSThe Dealing with Pressure and Stress composite found in the WPI closely resembles the Neuroticism composite found in the Five-Factor Model. Representing the tendency to remain calm, composed and free from worry in stressful situations, other common labels for this construct include Emotional Stability, Negative Emotionality, and Worrying.The two primary scales that reflect the key aspects of Dealing with Pressure and Stress are Self-Control and Stress Tolerance.
This chapter covers the administration guidelines that describe the settings and populations where the Work Personality Index assessment can be used effectively. The most important consideration when having people complete self-report measures of personality is to create an atmosphere where the individual feels at ease and free to accurately report their traits. This is usually best achieved when the administration is clear and consistent, and respondents’ questions are handled appropriately.
The WPI is largely self-administered, and can be completed individually or in groups. The instrument is administered mainly on-line through a web-based test administration system. While there is no time limit to the administration, most people complete the assessment in approximately 20–30 minutes. Those who take longer may be encouraged to work more rapidly and not study the items at length.
No rigorous controls are required to establish dependable, reliable results. The WPI has been used in a variety of conditions, including online testing, formal testing, individual administrations, and take home administrations. While a standard, supervised administration is ideal, the reliability and validity of individuals’ results have not been negatively affected through less stringent administration conditions.
When conducting a supervised group assessment, the administrator should ensure that the assessment environment is relatively free from distractions, is quiet, and well lit. It is important to create an environment that makes the individuals taking the WPI as comfortable as possible. It is useful to follow the steps below.
AgeThe WPI measures personality traits immediately applicable to work settings. The test items revolve around typical work experiences, and for people to respond to them in an informed manner they need to be able to relate to the situations presented in each item. For this reason, the WPI is appropriate for people in the working population. Younger adults may not have enough work experience to respond to the items in a valid way. Therefore, it is recommended that the WPI be used with people who are older than seventeen years of age.
Ethnic GroupsThe WPI has been normed with individuals from a variety of ethnic and cultural backgrounds in the United States and Canada. Research results presented in Chapter 5 suggest that the WPI may be used effectively with people from different cultures with little probability of adverse impact.
Required Reading LevelThe reading level of the WPI questionnaire was assessed using two popular methods. The Flesch Reading Ease score rates text on a 100-point scale using the average sentence length and the average number of syllables per word. The score for the WPI questionnaire was 51.6, indicating that it is easily understandable by individuals 12 years or older. The Flesch-Kincaid Grade level score rates text on a U.S. grade-school level. For most standard documents, a score of approximately 7.0 to 8.0 indicates it can be read by the majority of the population. The WPI questionnaire is rated at 8.0, meaning that a typical eighth grader can understand the items.
The 21 Primary Scales measure precise dimensions, providing a specific analysis of an individual’s personality. In order to examine the many variations in how people approach and complete their work, spending some time carefully reviewing the 21 scales is often necessary. The 21 Primary Scales are bipolar, and therefore both low and high scores have meaning. Low scores are not “bad” and high scores are not “good.” Instead they describe an individual’s personality style that should be interpreted in the context of his/her current or potential work environment. Very low (sten scores of 1-2) and very high (sten scores of 9-10) scores on any of the scales are associated with both positive and negative elements. Scores that fall in the average range (5-6) suggest that the respondent shows a few of the strong tendencies and behaviors (both positive and negative) found at both ends of the scale. For example, people with high scores on the Teamwork scale work well with others. Yet when required to work alone, they may quickly become dissatisfied and have difficulty motivating themselves. On the other hand, people with low Teamwork scores work well alone, but often struggle when required to work closely with others.
AmbitionThe Ambition scale measures an individual’s tendency to set high standards, establish tough goals, and work to achieve success.
• set difficult goals for themselves• be driven to achieve high standards• have high aspirations and work hard to achieve their goals• be described as driven and competitive• put a lot of effort into everything they do
InitiativeThe Initiative scale measures the individual’s willingness to take on new or additional work responsibilities and challenges.
FlexibilityThe Flexibility scale measures the person’s capacity to cope with a frequently changing work environment and adapt to emerging situations.
EnergyThe Energy scale measures an individual’s stamina and the tendency to maintain a high level of energy.
LeadershipThe Leadership scale measures the willingness to lead, take charge of situations, and offer opinions and directions to others.
PersuasionThe Persuasion Scale measures an individual’s comfort in negotiating, selling, influencing and attempting to persuade people or trying to change the point of view of others.
Social ConfidenceThe Social Confidence scale measures the individual’s tendency to be self-assured and at ease with people in all types of social situations.
Multi-TaskingThe Multi-Tasking scale measures the individual’s preference for dealing with several activities at a time.
WORK STYLE SCALES
The Persistence scale measures the preference for sticking with tasks, and the element of not giving up and overcoming obstacles in completing one’s task.
• persist in the face of obstacles• enjoy overcoming challenges• keep trying to solve problems, no matter how difficult• dislike leaving things unfinished• want to see things through to the end
• give up when things become difficult• be easily distracted and have difficulty motivating themselves to complete boring work• dislike work that requires overcoming too many obstacles• prefer straightforward tasks that can be completed quickly and with relative ease
Attention to DetailThe Attention to Detail scale measures the individual’s tendency to focus on details, work towards perfection, and approach work in a neat and organized manner.
Rule-FollowingThe Rule-Following scale measures the tendency to adhere to rules and strictly follow work regulations.
DependabilityThe Dependability scale measures the extent to which a person is reliable, responsible, dependable and fulfills obligations.
PlanningThe Planning scale measures the individual’s desire to plan their work and to follow their plan.
WORKING WITH OTHERS SCALES
TeamworkThe Teamwork scale measures an individual’s tendency to be cooperative with others, display a good-natured attitude, and encourage people to work together.
• enjoy helping others reach their goals• encourage people to work together• like cooperative work environments• be pleasant and good-natured• rarely get frustrated with colleagues
• prefer working alone• be more formal and reserved• dislike working in large groups• be described by colleagues as distant or withdrawn• prefer environments where there are many opportunities to work independently
Concern for OthersThe Concern for Others scale measures how sensitive and understanding an individual is to the needs and feelings of others.
• be sympathetic and show concern for others• quickly pick up on the feelings of others• lend a helping hand and support colleagues• show a willingness to help people with their problems• be described as caring and understanding
• pay less attention to the personal feelings of others• be reluctant to get involved with individuals’ personal problems• miss minor cues that indicate what a person is feeling• take a more logical approach to solving problems• be more task oriented than people focused
OutgoingThe Outgoing scale measures the preferences for interacting with others and establishing personal connections with people.
DemocraticThe Democratic scale measures the preference for making decisions through consultation, collaboration, and working with close supervision.
PROBLEM SOLVING STYLE SCALES
InnovationThe Innovation scale measures the degree of creativity and open-mindedness when addressing work issues.
• have many original ideas• enjoy solving problems• enjoy work that requires creativity• be curious about many things, and enjoy learning• be open to new things• focus on new opportunities
• use established ways of working• prefer focusing their energy on narrow topics• be suspicious of new or unconventional ideas• keep to the task at hand
Analytical ThinkingThe Analytical Thinking scale measures the tendency to carefully analyze information and use logic to address issues and problems.
• enjoy discussing theoretical concepts• like solving complex problems• analyze problems from many different angles• be critical and take an impartial, logical approach to evaluation• be seen as calculating, cautious and deliberate• make decisions only after careful analysis
• make quick decisions• be more spontaneous and willing to make choices with limited information• dislike analytical tasks• rely on their intuition when making decisions
DEALING WITH PRESSURE AND STRESS SCALES
Self-ControlThe Self-Control scale measures the extent to which individuals maintain their composure, keep emotions in check, and control their anger.
• be slow to anger, and rarely lose their temper• maintain their composure, even in difficult situations• deal with problems in a calm and relaxed manner• rarely experience feeling angry with others• keep their emotions in check
• readily feel anger and frustration• be easily annoyed and quickly become upset• act impatiently when things do not go their way• let others know what they are feeling
Stress ToleranceThe Stress Tolerance scale measures the tendency to be accepting of criticism and to deal calmly and effectively with high stress situations.
• have difficulty relaxing• become tense when faced with many tasks• dislike high pressure work• take criticism personally• quickly become nervous and tense• worry about things that are beyond their control
The design criteria guided the development of the WPI in several ways. The personality traits measured by the WPI have been found to relate to work preferences and job performance in a direct way. This ensures that users are provided information that directly helps in the accurate selection or informed development of personnel. In traditional personality assessments, job relevant traits are measured alongside non-job relevant traits, making the interpretation of the results more difficult and thereby increasing the chances for making errors.
For example: People describe me as understanding.
In addition to the items from the first version of the WPI more than 300 items were written and field-tested over a period of 18 months. Over 15000 individuals took various forms of the assessment during this time period. A combination of classical test analysis and Rasch analysis of the items for each scale were examined.
The Rasch model (Rasch, 1980; Wright, 1979, 1982) is a mathematical formula that indicates the relationship between persons and scores that define a trait. The model is usually referred to as a one–parameter model, but in fact looks at two parameters: people and items. These are usually defined as person logits and item logits. The analysis provides a number of fit statistics that are indices of how well the data fit the model. The fit statistics include average fit and individual item fit. Fit statistics include “infit”, a chi-squared based statistic that is sensitive to unexpected patterns of observations by persons on items; and “outfit” statistics a chi-squared based statistic that is sensitive to unexpected observations by persons on items. High outfit or outfit mean squares greater than 1.5 indicate that an item is not fitting the model reasonably well. In addition, an examination of the mean infit and outfit statistics presents overall information about whether the data acceptably fit the model. It is expected that the mean infit and outfit for both persons and items should be about 1.0. At the level of an individual item the accepted general rule regarding infit and outfit mean square fit statistics is they should be between 0.5 and 1.5 (Linacre, 1995, 2002, 2004, 2010).In addition, the Rasch model allows the examination of how the items perform for different groups of people. This is known as Differential Item Functioning (DIF) (Zwick and Thorpe, 1996). DIF is carried out by comparing two groups of interest (for example, males and females). One set of parameters used to judge this are called DIF Contrast Logits. In general logits under 0.43 are considered to indicate neglible differences between groups: those between 0.44 and 0.64 to be slight, and those over 0.65 to be moderate or large (Linacre, 2010).Through this process, the research questionnaire consisting of 221 items (10 items per scale) was used to collect normative data. Examination of the large data pool from the normative study was once again analyzed using Rasch technology. This analysis suggested that the scales could function as well if not better by being reduced by 1 item per scale. Internal consistency and item reliability analysis resulted in the dropping of 21 items, bringing the total to 189 or 9 items per scale. The items that were retained contributed positively to coefficient alpha reliability and ensured that the number of items on the questionnaire remained manageable for users. Internal consistency reliabilities range from .70 to .89 (for complete internal consistency results see page 27). The Rasch analysis for the final 189 item version of the Work Personality Index assessment is summarized in Table 5.1.
The table above indicates that each of the 21 WPI scales shows very good model fit with average mean square outfit and average mean square infit statistics being extremely close to 1.00. Table 5.1 also displays the number of items that do not fit the model as measured by the item mean squared (infit or outfit). One item on the Ambition scale has a marginally high Outfit mean square. Considering that there are 198 item items in the total test this is an extremely positive result. Two items show DIF that is above the negligible level (both would be classified as slight to moderate): one in the Dependability scale (0.44) (Females higher than Males) and one on the Ambition scale (-0.51) (Males higher than Females).
WPI Scale
As part of the development process we did not wish for the new version of the WPI to fundamentally deviate in psychometric characteristics from the previous versions. For the common scales we therefore expected that the alternate form reliability would be high, as would the factorial structure of the two assessments. The previous version of the Work Personality Index consists of 153 items representing 17 scales. The new version consists of the same 17 scales with the addition of Multi-Tasking, Planning, Social Confidence, Persuasion and an 8 item validity scale. The first step in judging the comparability of the new and previous versions of the WPI was to look at alternate forms of reliability for the scales. This is a measure of how the individual scales compare across versions. 1852 individuals completed both versions of the WPI. In general, observed correlations between the two forms are shown in Table 5.2. Correlations are very high, indicating that the traits measured by both versions are similar and that any resulting interpretation would also be similar.
Additionally, the factor structures of the 17 common scales across the two forms were compared. The inter-correlation matrices of the 17 WPI scales for each version of the WPI were calculated. A principal components analysis was applied to each correlation matrix with varimax rotation. A comparison was made between the factor structures using procedures and methods outlined by Barrett, Petrides, Eysenck, & Eysenck (1998). Comparison of the factor solutions with varimax rotation for the two assessments was made using congruence analysis, following the procedures outlined by Barrett (1986). In the analysis the North American norming sample was used as the target matrix with comparisons made to data gathered on the new version of the Work Personality Index. Results are shown in Table 5.3. All congruence coefficients are above 0.90 which is indicative of congruence between factors (Barrett, 1986; Ten Berge, 1986). The overall coefficient of congruence (0.99) for the analyses demonstrates high similarity for all five factors. The present study sought to demonstrate the initial validity and factor invariance of the Work Personality Index scales across the two versions of the assessment. The number and content of the factors are similar. All five factors showed near perfect equivalence across versions. The results suggest that the factor structure of the WPI is consistent across versions. At this level, participants responded to the two versions of the WPI in a highly similar fashion. Overall, this study supports the validity of the WPI factor structure. This gives administrators confidence that the WPI may be utilized across versions with similar interpretations.
Norming is a key step in test development. The norms establish the baseline by which all test results are measured against, which allows the comparison of different individual scores. Norms identify the below average, average, and above average performance on the test, and help the test user appropriately interpret a person’s results and make decisions. The more people that are included in the norm sample the more we can ensure that the test norms represent the actual distribution of personality traits of the people in the population. In turn, this allows the test results to be more accurate and informative when comparing different individuals.
A person’s results on the WPI are reported in a standard score format known as Sten Scores. Standard scores are converted raw scores that help with the interpretation of the test results by allowing the comparison of an individual’s results with the norm group. Standard scores also help compare a person’s primary scale scores against each other. This allows us, as an example, to determine if the person scores higher on Flexibility than on Persistence. Sten scores range from 1 to 10, have a Mean of 5.5, and a Standard Deviation of 2. This means that an individual with a Sten score of 5.5 precisely reflects the average score of the norm population. As a result, 50 percent of the norm sample would score above and below the individual.
Table 5.4 provides a detailed description of the WPI raw scale scores for the norming sample. The means and standard deviations shown provide the norms which individuals who complete the WPI are compared against. The mean raw score for each scale represents the “average” score of people in North America. The standard deviation indicates the spread of scores found among people in the normative sample. Approximately 68% of the population will obtain scores within one standard deviation above and below the mean, while 95% of the population will score within two standard deviations of the mean.
STANDARD ERROR OF MEASUREMENTEvery time someone completes the WPI, their scores are comprised of two components. The first component is their true score – the amount of their score represents the personality trait being measured. The second component is a random score – the proportion of their score that is attributable to external conditions which have nothing to do with the trait being measured, known as chance or external influence. For example, if an individual is distracted or tired, it may impact their responses, and thus the score they obtain on the WPI. Since this random score has a negative influence on the consistency of an individual’s results, it is usually called random error. The Standard Error of Measurement (SEM) gives an indication of how much an individual’s obtained score might vary from their true score. The SEM can be interpreted as meaning that an individual’s true score will fall within +1 or- 1 SEM 68% of the time, and within +2 or -2 SEM’s 95% of the time. To illustrate, consider the following. Paul receives a Sten score of 6 on the Ambition scale. We know that 68% of the time, his true score will lie in between 5.09 and 6.91 (+-1 SEM). We also know that 95% of the time, his true score will lie between 4.17 and 7.83 (+-2 SEM).
DEMOGRAPHICS OF THE NORMING SAMPLEThe final sample consists of 8360 individuals (4180 females and 4180 males). A detailed breakdown of the Age, Education Level, Ethnicity, Employment Status, Position Level, Work Experience and Occupational Area of the norming sample is provided in Tables 5.6 to Table 5.12.
Agriculture, Fishing and Forestry
Installation, Maintenance and Repair
DESCRIPTIVE STATISTICS FOR WPI STEN SCORES OF THE NORMING SAMPLE DEMOGRAPHIC GROUPSThe following tables (5.13 to 5.20) present the means and standard deviations of the Work Personality Index scales for each demographic category.
Since the WPI is used to compare people, including both males and females, it is important to have an understanding of the gender differences found on the 21 scales. A number of minor gender effects were discovered when comparing the mean scores of males and females. Most of the differences are quite small in magnitude. In general, females tended to receive higher scores on Concern for Others, and Multi-Tasking. Males tended to receive higher scores on Analytical Thinking and Persuasion. As demonstrated earlier only 2 items show DIF above the negligible level (both would be classified as slight to moderate); one in the Dependability scale (Females higher than Males) and one on the Ambition scale (Males higher than Females). Since the differences between the other WPI scales were minimal these outcomes should not influence test interpretation.
EthnicityThe norm sample for the WPI contains a significant number of minorities, allowing for the examination of ethnic differences. A test which finds significantly large variations among ethnic groups can result in adverse impact, making it inadvisable to use such a test in the selection process. These types of differences are commonly observed when using cognitive ability tests. However, these variations tend to be less frequent and less pronounced for measures of personality such as the WPI. Table 5.14 lists the mean score for the WPI dimensions for seven ethnic groups. While minor differences can be observed between the ethnic groups, their effect on test interpretation is minimal.
Age GroupsSome psychological traits can change throughout a person’s lifetime, and others tend to remain very stable. Examining the differences among people of different ages can help determine the effects that age may have on a person’s reported personality traits. Age effects for the WPI are summarized in Table 5.15, where age is divided into nine categories. While some differences can be found between the different age groups, the overall similarity of the mean scale scores across the nine age groups indicates the WPI can be useful in comparing the personality traits for people from ages that range from 15 to over 60.
Educational LevelVariation of WPI scores for people with different levels of education also provides evidence of the usability of the WPI for individuals with different educational levels. Some of the personality traits measured by the WPI would appear to interact with the type and level of education achieved by people. For example, individuals with less than a high school education tend in general to be lower than the average in all traits except Democratic, Persuasion, and Rule-Following. People with post-graduate degrees tend to be higher than average on Analytical Thinking, Concern for Others and Initiative and lower on average on Rule-Following.
Employment StatusThere is little variation in WPI score across employment status. Interestingly, Homemakers are above average in Attention to Detail, Concern for Others and Rule-Following; the Self-Employed are above average in Flexibility, Initiative and Innovation and below average in Rule-Following.
Position LevelVariation of WPI scores for individuals with different positions in organizations also provides evidence of the usability of the WPI. Some of the personality traits measured by the WPI would appear to interact with the position level that individuals currently have. For example, Executives and Top Executives tend to be above average on a number of traits including: Ambition, Analytical Thinking, Energy, Flexibility, Initiative, Innovation, Leadership, Multi-Tasking, Persuasion and Social Confidence, as well as below average in Democratic and Rule-Following.
Years of ExperienceThere is little variation in WPI score with years of experience. Few of the groups show average sten scores of more than 0.5, indicating that the groups are highly similar overall.
Occupational GroupsTable 5.20 lists the average sten scores for 34 different occupational groups. Key findings are listed below. For each scale, the occupational groups with the highest scores (generally one-half of one sten above the normative group mean) are listed.
Key Findings
When developing assessments for psychological constructs such as personality, it is necessary to conduct research to determine the reliability and validity of the tool. Reliability directly examines the consistency and stability of the assessment, while validity is concerned with the types of inferences (e.g. predicting work behavior, identifying better employees) that can be made from test results. For an assessment to be useful it needs to be both reliable and valid. The following chapter examines evidence for both the reliability and validity of the WPI.
Reliability is concerned with the consistency of test scores, and how free test results are from external, confounding influences. The higher the reliability of a test, the more likely it is consistently measuring actual differences between people. More reliable tests provide results that remain unaffected by irrelevant variations, or what is commonly called random errors. We assess the reliability of the WPI by examining how consistently the test measures personality traits. For example, a test that yields similar scores for a person who repeats the test at a later point in time is said to be reliable. However, if a person takes the same test twice and receives very different scores, the test is unreliable. In general, reliability refers to how dependable a test is.
Reliability is measured using correlation coefficients. A reliability coefficient is denoted by the letter “r”, and is expressed as a number ranging between 0 and 1.00 with r=0 indicating no reliability, and r=1.00 indicating perfect reliability. It is important to recognize that tests are never 100% accurate, so you will not find a test with a correlation coefficient of r=1.00. In general you will see the reliability of a test expressed as a decimal, for example, r=.80 or r=.93. The larger the reliability coefficient, the more consistent are the test scores. There are a number of reasons and/or conditions that lead to unreliable test results. Some of the possible reasons include the following:
These three factors are sources of chance or random measurement error in the assessment process. If there were no random errors of measurement, the individual would get the same test score (their “true” score) each time. The degree to which test scores are unaffected by measurement errors is an indication of the reliability of the test.
INTERNAL CONSISTENCY OF THE WPIOne of the main approaches used to assess reliability is through measures of internal consistency. A sophisticated form of internal consistency reliability is Cronbach’s alpha. It effectively splits the test items in every possible way and computes the average of all combinations. Consistency should be achieved in such a way that all the items on a scale measure the same thing to the same degree, and, therefore, the items for each test scale should have a high degree of correlation with one another. Most professionals agree that test scales with correlation coefficients above .70 are useful for most applications. The internal consistency reliability coefficients for each of the WPI scales are listed in Table 6.1. As shown in the table, the reliability coefficients range from 0.68 to 0.91 for the Norm sample. The average consistency across all scales is 0.83 for the norm sample. Most of the reliability coefficients for males and females in the sample are close – the largest difference being Flexibility (Females – 0.80 and Males – 0.72). The table also shows reliability coefficient for a number of different national groups: Australia, Canada, South Africa, United Kingdom, United States and for the French version of the assessment. Similar coefficients are found across all countries and for English and French. The strength of the reliability coefficients indicates that the Work Personality Index is relatively free from external errors that could negatively impact the measurement of personality traits. This allows the WPI to provide a close estimate of each individual’s true score on the 21 scales.
The internal consistency reliability coefficients for each of the WPI global scales are listed in Table 6.2. As shown in the table, the reliability coefficients range from 0.87 to 0.95 for the Norm sample. The average consistency across all scales is 0.91 for the norm sample. Most of the reliability coefficients for males and females in the sample are very close. The table also shows reliability coefficient for a number of different national groups, Australia, Canada, South Africa, United Kingdom, United States and for the French version of the assessment. Similar coefficients are found across all countries and for English and French. The strength of the reliability coefficients indicates that the Work Personality Index global scales are relatively free from external errors that could negatively impact the measurement of personality traits. This allows the WPI to provide a close estimate of each individual’s true score on the 5 global scales.
TEST-RETEST RELIABILITY OF THE WPITest-retest reliability is an estimate of how stable a characteristic is over time. It is calculated by correlating the results of a group of individuals tested on two separate occasions. The degree to which the scores are similar will indicate the test-retest reliability of the test. Evidence of this type of reliability is important for personality measures since personality traits are thought to be quite stable and change very little over time. A sample of 145 people completed the WPI on two separate occasions. The sample consisted of 100 females and 45 males. The median time interval was 62 weeks, and the time intervals ranged from 12 to 120 weeks. Table 6.3 lists the test-retest indices for the 21 WPI traits, which range from 0.78 to 0.90. The mean reliability index across all 21 scales was 0.85 for the total sample, 0.85 for the female sample, and 0.84 for the male sample and indicate that the WPI traits show consistency over time. Table 6.4 lists the test-retest indices for the 5 WPI global scales, which range from 0.87 to 0.91. The mean reliability index across all 5 scales was 0.89 for the total sample, 0.89 for the female sample, and 0.88 for the male sample and indicate that the WPI global scales show consistency over time.
When deciding to use a test, validity is arguably the most important consideration. While reliability focuses on how consistently a test measures a particular trait, validity examines the extent to which a test measures what it claims to measure. Validity describes the degree to which you can make specific conclusions or predictions about people based on their test scores. In short, validity shows the usefulness of a test. A test’s validity is established in reference to a specific purpose; the test may not be valid for all purposes. This means that a test is never valid or invalid. Instead, how the test is used can be classified as valid or invalid. For example, a ruler is a valid measure of a person’s height, but an invalid measure of that individual’s technical proficiency or leadership style.
CORRELATIONS WITH OTHER MEASURESOne primary source of validity evidence is known as construct validity. Construct validity evidence shows whether the test is appropriate to measure a particular psychological construct. The most popular method of examining construct evidence is to compare two different tests that are supposed to measure the same construct. For example, if the Outgoing scale on the WPI is valid, individual’s scores should relate with other tests that also measure extraversion and social orientation. A number of correlational matrices have been derived involving the WPI and other psychological tests and instruments. This research was conducted to provide further evidence of the validity of the WPI as a measure of personality traits. The findings listed below outline the relationship between the WPI and other established measures of personality, values and interests.
A sample of 455 individuals completed the WPI and the MBTI® Step I assessment (Myers, McCaulley, Quenk, and Hammer (1998). The MBTI® measures four dichotomies: Extraversion vs. Introversion (EI), Sensing vs. Intuition (SN), Thinking vs. Feeling (TF), and Judging vs. Perceiving (JP). The analysis examined the correlations between the 21 WPI scales and MBTI® preference scores. In order to compute these correlations, the MBTI® dimensions were arranged as single continuous variables ranging from low to high: Extraversion-Introversion (EI), Sensing-Intuition (SN), Thinking-Feeling (TF), Judging-Perceiving (JP). As a result, people with low scores on these variables would identify with Extraversion, Sensing, Thinking, and Judging, while people with high scores would identify with Introversion, Intuition, Feeling, and Perceiving. The relationships between the 21 WPI scales and the MBTI® preference scales are shown in Table 6.5.
WPI Correlations with E-I preference scoresThe MBTI® Extraversion-Introversion scale correlates most strongly with Outgoing (-.79), Social Confidence (-.67), Teamwork (-.57), Energy (-.47), and Persuasion (-.40). These findings indicate that people who indicate a preference for Extraversion on the MBTI® tend to be outgoing, comfortable in social situations, have high energy, enjoy working closely with others, and like situations where they can be persuasive as measured by the WPI. These results suggest that these WPI scales are measuring some of the same personality characteristics as the Extraversion-Introversion scale on the MBTI®.
The Sensing-Intuition dichotomy describes what an individual pays attention to. People with a preference for Sensing tend to focus on information that can be gathered through the five senses. As a result, they tend to be practical, and focus on factual details. People with a preference for Intuition tend to focus on future possibilities and novelty. As a result, they tend to be theoretical, creative, driven by insights, and enjoy variety. The MBTI® Sensing-Intuition dichotomy correlates with Innovation (0.65), Flexibility (.53), Rule Following (-.50), and Attention to Detail (-.50). These findings indicate that people who indicate a preference for Intuition on the MBTI® tend to be identified as flexible and innovative by the WPI. Those individuals who score toward Sensing on the MBTI® are identified as preferring to pay close attention to detail and rule-following on the WPI.
WPI Correlations with T-F preference scoresThe Thinking-Feeling dichotomy describes the process an individual uses to make decisions. People with a Thinking preference tend to structure and organize information to make decisions in a logical, objective way. Individuals who have a preference for Feeling organize and structure information to decide in a personal, values-oriented way. The MBTI® Thinking-Feeling dichotomy correlates with Concern for Others (.53), Analytical Thinking (-.43), Leadership (-.33), Ambition (-.33), and Democratic (0.25). As expected, people with a Feeling preference score high on the Concern for Others and Democratic scales on the WPI. Those individuals who score towards Thinking tend to be analytical, ambitious and have leadership aspirations.
WPI Correlations with J-P preference scoresThe Judging-Perceiving dichotomy describes the type of lifestyle a person adopts. People with a preference for Judging live planned, organized lives. Individuals with a preference for Perceiving enjoy living a spontaneous, flexible life. MBTI® Judging-Perceiving scores correlate with Planning (-.72), Flexibility (.60), Attention to Detail (-.56), Rule-Following (-.51), Dependability (-.45), Innovation (.38), Multi-Tasking (.37), and Persistence (-.35). This indicates that individuals who show a preference for Perceiving also tend to be flexible, innovative and enjoy multi-tasking. Those individuals prefering Judging tend to be planful, to follow guidelines closely, are persistent and dependable, and focused on details.
WPI AND MBTI® FORM Q STEP IIIn this research study, 369 adults completed the Work Personality Index and the Myers-Briggs Type Indicator® Form Q (Quenk, Hammer, and Majors (2001), which is commonly referred to as the MBTI Step II. The MBTI® Form Q was designed to measure facets of the four Personality Type dichotomies of Extraversion-Introversion, Sensing- Intuition, Thinking-Feeling, and Judging-Perceiving. There are 20 subscales that identify components of each of the four type dichotomies and provide a finer grained interpretation of an individual’s personality preferences. The MBTI® Form Q facet scales measure aspects of personality that are similar to those identified through the 21 scales on the Work Personality Index. Therefore, one would expect to find a number of significant correlations between the scales of these two assessment instruments. Table 6.6 lists the 20 facet scales from the Form Q. The purpose of this study was to identify the relationships between the WPI and MBTI® Form Q scales, and examine whether these relationships fit with current knowledge and understanding of personality. Correlating the scores of an assessment like the WPI with those of other instruments is a common method of establishing validity. This type of comparison provides validity evidence by determining if the WPI scales correlate with Form Q scales in ways that would be expected. For example, one would expect that the WPI scale of Outgoing demonstrated significant relationships with the Form Q scales that measure aspects of Extraversion.
The most detailed study of the relationship between the Work Personality Index assessment and the MBTI® Form Q is obtained when correlating the results from the 21 WPI traits and the 20 Form Q facets. Since each of these scales measures distinct personality traits, rather than global traits, the comparison between the two tests at this level provides the best indication of their relationship. Tables 6.7 to 6.10 show the correlations between the 21 WPI traits and the 20 Form Q facet scales.Extraversion – Introversion Facet ScalesThe first five scales on the MBTI Form Q measure facets of the Extraversion-Introversion dichotomy. They would be expected to correlate with the scales on the WPI that measure aspects of social outgoingness and activity. The results shown in Table 6.7 are in line with expectations. Some significant results include:
Sensing – Intuition Facet ScalesThe next five facet scales on the MBTI Form Q measure aspects of the Sensing-Intuition dichotomy. These would be expected to correlate with WPI scales that measure whether people are practical and detail-oriented, or innovative and focused on possibilities. The relationships between the WPI and the MBTI Form Q appear to be consistent with the hypothesis and are shown in Table 6.8. Significant findings in this area include:
Thinking – Feeling Facet ScalesThe Thinking-Feeling facet scales help identify typical ways in which people make decisions and the standards they use to maintain relationships. One would expect that these facet scales show some relationships to the WPI scales that measure aspects of how people relate to others, and how they deal with information. Table 6.9 lists the correlation coefficients. Some of the more interesting results include:
Judging – Perceiving Facet ScalesThe final five facet scales on the MBTI Form Q measure personality traits that are considered part of the Judging-Perceiving dichotomy. These facets reflect whether an individual is planful and organized, or spontaneous and flexible. The relationship between the J-P facets and the WPI scales are also in line with expectations, and are provided in Table 6.10. These results include findings wherein:
ConclusionThe results of this study indicate a number of significant relationships between the WPI and the MBTI® Form Q. These relationships demonstrate that both tests measure some of the same personality characteristics. All of the MBTI Form Q facet scales are significantly correlated with at least one of the WPI scales. These findings lend support to the use of the WPI as a valid measure of personality traits that play an important role in how people approach their work.
WPI AND THE CAREER VALUES SCALE (CVS)The Career Values Scale (Macnab, Bakker, and Fitzsimmons, 2001) is an assessment of career values that identifies a hierarchy of importance of particular values in a person’s life and work. Two versions of the CVS are available. The original CVS inventory that consists of 88 Likert-based questions. The paired-comparison version of the CVS contains 45 paired-comparison items. These values are categorized into three main areas:
WPI AND THE CAREER VALUES SCALE (CVS) LIKERT VERSIONA sample of 882 individuals completed both the Work Personality Index assessment and the Career Values Scales – (Likert-version). The correlations between the Work Personality Index and the Career Values Scale are shown in Tables 6.14, 6.15, and 6.16.
WPI AND THE CAREER INTEREST PROFILER
A sample of 1170 individuals completed both the Work Personality Index assessment and the Career Interest Profiler (Bakker and Macnab, 2004). Table 6.17 shows the correlations between the WPI and Career Interest Profiler. The Career Interest Profiler is a measure of occupational interests that uses John Holland’s theory of vocational personality. This theory, which has become one of the most widely accepted approaches for helping people make informed occupation choices, is based on six vocational personality types. Holland believed that people could be described by one of the six types:
THE WPI AND THE LEARNING STYLES INDEX (LSI)
Table 6.19 shows the correlations for the WPI and the LSI for a sample of 68 learners.
THE WPI AND THE SALES ACHIEVEMENT PREDICTOR (SALESAP)The SalesAP (Friedland, Marcus and Mandela, 1995) is an objective measure of characteristics that are considered critical for success in sales. The test consists of separate measures for overall Sales Disposition, Cold Calling, and Sales Closing, in addition to many characteristics related to sales potential and performance such as Assertiveness, Personal Diplomacy, and Patience. A sample of 3671 people took both the SalesAP and the WPI. Table 6.20 shows the correlations between the WPI and the three components of the SalesAP that measure sales related characteristics. The Sales Disposition score (SAL) indicates the degree to which an individual’s SalesAP results are similar to those observed for people who are successful in sales careers; the Cold Calling (CC) scale that reflects characteristics necessary for success in cold calling activities; and the Sales Closing (CLS) score indicates the degree to which an individual’s SalesAP results are similar to those observed for people who are successful in closing sales.
THE WORK PERSONALITY AND THE NEO PI-RThe NEO PI-R (McCrae and Costa, 2010) is a measure of five major factors of personality and some of the important facets that define each factor. The five factor scales and thirty facet scales provide a general assessment of adult personality. Thirty participants completed both the WPI and the NEO PI-R Form S (the self-report version of the assessment). The correlations between the WPI scales and the various NEO PI-R scales are shown in Tables 6.21 and 6.22. There are many conceptual similarities between the WPI and the NEO PI-R dimensions which the data analysis confirms.NeuroticismNeuroticism reflects the amount of Anxiety, Hostility, Depression, Self-consciousness and Impulsiveness an individual displays. Individuals who score highly on Neuroticism also score higher on the WPI scale of Attention to Detail and lower on the Energy, Stress Tolerance and Self-Control scales.ExtraversionExtraversion indicates the amount of Warmth, Gregariousness, Assertiveness, Activity, Excitement-Seeking and Positive Emotions an individual has. Individuals who score highly on Extraversion scored higher on the WPI scales of Outgoing, Teamwork, Initiative, Flexibility, Dependability, Concern for Others, Innovation, Leadership, Energy and Ambition.OpennessOpenness indicates the amount of Fantasy, Aesthetics, Feelings, Actions and Ideas an individual has. Individuals who score highly on Openness scored higher on the WPI scales of Innovation, Analytical Thinking and Flexibility and lower on Rule-Following.AgreeablenessAgreeableness indicates the amount of Trust, Straightforwardness, Altruism, Compliance, Modesty and Tender-Mindedness an individual has. Individuals who score highly on Agreeableness scored higher on the WPI scales of Self-Control and Concern for Others.ConscientiousnessConscientiousness indicates the amount of Competence, Order, Dutifulness, Achievement-Striving, Self-Discipline and Deliberation an individual has. Individuals who score highly on Conscientiousness scored higher on the WPI scales of Dependability and Persistence.
The correlations shown in Table 6.22 between the WPI Global Scales and the NEO scales indicate expected relationships between the two inventories: Neuroticism and Dealing with Pressure and Stress; Extraversion with Working with Others and Energy and Drive; Openness with Problem Solving and Conscientiousness with Work Style. Agreeableness does not correlate significantly with any of the WPI Global Scales.
WPI AND THE EMOTIONAL QUOTIENT INDEXThe Emotional Quotient Inventory (EQi; Bar-On, 2002) is an inventory designed to measure Emotional Intelligence by means of 133 items on a five-point response scale. The assessment aims to measure an array of non-cognitive abilities relating to an individual’s coping ability and general psychological well-being. The EQi comprises of five composite scales, fifteen subscales, four validity scales, and also renders a total EQ score. The five composite scales are: Intrapersonal, Interpersonal, Adaptability, Stress Management and General Mood. The EQi presents scales for each of the composite scales. These are shown in Table 6.23 below
A sample of 361 adults took both the EQi and the WPI. Correlations between the EQI Scales and the WPI Global Scales are shown below.
• Emotional Self Awareness assesses the “ability to recognize one’s feelings” (Bar-On, 1997). Individuals who score highly on Emotional Self Awareness scored high on the WPI scales of Concern for Others (0.44) and Teamwork (0.33). • Assertiveness assesses the ability to “express feelings, beliefs and thoughts and defend one’s rights in a nondestructive manner” (Bar-On, 1997). Individuals who score high on Assertiveness scored high on the WPI scales of Leadership (0.53), Stress Tolerance (0.42), and Initiative (0.39). • Self-Regard measures the ability to “accept oneself as basically good” (Bar-On, 1997). Individuals who score high on Self-Regard scored high on the WPI scales of Stress Tolerance (0.52), Dependability (0.46), Initiative (0.45) and Self Control (0.45). • Self-Actualization measures the ability to “realize one’s potential capacities” (Bar-On, 1997). Individuals who score high on Self-Actualization scored high on the WPI scales of Initiative (0.57), Ambition (0.52), Dependability (0.52), and Persistence (0.52).
Interpersonal Scales Summary • Empathy measures the ability to “be aware of, to understand, and to appreciate the feelings of others” (Bar-On, 1997). Individuals who score highly on Empathy scored higher on the WPI scales of Concern for Others (0.74), Teamwork (0.41) and Outgoing (0.36). • Interpersonal Relationships measures the ability to “establish and maintain satisfying relationships” (Bar-On, 1997). Individuals who score highly on Interpersonal Relationships scored higher on the WPI scales of Outgoing (0.56), Teamwork (0.54), Concern for Others (0.42), Initiative (0.42) and Energy (0.40). • Social Responsibility measures the ability to “demonstrate oneself as cooperative, contributing, and constructive member of one’s social group” (Bar-On, 1997). Individuals who score highly on Social Responsibility scored higher on the WPI scales of Concern for Others (0.67), Teamwork (0.40) and Dependability (0.39).
General Mood Summary • Happiness measures the ability to “feel satisfied with one’s life” (Bar-On, 1997). Individuals who score highly on Happiness scored high on the WPI scales of Initiative (0.44), Stress Tolerance (0.43), Energy (0.42), Self-Control (0.41) and Outgoing (0.40). • Optimism measures the ability to “look on the brighter side of life” (Bar-On, 1997). Individuals who score highly on Optimism scored high on the WPI scales of Initiative (0.60), Ambition (0.56), Persistence (0.54), Stress Tolerance (0.54), Self-Control (0.52), and Energy (0.52).
Adaptability Summary • Problem Solving assesses the ability to “identify and define problems as well as to generate and implement potentially effective solutions” (Bar-On, 1997). Individuals who score highly on Problem Solving scored high on the WPI scales of Persistence (0.61), Ambition (0.57), Analytical Thinking (0.54), Dependability (0.53), and Attention to Detail (0.49). • Reality Testing measures the ability to “assess the correspondence between what is experienced and what objectively exists” (Bar-On, 1997). Individuals who score highly on Reality Testing scored high on the WPI scales of Persistence (0.55), Dependability (0.50), Attention to Detail (0.48), Self-Control (0.47), and Ambition (0.46) • Flexibility is the ability to “adjust one’s emotions, thoughts and behavior to changing situations and conditions” (Bar-On, 1997). Individuals who score highly on Flexibility scored high on the WPI scales of Initiative (0.59), Ambition (0.57), Flexibility (0.54), and Stress Tolerance (0.51).
Stress Management Summary • Stress Tolerance measures the ability to “withstand adverse events and stressful situations” (Bar-On, 1997). Individuals who score highly on Stress Tolerance scored high on the WPI scales of Stress Tolerance (0.72), Self-Control (0.64), Initiative (0.56), Ambition (0.55), and Persistence (0.53). • Impulse Control assesses the ability to “resist or delay impulse, drive or temptation to act” (Bar-On, 1997). Individuals who score highly on Impulse Control scored high on the WPI scales of Dependability (0.52), Persistence (0.50), Self-Control (0.49), Initiative (0.42), and Stress Tolerance (0.40).
ANOTHER STUDY COMPARING THE WPI AND THE EQ-IA recent study van Zyl and Taylor (2010) compared the WPI and the EQi with a sample of 1602 South African working adults. Significant correlations were found between most WPI scales and those of the EQi. The authors highlight that high correlations between the assessments are as one would expect, with scales such as Self-Control and other WPI scales correlating with Impulse Control on the EQi (0.52), Stress Tolerance on both scales correlated 0.59. Some notable correlations are highlighted below.• Energy, Initiative, Persistence and Stress Tolerance with Total EQi scores.• WPI scale Concern for Others with EQi Interpersonal Composite (0.57) as well as with Empathy, Social Responsibility, Interpersonal Relations and General Mood.• The Mood Scales on the EQi had high correlations with Dependability, Energy, Initiative, Persistence and Stress Tolerance.• Of 340 possible correlations only 35 are not significantly correlated.
SUMMARY OF KEY CORRELATIONS OF THE WPI AND OTHER MEASURESThe Table below shows a summary of the key correlations between Work Personality Index scales and other assessments. Table 6.25 below summarizes the correlations between the Work Personality Index and other aforementioned measures of personality and is organized by WPI scale.
CRITERION VALIDITY FOR TAX COLLECTION OFFICERSThe following statistical analyses were conducted to explore the relationships between the personality traits measured by the WPI and job performance ratings for 24 collection officers from a government tax collection agency. These employees were rated on 4 scales: Investigation Skills, Timeliness of Actions, Use of Collection Tools, Maintenance and Control. Given the small sample size, the relationships should be viewed as general trends or hypotheses. There are a number of significant correlations, indicating a strong relationship between the WPI and job performance for this sample. Investigations Skills correlates with Dependability (.65), Rule-Following (.61) and Persistence (.45). Timeliness of Action correlates with Dependability (.66), Ambition (.60), Energy (.59), Persistence (.54), Stress Tolerance (.47) and Initiative (.44). Use of Collection Tools correlates with Dependability (.71), Persistence (.60), Ambition (.59), and Rule-Following (.49). Maintenance and Control correlates with Rule-Following (.63), Dependability (.59), Persistence (.49) and Attention to Detail (.41).The strength of these correlation coefficients indicate that the WPI scales of Ambition, Initiative, Energy, Persistence, Rule-Following, and Dependability show promise in being able to predict aspects of job success for this occupation.SELECTING TRAINEES FOR TECHNICAL MANAGERIAL POSITIONSIn a large national organization, 30 trainees and 17 employees completed the Work Personality Index and their job performance was rated by supervisors. The participants were rated for Safety, Attention, Prioritizing, Communication, Multi-Tasking, Dealing with Stress, Flexibility, Technical Knowledge, Teamwork and Overall Effectiveness. WPI Job Match scores were calculated from the score ranges and importance rating provided by the expert raters. The WPI Job Match score is based on subjectmatter expert defined score ranges and importance rating for each scale. This information is used in a scoring algorithm that provides a score for an individual that ranges between 0% and 100%. A candidate who receives a Job Match Score of 100% has trait scores that fall within the benchmark for all of the traits. In essence, the Job Match Score shows how well an individual’s personality results fit with the benchmarks.THE JOB MATCH SCORE WAS THEN CORRELATED WITH THE PERFORMANCE RATINGS FOR THE TRAINEE GROUP.Table 6.26 shows the correlations between the Job Match Score and performance ratings. These range from .085 for Technical Knowledge to .542 for Overall Effectiveness. It is expected that the WPI would have no relationship with measures of Technical Knowledge, as knowledge and personality have shown little to no relationship in previous research. Areas of significant correlation, such as Communication, and Multi-Tasking relate more to aspects of personality which are measured by the WPI. A correlation of .542 with Overall Effectiveness accounts for 29% of the variance in ratings of trainees overall job performance, a statistically significant result that can indicate that the WPI may be useful for making hiring decisions.In addition to correlation data, we can illustrate how the Job Match Score fits with classifications of job effectiveness. The trainees were split into three groups based on their overall effectiveness ratings. Trainees with ratings from 5 to 7 were assigned to a high group, trainees with ratings of 4 to a middle group, and those with ratings of 1 to 3 to a low group. The Table 6.27 shows the mean Job Match Scores and percentage who reached the established cut-score (85% Job Match Score) for participants in each groups.
SELECTING EMPLOYEES IN A LARGE RETAIL ORGANIZATIONA large retail organization was interested in creating a method for identifying effective customer service agents and refining their intake procedures so that the number of poor performers hired could be reduced. 228 incumbents completed the WPI and were rated by their managers on 12 specific areas of performance, and 2 global performance categories. These areas included job knowledge, attitude, reliability, sales skills and overall performance. The number of employees whose performance was unacceptable or requires improvement ranged from 2.2% for Safety to 13.9% for Selling Skills. The Managers ratings indicated that the majority of candidates hired were performing at levels rated good or better. The number of good or better employees ranges from 52.9% on Selling Skills to 77.7% on Teamwork. Managers also responded that they would likely rehire 73% of the rated individuals if these employees ever left the organization.Table 6.28 shows the correlations between the WPI Job Match Score and performance ratings. These range from .114 for Selling Skills to .322 for Reliability. All correlations are significant except for Selling Skills.
In addition, employees were categorized based on their job performance into three areas: high, average and poor. Using information from subject-matter experts a WPI cut-off score was established. The percentage of employees in each performance category that would be screened out if the WPI Job Match cut-off score was used would be: 6.5% of High Performers; 12.1% of Average Performers; 50.0% of Poor PerformersThe cut-off score was also used to evaluate its impact on voluntary and involuntary turnover. The cut-off would eliminate 51% of employees whose turnover fell in the voluntary category, and 60% of those employees whose turnover was classified as involuntary (poor performance, failed training, breach of conditions etc.). As a result, the cut-off score could also have an impact on the prediction of those candidates who are likely to leave either in an involuntary or a voluntary manner.Upon investigation of historical job candidate performance, 36% of job applicants did not meet the job fit cut-offs. If the cut-off score had been used as a yardstick with current employees, it would have eliminated 69% of employees whose performance was rated as Unacceptable or Requires Improvement, while only eliminating 14% of current employees whose performance was rated as Very Good or Excellent.
SELECTING SALES AND CUSTOMER SERVICE EMPLOYEES IN A LARGE RETAIL ORGANIZATIONCorrelations between the employees’ assessment results and managers’ ratings of the selling skills of 109 Sales and Service employees’ performance in a large retail organization showed that there was a significant relationship between performance ratings and WPI Job Match scores. The Job Match score, which incorporates the personality benchmarks, correlated significantly with managers’ ratings of Selling Skills. This indicates that employees with higher Job Match scores are rated as having better Selling Skills.
Correlations between the employees’ assessment results and managers’ ratings of the sales ability of 81 sales staff in a large retail organization showed that there was a significant relationship between performance ratings and several WPI scales. The scales with the highest relationship were Energy (0.34), Multi-Tasking (.30), and Initiative (0.28). It should be noted that the scales normally associated with sales performance – Persuasion and Social Confidence were not highly correlated with performance. The reason for this is that all of the sales people scored high on both these scales in comparison to the norming sample leading to the lack of variation in scores and non-significant correlations.WPI – SALESAP CONCURRENT VALIDITY STUDYThe Work Personality Index assessment measures many components that are highly correlated with sales ability and other sales characteristics. The SalesAP assessment classifies candidates into different categories: Highly Recommended for Sales, Basically Recommended for Sales, and Not Recommended for Sales. Individuals with an outcome score greater than the 80th percentile are categorized as Highly Recommended; those with scores under the 50th percentile are classified as Not Recommended. This analysis looked at the ability of the WPI to predict the correct SalesAP groups to further establish its criterion validity.
Classification matrices were produced for both the estimation sample and the holdout sample. These classification or confusion matrices provide a convenient summary of the number of correct and incorrect classifications made by the discriminant procedure, and provide an index of the concurrent validity for the cross-validation holdout sample. Table 6.31 summarizes the results of the classification analysis. The first part of the table shows the percentage of correct and incorrect classifications based on the classification equation for the Estimation sample. As the table indicates, 98.0% of the original grouped cases were correctly classified. When the function was applied to the validation sample the overall correct classification was 97.9% indicating that the WPI is extremely good at predicting membership in different sales groups.
INTER-CORRELATION BETWEEN WPI SCALESThe correlations in Table 6.32 show the relationships between the 21 WPI scales. Reviewing the table shows there are significant relationships among some of the scales. This suggests that scores on some of the scales could be combined to form a broader level of analysis. In general, the scales that correlate highly together can be grouped into five global areas that closely represent the Big Five personality actors. For more information on how these scales relate to each other, read the following section on Factor Analysis.FACTOR ANALYSISFactor Analysis is a statistical process used to examine the WPI’s subscale structure and assess the extent to which the WPI Personality Model outlined in Chapter 2 is justified. In simple terms, factor analysis helps determine how the 21 WPI scales relate to each other. Principal components analysis was used to assess the factor patterns underlying the Work Personality Index. Factors were extracted on the basis of having eigenvalues greater than 1.0. These were rotated to orthogonal simple structure by the Varimax procedure. The results of the principal components analysis for the WPI are presented in Table 6.33. Five factors were extracted and rotated, accounting for 66% of the total variance. The eigenvalues and percentage of variance accounted for are presented at the bottom of Table 6.33. The rotated factor loadings are presented in the top part of Table 6.33. The factors are described below in terms of their highest loadings and by comparison to previous studies.The five factors identified are quite similar to the Big Five personality traits that are best represented by the work of Barrick and Mount (1991) and by Macnab and Bakker (2001). Factor 1 – the Energy and Drive factor – is characterized by high positive loading on Leadership, Energy, Persuasion, Initiative, Social Confidence, Outgoing, Ambition, and Multi-Tasking. This factor is similar to the Achievement Orientation factor reported by Macnab and Bakker (2001) with the first edition of the Work Personality Index. Factor 2 – Work Style – is characterized by high positive loadings on Attention to Detail, Planning, Rule-Following, Dependability, and Persistence an high negative loadings on Flexibility. This factor is similar to the Conscientiousness factor reported by Macnab and Bakker (2001). Factor 3 – Dealing with Pressure and Stress – is characterized by high positive loadings on Self-Control and Stress Tolerance. This factor is similar to the Adjustment factor found by Macnab and Bakker (2001) in an earlier edition of the WPI. Factor 4 – Working with Others – is characterized by high positive loadings on Teamwork, Outgoing, Concern for Others and Democratic. This is similar to the Social Orientation factor found by Macnab and Bakker (2001) with the first edition of the WPI. Factor 5 – Problem Solving Style – is characterized by high positive loading on Analytical Thinking and Innovation. This is similar to the Practical Intelligence factor reported by Macnab and Bakker (2001) with the first edition of the WPI.
FACTORIAL INVARIANCE BETWEEN GENDERSA study was conducted to examine the factor structure of the Work Personality Index assessment across genders. The similarities between male and female samples were examined by means of factor analysis and congruency coefficients. The inter-correlation matrices of the 21 WPI scales for each sample were calculated. Principal components analysis was applied to each correlation matrix with varimax rotation.Comparison of the factor solutions with varimax rotation for males and females was made using congruence analysis, following the procedures outlined by Bartlett (1986). The congruence coefficient was 0.97 which is typically indicative of showing congruence between factors (Barrett, 1986; Ten Berge, 1986). This demonstrates high similarity for all five factors amongst both males and females.FACTORIAL INVARIANCE ACROSS COUNTRIESA study was conducted to examine the factor structure of the Work Personality Index assessment across countries and languages. The similarities between the Norming Sample, Australian, South African, UK, Canadian, US and French samples were examined by means of factor analysis and congruency coefficients. The inter-correlation matrices of the 21 WPI scales for each sample were calculated. Principal components analysis was applied to each correlation matrix with varimax rotation.
Comparison of the factor solutions with varimax rotation for the seven samples was made using congruence analysis, following the procedures outlined by Bartlett (1986). In each analysis the Norming Sample was used as the target matrix with comparisons made to each of the other six samples. All congruence coefficients are above 0.90 which is typically indicative of showing congruence between factors (Barrett, 1986; Ten Berge, 1986). The overall coefficient of congruence for each of the analyses is as follows: Norming Sample and Australian sample – .98, Norming Sample and South African Sample – .93, Norming Sample and UK Sample – .96, Norming Sample and US Sample – 0.98, Norming Sample and Canadian Sample – 0.98 and, Norming Sample and French -.95. This demonstrates high similarity for all five factors across all countries.
All five factors showed near perfect equivalence across countries and genders. The results suggest that the factor structure of the WPI is independent of the gender, country and translation of the inventory and support confidence in the invariance of the instrument across multiple samples. At this level, participants from different countries and different genders responded to the WPI in a highly similar fashion. Overall, this study supports the validity of the WPI factor structure. This provides administrators the first level of confidence that the WPI may be utilized across countries with similar interpretations. In addition, it suggests that the personality structure as measured by the WPI should hold up across gender, cultures and languages.
This section provides information on a number of Special Purpose Scales developed in the process of norming the Work Personality Index Assessment. These include a Managerial and Leadership Potential Scale; a Sales Potential Scale; and WPI Profile Validity Scale.
A close analysis of group differences revealed that there were a number of significant relationships among the Work Personality Index assessment and position level. 2198 people identified themselves as Managers, Executives and Top Executives. The means and standard deviations for this sub-sample are shown in Table 7.1. The table illustrates that this sample is higher than the average sten score (5.5) in a number of scales, specifically Leadership, Initiative, Multi-Tasking, Energy, Flexibility, Persuasion, and Social Confidence and lower on Rule-Following and Democratic scales.
All means differences are significant at p<.001 except Planning, Dependability and Self-Control. The largest differences between the Management group and the Non-Management Group are shown below:
On the basis of these findings it was decided that a Managerial and Leadership potential scale be developed. An analysis of the item data revealed that a selection of 53 items from 14 of the above WPI scales discriminated well between groups in different position levels. These items showed an internal consistency coefficient of 0.91 and a test-retest coefficient of 0.85. The total score for the Managerial and Leadership Potential scale was calculated. This score was then standardized and sten scores calculated. Tables 7.3 and Table 7.4 in the norming section show the means and standard deviations for the Managerial and Leadership Potential Scale for the norm sample by Position and Occupation. Table 7.3 shows a summary of the scores by position level. It is clear that the higher the position level the higher the Managerial and Leadership Potential scale score. Entry Level and Non-Supervisory participants are 2 to 3 Stens lower than Executives and Top Executives on the scale. This pattern is similar for males and females. There are no significant differences between males and females within each position level, as indicated by similarities in the means and standard deviations within each group.
Further evidence for the validity of the scale may be determined by examining selected group mean scores on the Managerial and Leadership Potential Scale. Table 7.4 shows that means for the Management group has the highest score on the scale, with the Sales Management group also being high on the scale. The lowest mean scores on the scale are for Personal Care and Service, Office and Administrative Support, Customer Service -Technical support and Food Preparation and Serving occupational groups.
Our research has found multiple relationships between sales characteristics and the 21 WPI scales. Along with significant correlations between measures of sales personality (SalesAP), the WPI was able to discriminate between individuals who were considered Highly Recommended for sales from those Not Recommended on the SalesAP. Based on these findings, a new scale was calculated – a Sales Potential scale. This scale is based on the discriminant function analysis described earlier. The best indicators for discriminating between the Not Recommended and Recommended groups were the WPI scales of Persuasion, Social Confidence, Leadership, Energy and Initiative. 50 items were selected and represent the Sales Potential Scale. The total raw scores for the scale were standardized and sten scores calculated. These items showed an internal consistency coefficient of 0.94 and a test-retest coefficient of 0.86.VALIDITY INFORMATION RELATED TO THE WPI SALES POTENTIAL SCALEIn a study that examined perceived competence and the Sales Potential Scale a group of Sales Managers and a group of Retail Sales Workers were examined. The participants were asked to rate themselves on a three item competency scale (I am one of the top performers at work; I am very skillful at what I do at work; I am very good at my job). They were then classified into a Low Competency group and a High Competency group.22 Sales Managers rated themselves as having low competency in their current occupation and 21 rated themselves as being highly competent. Table 7.5 shows the means and standard deviations for both groups. The high competency group scored significantly higher on the WPI Sales Potential Scale than the low competency group (t=-4.52, df=41, p<.001).
39 Retail Sales workers rated themselves as having low competency in their current occupation and 43 rated themselves as being highly competent. Table 7.6 shows the means and standard deviations for both groups. The high competency group scored significantly higher on the WPI Sales Potential Scale than the low competency group (t=-4.315, df=80, p<.001).
CORRELATIONS OF THE WPI SALES POTENTIAL SCALE AND THE SALESAP GLOBAL SCALESThe relationship between the SalesAP global scales and the WPI Sales Potential scale was examined. The WPI Sales Potential Scale correlates 0.69 with SalesAP – Sales Closing; 0.71 with Sales Disposition and 0.76 with Cold Calling (all correlations are significant at a 0.001 level, n=3671).
The WPI Profile Validity Scale has been adopted from the WorkSafe Predictor (Meen & Macnab, 2012). The WorkSafe Predictor was created to assess patterns of thinking and acting that predict safe behavior and the likelihood of remaining free from workplace safety incidents. It is intended to provide insight into safety behaviours that can be used in a developmental, needs assessment or selection context. The Profile Validity Scale consists of 8 items that are extreme behavioural statements such as: “I never make mistakes”.The Profile Validity scale was designed to identify candidates who respond to the questionnaire in an overly positive or unusual way. When this occurs, the candidate’s test results may not provide an authentic picture of his/her personal style. While the number of candidates who misrepresent themselves tends to be small, it is not insignificant considering the costly effects of a bad selection decision. The Profile Validity Scale is designed to flag these candidates who may be manipulating their answers in order to come across in a highly socially desirable manner. Identifying people who may be misrepresenting themselves gives the employer the opportunity to confirm the candidate’s results through other assessment methods, and confirm whether or not the candidate’s responses are valid or invalid. The Profile Validity Scale is only reported on the WPI reports related to selection.
RELIABILITY OF THE PROFILE VALIDITY SCOREThe internal consistency coefficients for the Profile Validity Score are shown in Table 7.7 for both an applicant sample and for the norming sample.
ITEM ANALYSIS OF THE PROFILE VALIDITY SCALEThe Rasch model is a mathematical formula that indicates the relationship between persons and scores that define a trait. The model is usually referred to as a one–parameter model, but in fact looks at two parameters: people and items. These are usually defined as person logits and item logits. The analysis provides a number of fit statistics that are indices of how well the data fit the model. Detailed explanation of the Rasch model is shown in the chapter on development. The average infit for the Profile Validity Scale is 1.02 and for the outfit is 1.01 displaying that the items fit the model. There are no items that display infit or outfit. In addition, a DIF analysis was carried out comparing males and females – none of the items show DIF that is above the negligible level.
Ashton, M. (1998). Personality and job performance: The importance of narrow traits. Journal of Organizational Behavior, 19, 289-303.
THE CURRENT STATE OF LEADERSHIP POTENTIALA critical component for any organization is a strong leadership team. However, the leaders in an organization today will not be those who will lead in the future. Many organizations are left with the question “How do we fill this role?” Successful organizations seek to identify their future leaders from their current individual contributors, and will help them develop the necessary leadership skills. Not only does this training provide the organization with benefits in the future, it also helps these individuals contribute more in the short-term (e.g., build better relationships internally and externally, bring new ideas to the table that can help improve the organization, provide a source of motivation and engagement for the high potential). However, the criteria used to identify leadership potential is subjective in nature. In many cases, current leaders look for those who are most like them. In other cases, they apply subjective evaluations (e.g., performance evaluations from managers) or abstract criteria (e.g., they’re humble, they communicate well with the leadership team, they are loyal to the organization). While many organizations try to quantify these evaluations, they lack an objective perspective.THE NEED FOR EFFECTIVE LEADERSHIP POTENTIAL IDENTIFICATION AND DEVELOPMENTWhen asked to define the qualities needed for an effective leader, many organizations produce traits that are vague or even mysterious. This is completely understandable, as leadership qualities are often hard to identify. Sometimes, it seems leaders have an undefinable quality about them. In response to this challenge, many organizations have tried to fit candidates to various models based on subjective evaluation. However, the belief that the best future leaders will be those who fit these frameworks can be a costly mistake.Effective leadership potential identification begins with understanding the traits and skills that a leader will need within the organization. Many of these qualities may differ from company to company.After understanding the qualities that a leader needs within the unique environment, candidates can be identified by objectively measuring their workplace traits and preferences. Each individual will have strengths that they can harness in a leadership role, as well as areas where they can develop to be a more effective leader. Organizations can develop their best future leaders by treating candidates as individuals, understanding where they currently are and identifying how they need to develop.EFFECTIVE LEADERSHIP STARTS WITH SELF-AWARENESSSelf-awareness is often cited as one of the key factors towards becoming a great leader. Successful leaders understand themselves and those around them, allowing them to manage relationships, influence behaviours, manage conflict, lead change and make better decisions.Self-awareness development is most effective when based on an objective measure of an individual’s traits, as this will identify the individual’s strengths and development needs when preparing for a leadership role. Incorporating these measured preferences into their development plan will help the candidate realize their full potential as a leader.
The Work Personality Index Leadership Potential Report is an objective measure of an individual’s traits and preferences that will influence their performance as a leader. Identifying, understanding and developing these qualities will help create the most effective future leaders.
The Work Personality Index® Leadership Potential report measures and describes 21 key traits of an individual’s style that will influence their performance as a leader.
THE LEADERSHIP SAMPLE AND THE LEADERSHIP RANGEIn order to create the Leadership Range mentioned above, a sample of Senior Leaders was compiled. 712 Directors, Executives (e.g., Vice Presidents, Senior Vice Presidents) and Top Executives (e.g., C-levels, Presidents) were included in this sample. Demographics of the sample can be found in Appendix A.
The Leadership Range was constructed by determining the average score and the standard deviation. The Leadership Range for each scale consists of a half standard deviation above and below the average, to ensure that candidates are being compared with the traits shown by the majority of the leaders in the sample.
USING THE LEADERSHIP POTENTIAL REPORT FOR SELECTIONOne potential use for the Leadership Potential Report is to assist during the identification phase of a leadership potential program. This report is designed to bring objective, data-driven information about a candidate’s leadership strengths and weaknesses, which can help to explore questions such as “Does this person have the qualities that we need for the role?” and “Does this person have the qualities that suit the organization?”
In order to tune your identification process and integrate the Leadership Potential Report, we recommend that you consider the following:
USING THE LEADERSHIP POTENTIAL REPORT FOR DEVELOPMENTAfter harnessing the information during the identification or selection process, the Leadership Potential Report can be used to guide or support a development plan for the candidates. Below are some highlights and options for doing so.
In order to better help further coaching efforts while using the Work Personality Index Leadership Potential Report, the Leadership Range is presented for each scale. This allows coaches and trainers to compare the candidate’s scores to Senior Leaders. This will help the candidate understand how they are more similar to other leaders and how they differ.
There are two topics that will help candidates get the most out of debrief session: Their “Strengths to Harness” and “Areas for Development”. Page 3 of the report contains a starting point for this information.
Where is the candidate most similar to other leaders? (Strengths to Harness)This section provides the three scales where the candidate is most like the Leadership Sample. Based on this information, the presented scales have been identified as the strengths that the candidate can harness to refine their leadership capabilities. For detailed feedback around each scale, please see the corresponding page highlighted to the right of the page.
Where is the candidate most different from leaders? (Areas for Development)This section provides the three scales where the candidate is least like the Leadership Sample. Based on this information, the presented scales have been identified as the best traits to develop in order to strengthen their leadership capabilities. For detailed feedback around each scale, please see the corresponding page highlighted to the right of the page.
Of course, these comparisons are not meant to be seen as the sole consideration when determining leadership potential. Individual organizations may have different needs within their leadership roles, and these needs should be considered when developing coaching and training programs for future leaders. Additionally, leaders can still be successful without fitting perfectly into the Leadership Range.
SUMMARY
THE BACKGROUNDThe retail organization has more than 70 stores and 6000 staff covering a large part of Canada. With their focus on providing superior customer service and a positive shopping experience for all their customers, the organization continually evaluates its recruitment and training processes.
THE CHALLENGEThe organization identified the following key performance areas where successful employees excel:
The majority of these key areas are impacted by personality – those behavioural traits which are measured by the Work Personality Index. The validation study was completed to determine the likelihood that the Work Personality Index could identify candidates who will be productive, dependable, team players.
THE RESULTSThe study showed that Work Personality Index traits correlate significantly with employees’ overall and specific areas of performance based on managers’ ratings. To help managers make selection decisions Psychometrics Canada developed a scoring algorithm that used candidates’ assessment results to calculate a Job Fit Score. Those with the highest Job Fit Scores were rated as much better in sales performance, customer service and dependability. High scorers also had much higher ratings on overall job performance, and were much more likely to be recognized as good hires.Candidates with High Job Fit Scores:
Work Personality Index scales with the strongest relationship to job performance include Ambition, Initiative, Energy, Persistence, Attention to Detail, Rule Following and Dependability.
NEXT STEPSTo help managers make selection decisions Psychometrics Canada developed a scoring algorithm that used candidates’ assessment results to calculate a Job Fit Score. The correlation between the Job Fit Score and employees overall rating of performance is .428. Figure 2 shows the average Job Fit Score for employees that were classified into 3 different performance groups. Those employees with the highest performance ratings had a significantly higher Job Fit Score than those employees whose performance was rated as or good.
The Work Personality Index was then incorporated into the organizations applicant assessment and tracking system, allowing each store manager to easily administer the assessment to job candidates and view their candidates’ results. This integrated system also provided the head office with the ability to track results from each of the stores – giving them the opportunity to examine individual store activity, while also looking at overall assessment results for the entire organization – all from a central interface.
EXECUTIVE SUMMARYThis validity study evaluated the effectiveness of the personality and ability pre-screening assessments used to select new employees at a client Organization. The analysis reviewed the assessment results and performance of:• 122 current employees, whose job performance was evaluated by their managers.• 1008 employees, hired since August 1, 2013 who have since been terminated.• 9144 job candidates.The key findings of our research indicate that the pre-screening assessments are performing very well. Highlights include:• 9 out of 10 employees who pass the assessments are rated as satisfactory or better on their job performance.• Personality assessment predicts job performance in 11 out of 13 key areas.• Ability assessment screens out 14% of job candidates who struggle the most with simple math, reading, and logical reasoning.• There are few measureable differences between voluntary and involuntary terminated employees.
RESEARCH RESULTS
Client Organization Store Managers rated 122 randomly selected employees on 12 specific areas of job performance, and 2 global performance areas. These employees have worked at Client Organization for a minimum of 6 months, in a number of roles, and at a variety of city/provincial locations. The random selection of employees from a representative sample of stores and locations helps to ensure that the results are representative of all Client Organization store employees.
Job Performance Categories:
The Overall rating score provides the best global evaluation of the assessment effectiveness. Data indicates that 9 out of 10 employees who pass the current assessment benchmarks perform at the satisfactory level or higher; 7 out of 10 employees are rated as good or better.Described another way, the assessment identifies satisfactory or better employees 9 out of 10 times, and good or better employees 7 out of 10 times. These odds for predicting successful job performance are significant. While no selection system is 100% accurate, the current assessments and scoring benchmarks are effectively identifying candidates who perform well 90% of the time.Other results:
PRE-SCREENING ASSESSMENTS VALIDITY RESEARCHThe next areas of analysis examine the statistical links between the employees’ assessment results and managers’ ratings of the employees’ performance. The correlations between the personality assessment, ability assessment and the job performance ratings are described below.
Personality Assessment Validity CorrelationsJob candidates complete the Work Personality Index (WPI), a measure of 18 personality characteristics that describe how individuals complete their work, interact with others, manage change, and deal with stress. Candidates’ scores on the scales are used to calculate a Job Fit Score – this Job Fit Score is used as the cut-off to determine a candidate’s desired fit with the job requirements.
Correlations between the employees’ results on the WPI and their ratings of job performance are shown in Tables 1-6 below. These correlations are called validity coefficients, and they indicate the magnitude of the relationship between the assessment and the measure of job performance. They can be interpreted using the guidelines below:
The first area we examined were the correlations between the WPI Job Fit score and the 13 job performance criteria ratings. These are shown in Table 1. The data shows that the Job Fit score predicts job performance in 11 of the 13 job performance criteria ratings. The sizes of the correlations indicate that the predictive power of the personality assessment is significant. At the present time, the global score on the Work Personality Index assessment is effectively predicting positive job performance among candidates.
The following tables contain the correlations between the individual scales of the Work Personality Index tool and the job performance ratings. The majority of the scales correlate significantly with some aspects of job performance. In other words, these scales which are used in the calculation of the Job Fit score, continue to predict almost all aspects of job performance. The only two areas of performance not accounted for by the individual scales of the WPI are Selling Skills and Reliability. While Reliability is correlated with the Job Fit Score, Selling Skills is not significantly predicted by the personality assessment. This is also the area in which employees’ ratings are currently the lowest.
Ability Assessment Validity CorrelationsThe ability assessment was originally developed in 2009 using publicly available marketing and policy information from Client Organization as the source material. Simple numerical and verbal reasoning problems were developed from this material, giving them a realistic context and direct job applicability.The goal was to develop problem solving items that were set in a realistic context and would identify job candidates with the basic level of skill required to successfully work at Client Organization. Previous validity research, and the current performance of job candidates, indicates that the assessment is relatively easy, with most applicants doing quite well on the task. The average score for job candidates is 13.75 out of 16.Unlike normative ability assessments, where the goal is to identify the full range of skills, this assessment was designed to measure a candidate’s ability to work with the types of problems encountered in a retail setting. Our previous research with Client Organization found that successful candidates did not require high levels of numerical or verbal skill; instead they required basic skills in these areas, being able to function at or above the 25th percentile when compared to the general working population.
There are no significant correlations between the job performance ratings of candidates and their performance on the ability assessment. Average scores by overall ratings are listed below:
While those employees rated as unacceptable score the lowest on the ability assessment, the difference is not enough to be statistically significant. Overall, since individuals who struggle with the ability assessment are not hired, the restriction of range makes correlation research difficult. At the present time the ability assessment is only screening out candidates with significant challenges for working with the types of verbal and numerical information used at Client Organization.TERMINATED EMPLOYEES RESEARCHThe third and final focus of our research was on 1008 employees hired since August 1, 2013 who have since been terminated. The reasons for termination were used in the analysis, allowing for the comparison of employees who were either voluntary or involuntarily terminated. We also examined the relationships between employee tenure and assessment results.Reasons for TerminationThe reasons for termination, the percent of employees whose termination falls into each classification, and the average number of days of employment for those in each classification are listed below.
Comparison of Assessment Results for Voluntary Terminated, Involuntary Terminated, and Job CandidatesThe table below lists the average assessments scores for employees who fall into the two main categories of termination and all job candidates over the past 18 months.
Statistically significant differences exist between the average scores of candidates that fall into these three groups; however, these differences are too small to utilize in differentiating among future job candidates. The primary challenge of research in this area is the very small number of employees whose termination is involuntary. Out of all voluntary/involuntary terminated employees only 1.8% were terminated for unsatisfactory performance and only 1.1% were deemed not suitable for the job. This equates to only 29 people out of 1008. Tenure – Length of EmploymentThe table below shows the correlations between assessment scales and length of employment for terminated employees.
The length of employment shows small but significant correlations with the Ability Assessment and the Analytical Thinking scale. Given the minimal size of these effects, they are not enough to be of use in evaluating future job candidates.
In order to explore the connection between personality traits and safety-related behaviours and thinking patterns in the workplace, we examined the relationships between the scales in the Work Personality Index (2nd Edition) and the WorkSafe Predictor. Much of the literature around this topic examines broad measures of personality (such as the Big Five), but the Work Personality Index contains 21 focused traits of personality. This allows for a deeper dive into which aspects of personality are most related to safety behaviours.
The WorkSafe Predictor is designed to assess patterns of thinking and acting that predict safe behaviour and the likelihood of remaining free from workplace safety incidents. The assessment contains 109 Likert-based items, and explores patterns regarding 6 safety behaviours, including:
The assessment also contains a measure of trust regarding safety in the work environment (Safety Trust), and a validity scale to guard against faking (Motivational Distortion). We expect to find no relationship between Safety Trust and personality, as it is designed to examine attitudes and beliefs about the current external environment, not internal motivators.A sample of 1308 individuals completed both the Work Personality Index (2nd Edition) and the WorkSafe Predictor. The correlations between the two assessments can be found in Appendix A, indicating the relationships between personality traits and safety behaviours.
Key Findings:
Consistent with newer academic research in the topic (Doerr, 2020), one of the main personality drivers of safety behaviour is Conscientiousness. Two traits that impact an individual’s conscientiousness is Dependability (a measure of reliability and willingness to take responsibility) and Rule-Following (a measure of one’s willingness to follow established rules). Both traits were found to correlate with multiple safety behaviours, with Rule-Following correlating well with safety behaviours that involve identifying and avoiding risks in the environment. Also, and perhaps unsurprisingly, one’s Stress Response (safety behaviour) is closely connected with their natural Stress Tolerance (personality trait).
REFERENCESDoerr, A.J. (2020). When and how personality predicts workplace safety: Evaluating a moderated mediation model, Journal of Safety Research, 75, 275-283.
The assessment can be used in a variety of ways and settings, but is most commonly employed in corporate, educational, clinical, medical and research settings. The EQi measures a clearly defined construct that provides valuable information about an individual’s ability to effectively manage environmental demands and pressures.
STUDY 1 – WORK PERSONALITY INDEX
Sample DescriptionThe WPI sample consisted of 1,886 South African working adults that completed the assessment between 2005 and 2010 for selection and development purposes. The South African data were extracted by the publishers from their database, compiled and sent to the author. From this sample, 284 cases were removed due the possibility that they may be duplicate cases, leaving a total sample size of 1 602. Table 1 shows the demographic composition of the sample in terms of age and gender representation. Age was not reported on the WPI, however 38.4% of this information could be retrieved from the related EQi data.
Table 2 contains the descriptive statistics for each of the primary scales of the WPI. The mean scores were relatively high, but not unexpected since the sample is based on candidates who completed the assessment for selection purposes. This can be compared to the sten scores of the North American sample. Sten scores have a mean of 5.5 and a standard deviation of 2. Scores for the S.A. sample were higher on each of the scales of the WPI.
RELIABILITY ANALYSISThe internal consistency reliability of a test reflects the degree to which a specified set of items are sampling the same domain. The internal consistency reliability for each of the primary scales of the WPI was calculated using Cronbach’s alpha (a) coefficient (Cronbach, 1951). The closer the internal consistency reliability is to 1, the more reliable the test. Table 3 shows the Cronbach alpha coefficients for the South African sample. All of the scales except for the Democratic scale had acceptable reliabilities of 0.70 and higher.
FACTOR ANALYSISExploratory factor analysis is a useful tool with which to evaluate the construct validity of a test. Factor analysis was used to investigate the construct validity of the WPI at a scale level. The 17 primary scales of the WPI were subjected to a principal components analysis. The theoretical model specifies five global constructs, and in order to try and replicate the factor structure reported in the manual, five components were specified for extraction. Five components were successfully extracted and rotated to a simple structure by means of the Varimax criterion. The five components that emerged closely matched the theoretical structure of the WPI and the extracted component structure is similar to the structure reported in the manual. Table 4 displays the rotated component matrix. Primary loadings are marked in bold and underlined whereas secondary and tertiary loadings are in bold only.
All five of the Achievement Orientation scales had salient loadings on this component, making it identifiable as such. Another seven scales also had loadings of >0.35 on this component. These include: Persistence, Attention to detail, Dependability, Innovation and Stress tolerance which could likely prove to be important aspects of any Achievement Orientation construct. Concern for Others also had a salient loading of 0.35 and is further discussed in Component 2. A negative loading of -0.32 was found for the Democratic scale on this component. Considering the additional scales that also load on this component, positively and negatively, it becomes possible to speculate that this component may be indicative of a well adjusted worker or perhaps an expanded concept of important leadership qualities.Component 2This component was identified as Social Orientation since all of the scales for this global construct had salient loadings on this component. Interestingly, Concern for Others had a primary loading on component 3, a secondary loading on component 1 and only a tertiary loading on its expected component. When evaluating the items of the other Social Orientation scales of Teamwork, Outgoing and Democratic, it seems plausible that the weaker loading of Concern for Others may be the result of more intrapersonal types of questions as opposed to more interpersonal questions in the other scales.Component 3This component was identified as Adjustment with primary scale loadings >0.58 on Self- Control and Stress Tolerance. Stress Tolerance also had a secondary loading on component 1 of 0.49. Concern for Others had its primary loading on this component as mentioned above. When considering Self-Control at an item level it becomes clear that this scale largely measures controlled behaviour when interacting with others, suggesting an absence of hostile or cold behaviours. This might explain the strong loading of Concern for Others on this component since these constructs are more conceptually similar than different at item level. Persistence also had a tertiary loading on this component of 0.31. It is conceivable however that individuals characterised by high levels of Self-Control would also be persistent at their tasks.Component 4All the scales for Conscientiousness had salient loadings on component 4. Two of the four scales comprising this construct had primary loadings on component 1. These scales are Persistence and Dependability with loadings of 0.71 and 0.67 respectively. Component 1 was identified as Achievement Orientation, and on the WPI this construct stands alone, whereas on other personality inventories Achievement Orientation is often included as a subscale of Conscientiousness. The double loadings therefore make sense, since one would expect some level of overlap between the two constructs.Component 5The Innovation and Analytical Thinking scales had primary loadings on component 5 as postulated by the theoretical model for Practical Intelligence. The Innovation scale however had a secondary loading of 0.42 on component 1. This provides support for the notion that component 1 may be measuring a larger leadership construct. Both Initiative 0.32 and Leadership 0.37 had secondary loadings on this component. On a theoretical level, Initiative and Leadership might be expected constructs in any consideration of Practical Intelligence.The theoretical model of the WPI dictates that five components should be extracted. However, when inspecting the scree plot (Cattell, 1966) and considering Kaiser’s (1970) criterion of eigenvalues greater than unity, it appears that the extraction of three components might be more appropriate. If the WPI was not developed based on a theoretical model, and it still approximates the Five Factor Model, it seems likely that the three component structure extracted could also be explained by Eysenck’s (1992) 3-factor structure of personality. A three component solution accounts for 59.20% of the variance on the component matrix.
RASCH MEASUREMENTThe Rasch model is known as a fundamental measurement model, and is based on the assumption that the probability of achieving higher scores on a test increases as individuals possess more of a latent trait, and decreases as they possess less of the trait, an indication that items become more difficult to endorse (Green & Frantom, 2002). In other words, the probability of endorsing an item on a test is a function of the difficulty of the item and the ability of the person. The Rasch model is a method of logistic probability modelling that estimates item locations independent of the sample characteristics, allowing the researcher to make inferences about the test regardless of the distribution of the sample (Bond & Fox, 2001). The unit of measurement in Rasch analysis is the logit (or log-odds unit), and is the same for item location parameters as it is for person location parameters. The mean logit score is set at 0, with higher scores indicating greater difficulty and greater ability, and negative scores indicating lesser difficulty and lesser ability (Bond & Fox, 2001).Rasch AnalysisThe person separation reliability, the number of items that misfit (underfit and overfit) as well as the number of items displaying DIF for each of the scales of the WPI are reported in Table 5. The person separation reliability is a similar estimate to Cronbach’s alpha reliability estimate, and indicates how well individuals’ level of ability was estimated. Misfit occurs when items do not behave according to the stringent requirements set by the model. Items that do not adhere to the model parameters are classified as items that either underfit or overfit the model, depending on the relevant statistical value. Underfit means that the specific item behaves in an unpredictable way and may be measuring something else. Overfit means that the item is too predictable and may be considered superfluous. Differential Item Functioning (DIF) examines the extent to which various groups may be responding significantly different to items on the scale. DIF was investigated for men and women on items of the WPI.
MEAN DIFFERENCESThe means of the 17 scales on the WPI of the South African sample were compared to those of the North American normative sample (Sten score = 5.5). Table 6 contains the results of a one sample t-test for mean differences between the SA sample and the U.S. mean of 5.5. Cohen’s d effect sizes are also reported.
The S.A. sample scored significantly higher than the North American mean on all of the scales of the WPI. Cohen’s d effect sizes were calculated to further explore these mean differences. Medium to large effect sizes were found for 11 of the 17 scales (Cohen, 1988). These are highlighted in bold in Table 6. Some of the largest effect sizes were found for Rule Following (1.05) and Stress Tolerance (1.04). This significant score differences and noteworthy effect sizes highlights the need for locally developed norms for applicants in South Africa.
GENDERThe WPI sample of 1600 respondents was also subjected to tests for mean differences across gender. Table 7 reports the test statistics for each of the scales on the WPI. The mean differences in the subgroups should not be considered as evidence of test bias. Statistically significant differences between men and women were found for Leadership, Attention to detail, Dependability, Teamwork, Concern for Others, Outgoing, Innovation, Analytical Thinking and Stress Tolerance. However when considering the effect sizes for these differences, they are all found to be quite small (Cohen, 1992). The statistical significance may be due to the large sample size, and it appears that there were no real differences between men and women on the WPI.
AGEThe WPI sample was also subjected to mean difference tests, with the aim of comparing means across the different age categories used in the sample. The results of these tests are reported in Table 8 and Table 9. Statistically significant differences were found only for the Persistence, Dependability, Teamwork and Democratic scales of the WPI. The effect sizes were also calculated to further examine the magnitude of these differences. The partial eta squared values for each of the differences indicated that membership in a particular age group accounted for less than 1% of the variance in scores. Overall, no real differences were found with regard to age differences and scores on the WPI.
WPI AND EQ-ISample DescriptionThe EQ-i sample is based on a data-extraction from the JVR database comprising of 38285 working adults that completed the assessment from 2005 to 2010 for selection purposes. The data from the EQ-i was matched with the WPI data, rendering a total sample of 615 individuals who completed both the WPI and the EQ-i simultaneously or within a six-month period. Table 10 shows the demographic description of the combined sample differences and scores on the WPI.
Descriptive StatisticsThe Table 11 contains descriptive statistics for the South African EQi sample. Mean and standard deviation scores for each of the 15 subscales, 5 composite scales, as well as total EQ are reported. The descriptive statistics for the scales of the WPI are reported in Table 12.
CorrelationsThe correlations between the scales on the WPI and EQi are reported in Table 13. Significant correlations were found between most of the primary scales on the WPI and the scales of the EQi. As expected, some of the highest correlations included constructs such as Self-Control on the WPI with its conceptual counterpart of Impulse Control on the EQi, with a correlation of 0.52. Similarly, the Stress Tolerance subscales on both assessments correlated 0.59 with one another. Other large correlations of >0.5 were mostly between the WPI and composite EQi scales. For example, Energy, Initiative, Persistence and Stress Tolerance correlated with total EQ (r=0.53), (r=0.51), (r=0.53) and (r=0.57) respectively. Some other interesting correlations are briefly discussed in the following section.Concern for Others on the WPI correlated with the Interpersonal composite 0.57 on the EQi as expected. At the subscale level, Concern for Others also correlated with the Empathy 0.48, Social Responsibility 0.48 and Interpersonal Relationships 0.45, suggesting that individuals characterised by a concern for others, were likely to be empathic, have good relationships and would also have a tendency toward more pro-social behaviours. Although these scales appear to measure a very similar construct, the correlations suggest that there is some overlap however they are sufficiently different in what they measure and as such cannot be used interchangeably. General Mood also correlates with Concern for Others 0.40, suggesting that people with a higher Concern for Others are likely to be more optimistic and happy as well. Interestingly, the Dependability scale on the WPI correlates well with the Problem-Solving 0.40,General Mood 0.43 and Optimism 0.41 scales of the EQi, thereby linking a person’s ability to effectively deal with life’s challenges to an individual’s general well-being. The Mood scales on the EQi (General Mood, Optimism and Happiness) had good correlations with Dependability 0.43, Energy 0.44, Initiative 0.50, Persistence 0.46 and Stress Tolerance 0.44, suggesting that these are important aspects of personality related to an individuals’ level of general well-being at work.Thirteen of the EQi scales had correlations of >0.30 and higher with the Stress Tolerance scale of the WPI. In effect, this included Total EQ, the Intrapersonal (except Emotional Self Awareness), Adaptability, Stress Management, and General Mood composites as well as all of their subscales. The Interpersonal composite on the EQi was also significantly correlated to Stress Tolerance on the WPI however it did not display the same robust relationship as the other EQi scales.
Factor analysisIn order to determine whether any special factors could be identified from both the WPI and EQ-i, the scales from both assessments were included in a principal axis factor analysis with Oblimin rotation. Based on Kaiser’s criterion of eigenvalues-greater-than-unity, six factors were extracted that explained 62.10% of the variance in the correlation matrix. The pattern matrix is shown in Table 14.The first factor was easily identified as an emotional intelligence factor, as all of the scales of the EQ-i had salient primary loadings on this factor, apart from Empathy, Social Responsibility and Impulse Control. This is an important finding in that it shows that emotional intelligence remains a separate construct from personality, although there are relationships with many of the WPI scales.The second factor was made up of the Empathy, Social Responsibility and Interpersonal Relationship scales of the EQ-i and the Concern for Others scale from the WPI. All scales had salient loadings on this factor, although Social Responsibility had a secondary loading. These scales appear to represent the Big Five factor of Agreeableness. For the ease of reference, it was proposed that this factor be called Empathy.The third factor consisted of most of the WPI scales. Only Concern for Others, Democratic, Outgoing, Rule Following, Self Control, and Teamwork did not have salient loadings on this factor. It is proposed that this represents a cluster of what may be leadership characteristics, and was therefore tentatively called Leadership.The fourth factor was negatively correlated with the other factors, and consisted of the Impulse Control, Flexibility and Reality Testing scales of the EQ-i and Self Control and Stress Tolerance scales of the Adjustment global scale of the WPI. All of these scales had salient negative loadings on the factor, although the loadings of Flexibility, Reality Testing and Stress Tolerance were secondary loadings. The content of the factor appears to represent that of the Big Five factor of Neuroticism. However, if a person were to obtain high scores on each of these scales, they might be described as resilient or emotionally adjusted. It is proposed that this factor be named Adjustment.The Democratic, Outgoing and Teamwork scales of the WPI all had primary salient loadings on the fifth factor. These scales all form part of the global Social Orientation scale, although Concern for Others does not load on this factor. While it might be more appropriate to name this factor Working with Others, Social Orientation was retained as a name for this factor.The last factor was made up of the Attention to Detail and Rule Following scales of the Conscientiousness global scale of the WPI, which had salient negative loadings, and the Flexibility scale from the Achievement Orientation global scale, which had a salient positive loading. While it does appear to capture elements of Conscientiousness, it does not appear to represent the entire scale. It is suggested that it be called Diligence, characterised by high scores on Attention to Detail and Rule Following and low scores on Flexibility.
SUMMARYThe results presented in this Technical Report show good evidence for the construct validity of the Work Personality Index in the South African context. However, users of this instrument should use caution when interpreting scores, since only North American norms are presently available and the current study demonstrates that South Africans typically score higher on all of the WPI scales*. Scores may therefore seem artificially inflated, but this will not adversely influence relative comparisons between South African incumbents.By means of a Principal Components Analysis it was possible to replicate the factor structure reported in the manual. Four of the WPI scales did not have primary loadings on its expected components, but all had salient loadings on those components.Rasch analysis was conducted and the number of items that misfit in every scale was identified and Differential Item Functioning for men and woman was investigated. In addition, category probability curves were generated to examine the response styles on each of the scales of the WPI.With regard to differences between gender and age groups, no real differences in performance on the WPI between men and woman as well as between different age categories were found.Correlations between the EQi and WPI found good relationships between the scales as expected. The level of correlation between the two instruments was such that it demonstrated differentiated and independent construct measurement. The combined factor analysis also demonstrated that emotional intelligence remains distinct from personality. Certain areas of overlap were identified between the two assessments that could prove fruitful when interpreting both assessments together.*South African norms were developed and implemented in 2010
