What an HPI Percentile Compares, and What It Cannot Predict

What does an HPI percentile tell me at a glance?

An HPI percentile tells you where one scale score falls in the norm distribution used for that report. It does not mean you show a trait in that percentage of situations or have that percentage chance of succeeding at work. Those are different claims. First identify the score being ranked and the reference group. Then ask what evidence supports an interpretation about behavior or a work outcome. For example, an illustrative 75th-percentile Sociability result places the score above roughly three quarters of the reference distribution, subject to the scoring convention. It does not mean the person is sociable 75 percent of the time, is 25 percent more sociable than someone at the 50th percentile, or will excel in an interactive role. The number reports relative position, not a share of a trait or a forecast. Hogan’s “Personality Assessment FAQs” describes percentile comparisons against normative data. Its sample HPI report displays separate percentiles for individual scales, so the number belongs to a scale, not an overall ranking of the person. Use a two-stage reading rule: identify the scale and norm before interpreting the rank; treat a claim about likely behavior or job success as a second inference requiring its own evidence. Next ask who was included in the norm, how the score was produced, how much uncertainty surrounds it, and whether evidence supports the prediction. A percentile can begin reflection, but it cannot settle those questions or serve as a verdict about you.

Sources: Personality Assessment FAQs | Hogan Assessments; HPI Career Report UK English Sample

Who are the people behind the comparison?

The comparison group is the norm selected when the score was interpreted. A norm is a reference distribution used to give a score meaning by showing how it ranks among a specified group. It does not automatically mean people who share your job, age, location, language, or employer. Hogan’s public FAQ says the HPI and HDS are normed on samples of more than 100,000 working adults. It does not identify the norm used for a particular report. That distinction matters because “compared with working adults” and “compared with managers in this industry” are different questions. These answer different comparison questions. Hogan’s 2015 “FAQ Blog Series: Norms and Reliability” says general working-population norms are used for most selection and development applications, while some clients request supplemental subgroup norms, such as for managers in an industry, organization, or location. Such norms require a sufficiently large and diverse dataset, with requirements varying by group. This does not establish which norm your report used. A report’s percentile therefore cannot answer “How do I compare with people like me?” unless “like me” is made specific and the chosen norm actually matches that question. A broad norm suits a broad comparison; a local norm may suit a specific development conversation. Neither comparison by itself says what a person will do in a role; the norm identifies the reference for rank, while interpretation and prediction require separate reasoning. Hogan’s “Global Norm Update” page illustrates why readers should ask which reference was selected. It describes a global norm update scheduled for January 2019 and says that an administrator selects a norm before generating a participant report; the selected norm could be looked up in Hogan’s administrative portal. This historical page cannot identify the norm or scoring version on a current report, but shows that “the Hogan percentile” alone cannot reconstruct the comparison frame. If your report does not name its norm, ask the administrator or issuing provider: “Which norm set was used to calculate this percentile, and where is that documented?” Confirm the scale and date, too. Do not infer a local, age-matched, or occupation-matched comparison from the fact that you took the assessment at work. Until the reference group is identified, the percentile’s safest meaning is simply relative standing under the scoring used, with the comparison population still unknown.

Sources: Personality Assessment FAQs | Hogan Assessments; FAQ Blog Series: Norms and Reliability | Hogan Assessments; Global Norm Update | Hogan Assessments

What exactly is being ranked?

An HPI percentile belongs to a particular scale score, not to an overall ranking of your personality. The publisher’s sample report displays separate percentiles for seven primary scales, so read the scale label and its description before attaching meaning to the number. Hogan’s current HPI product page also describes seven primary scales, while the sample report shows how those separate results can appear together on a page.

A raw score is the score produced from a person’s responses under an assessment’s scoring method. A percentile then places that score in relation to scores in a reference distribution. If a report showed the 75th percentile on a named scale, the ordinary interpretation is that the score is at or above roughly three quarters of the reference distribution on that scale, subject to the test’s scoring convention and treatment of ties. That is a statement about relative position. It does not mean a person possesses 75 percent of a trait, shows a behavior in 75 percent of situations, or has 75 percent more of a quality than someone at the 50th percentile.

The distinction matters because the seven scale labels refer to different kinds of tendencies. In the sample report, Sociability and Prudence have separate percentile entries, as do Adjustment, Ambition, Interpersonal Sensitivity, Inquisitive, and Learning Approach. A relatively high rank on one scale cannot be rolled into a general claim that the person is “high” in personality overall; it says nothing by itself about rank on another scale. Its meaning depends on what that scale is intended to describe and on the report’s explanatory material. Hogan’s FAQ likewise says that there is no ideal HPI score or profile and that scores should be considered in light of job context and future goals.

The sample report is a dated UK English illustration, not a current universal scoring guide. It groups its example scores into low, average, and high bands and gives specific percentile boundaries for those bands. Those cut points belong to that example’s format; they should not be assumed to apply to every HPI report, scale, language, or scoring version. If your report uses bands, read its definitions and ask which scoring and norm version was used if unclear. First identify the scale, then the report version, then the comparison frame. That sequence keeps a percentile in its proper role: a relative position on one measure, not a complete description of a person.

Sources: HPI Career Report UK English Sample; Hogan Personality Inventory | Hogan Assessments; Personality Assessment FAQs | Hogan Assessments

What does the number not tell me about distance?

A percentile rank tells you the order of a score within a comparison distribution; it does not measure distance on an equal-step scale. The Hogan FAQ says HPI scores are reported in percentile format against normative data. In its sample HPI report, seven separate scales each have their own percentile, shown on a chart. The sample is a 2013 UK English illustration, so its values and low, average, and high bands should not be treated as current universal cutoffs. The general reading rule still holds: the percentile marks relative position on a particular scale, not an amount of a trait. The number answers “where in this norm distribution?” It does not answer “how much more?” or “how often?” Those questions require different evidence and should not be supplied by mental arithmetic on the rank.

That distinction matters when comparing ranks. A ten-point gap in one part of a distribution need not correspond to the same raw-score gap as ten points elsewhere. Raw scores are converted into ranks using the distribution of scores in the selected norm; where more scores cluster or spread out, equal percentile intervals can cover different raw-score intervals. Nor does moving from the 50th to the 75th percentile mean a person has 25 percent more of a characteristic. It means the score is at a different relative position in the relevant comparison distribution.

Keep two kinds of comparison separate. Two people can occupy different percentile positions without that rank gap measuring the size of a difference in their day-to-day tendencies. Retesting raises another question. A changed percentile might reflect a changed response or score, but the rank alone cannot show whether any underlying tendency changed proportionally. Without compatible report versions and evidence about score precision, small shifts do not warrant arithmetic stories about personal change. Read the scale label, the comparison norm, and the report conditions first; use the number as a location, not a measurement of distance or intensity.

Sources: HPI Career Report UK English Sample; Personality Assessment FAQs | Hogan Assessments

Why can two HPI reports be hard to compare?

A difference between two HPI percentiles is not automatically evidence that your personality changed. It may reflect changed answers, but it can also reflect a different response format, scoring method, norm, language, or report version. Before reading the numbers as a before-and-after record, establish whether both reports were produced on comparable terms. Hogan’s “Four-Point Assessments and Norm Upgrades” documents changes relevant to that comparison. The publisher says HPI and HDS items historically used True/False responses, then moved to four options in September 2018: Strongly Disagree, Disagree, Agree, and Strongly Agree. After four years with the expanded format, Hogan says it updated its global norm and all 39 local norms. It also introduced a scoring method intended to make main- and subscale scores more precise. The comparison has two moving parts. A scoring method turns responses into a scale score; a norm distribution locates that score relative to a reference group. If the norm changes, the same underlying score could occupy a different percentile because the comparison distribution changed. If response options or scoring changed, the scale score itself might also be produced differently. An apparent shift could therefore combine a person-side change with a measurement-system change. That is a reason to check comparability, not evidence that a particular score shifted because of an upgrade. Norm changes have affected interpretation before. Hogan’s “Global Norm Update” says a 2019 update replaced its prior global norm with one representing 171,132 working adults from 180 countries and 46 languages. Hogan reported that most scales shifted less than five percentile points and nearly 90 percent stayed in the same interpretation range. Those figures describe that 2019 update; they should not be generalized to 2023 or assumed to describe your reports. The page says administrators select a norm before generating a report and can look up which norm was used. Ask for that name rather than guessing. When scores differ, consider changed responses or circumstances, ordinary fluctuation, and changed measurement conditions. Hogan’s “Personality Assessment FAQs” says individual HPI and HDS scores may fluctuate slightly over time and characterizes significant shifts as rare. This general guidance does not explain any one pair of results. Do not assign a difference to personal development until conditions are checked. Ask the administrator or provider to confirm each assessment date, language and form, scale name, norm, and scoring version. Ask whether the earlier responses can be scored under the same method as the later report; do not assume this is available. If conditions match, the difference is more interpretable, though it still needs context. If conditions differ or cannot be established, describe the comparison as uncertain rather than calling it personal growth, decline, or a stable trait shift.

Sources: Four-Point Assessments and Norm Upgrades | Hogan Assessments; Global Norm Update | Hogan Assessments; Personality Assessment FAQs | Hogan Assessments

How much precision should I attach to the percentile?

Treat a reported percentile as an estimate from an assessment and scoring process, not an exact description of one person. Three questions are often collapsed: score consistency, support for an interpretation, and uncertainty in this result. None is answered by the percentile alone. If a narrow difference could change a consequential decision, ask for technical documentation. For ordinary reflection, keep the interpretation provisional and test it against observed situations.

Reliability concerns consistency; validity concerns whether evidence supports a specific interpretation or use. Hogan’s 2017 explainer, “A Quick and Dirty Guide to Validity and Reliability,” distinguishes consistency from whether an assessment measures what it is intended to measure. That distinction is useful, but its dated coefficients should not be treated as current properties of a report. A reliable measure can be unsuitable for the question, and evidence for one interpretation does not automatically support another. A percentile also depends on its norm distribution. These are separate parts of the argument, not interchangeable signs of accuracy.

Precision asks how narrowly a score locates someone’s standing, given the instrument, version, scoring method, and measurement-error information. Hogan’s “Personality Assessment FAQs” says HPI scores may fluctuate slightly over time and describes personality as relatively stable. That does not establish uncertainty around one person’s percentile or whether a small change matters. Ask whether technical materials for the relevant report provide a standard error, confidence interval, or other guidance; do not assume every report includes one. Without it, there is no basis for drawing a precise boundary around the result.

Version matters when judging change. In “Four-Point Assessments and Norm Upgrades,” Hogan describes moving HPI and HDS responses from True/False to four options in 2018, updating its global and 39 local norms, and implementing a new scoring method in the 2023 upgrade. This account establishes changed scoring conditions, but not the method used for a reader’s report or an individual effect. A modest percentile difference could reflect response fluctuation, a changed norm or scoring process, or both. Do not call it meaningful personal change until comparability and uncertainty are established.

When a close distinction affects an important conclusion, ask which report version and norm were used and what evidence describes its precision. Then ask whether the interpretation is supported for its intended purpose. If details are unavailable, describe the result as a tentative relative pattern, not an exact personal boundary. The percentile remains useful without carrying more certainty than the evidence supports.

Sources: A Quick and Dirty Guide to Validity and Reliability | Hogan Assessments; Personality Assessment FAQs | Hogan Assessments; Four-Point Assessments and Norm Upgrades | Hogan Assessments

Why is a work prediction a different claim?

A percentile and a prediction answer different questions. A percentile locates a score in a reference distribution: it tells you how that score ranks among the people represented by the norm. A prediction asks whether scores are related to a defined outcome in a particular setting. That outcome might be a specified measure of job performance, but the claim needs evidence connecting the assessment to that criterion. It does not give one person's odds of performing well or advancing. This distinction does not mean personality measures can never contribute to work-related prediction. It means that the evidence must match the claim. The 1991 meta-analysis “Personality Measures as Predictors of Job Performance: A Meta-Analytic Review” reviewed 494 studies and found usable results in 97 independent samples, comprising 13,521 people. Across the measures and criteria in those studies, the authors reported corrected mean validity of .29 for confirmatory strategies, .12 for exploratory strategies, and .38 when studies explicitly used job analysis to select measures. The pattern suggests that job-related selection of measures matters. Job analysis is a structured account of a role’s tasks and demands; it gives researchers a basis for deciding which measured characteristics might be relevant, rather than assuming relevance from a broad label. Those figures do not tell a reader what their HPI percentile predicts. The review combined different personality measures, samples, jobs, and performance criteria. It was published in 1991, and its reported corrected mean validity values summarize relationships across studies; they are not percentages of accuracy, odds of success, or forecasts for one person. Nor does evidence that some measure-and-job combinations show a relationship establish that every HPI scale is useful for every role. An individual forecast needs evidence specific to the instrument, population, role, outcome, and uncertainty. A concrete distinction helps. “This score is at the 75th percentile in the norm used for this report” is a statement about relative standing. “People with this score have a given chance of meeting this role’s performance standard” is a probability claim. The second statement needs a defined performance standard and data showing how scores relate to that standard in an appropriate setting. That relationship is what criterion-related evidence addresses: it tests whether scores vary with a specified outcome, not whether a rank is itself an outcome. The percentile alone supplies neither. A publisher’s general description of an assessment’s work relevance cannot fill those missing links for a particular employer or person. Hogan’s FAQ says there is no ideal HPI score or profile and advises interpreting results in relation to job context and goals. That guidance is consistent with the distinction, but it is not itself evidence that a particular score predicts a particular outcome. The careful conclusion is therefore neither “personality scores predict careers” nor “personality scores tell us nothing about work.” Some personality measures may contribute to group-level prediction when selected and evaluated for a relevant purpose. Whether that evidence supports a specific employment decision is a separate question, and a norm percentile by itself cannot answer it.

Sources: Personality Measures as Predictors of Job Performance: A Meta-Analytic Review | Personnel Psychology / Wiley; Personality Assessment FAQs | Hogan Assessments; A Quick and Dirty Guide to Validity and Reliability | Hogan Assessments

What evidence would make a work claim more defensible?

A work claim becomes more defensible when it names the intended decision, role, personality scale, people and setting studied, and outcome the score is meant to relate to. It should explain uncertainty and how the result is weighed alongside other relevant evidence. A percentile answers a narrower question: where a score sits in a chosen norm distribution. It cannot establish that a particular person should be hired, promoted, or steered toward a career. Hogan’s description of its assessments as useful for work does not substitute for evidence about a specific use.

The first link to examine is between the role and the proposed interpretation. Job analysis means examining important tasks and work behaviors, including their relative importance, so an assessment claim can be judged against what the role requires. If a scale is said to matter because a role involves frequent coordination, the argument should explain how it relates to that requirement and what outcome counts as effective coordination. A broad label such as “leadership potential” can conceal claims about communication, judgment, delivery, or advancement. These outcomes are not interchangeable; evidence for one does not automatically establish another.

The 1991 meta-analysis, “Personality Measures as Predictors of Job Performance: A Meta-Analytic Review,” is counterevidence to the claim that personality measures can never inform work outcomes. It reviewed 494 studies and found usable results from 97 independent samples, totaling 13,521 people. Reported corrected mean validity was .29 for confirmatory strategies, .12 for exploratory strategies, and .38 where job analysis guided measure selection. These aggregated findings cover varied personality measures and performance criteria; they are not HPI-specific, individual probabilities, or guarantees for another role. The pattern supports a narrower lesson: choosing a measure in relation to the work matters.

For US employment selection, the EEOC-hosted Uniform Guidelines on Employee Selection Procedures say validity studies should review job information and include job analysis, subject to a stated exception. They also caution that content validity alone is not appropriate for procedures based on inferred personality traits. Content validity means showing that a procedure samples work behavior or products; a personality questionnaire does not become a job sample because its wording sounds relevant. This guidance is US-specific, not legal advice everywhere. It does not establish whether a particular employer has suitable evidence; it shows which questions matter.

Ask an employer or provider: what scale-to-outcome evidence supports this use for this role, and how is the result combined with other job-relevant information? Ask what performance criterion was used, how the study population and setting compare with the role, and what uncertainty or fairness checks were considered. Evidence from another occupation, measure, or outcome may inform the question without settling it. For coaching or private development, a report can prompt discussion of observed habits without proving they cause success. The evidence standard rises when a score can determine access to work.

Sources: Uniform Guidelines on Employee Selection Procedures | U.S. Equal Employment Opportunity Commission; Personality Measures as Predictors of Job Performance: A Meta-Analytic Review | Personnel Psychology / Wiley; Personality Assessment FAQs | Hogan Assessments

Why is high or low not the same as good or bad?

A high or low HPI percentile is not a universal judgment of quality. Hogan’s “Personality Assessment FAQs” says there is no ideal HPI score or profile: a result may point to strengths and challenges, and should be interpreted in light of a person’s job context and goals. The useful question is not whether a score is good, but what tendency it may describe, in which situation, and with what observable consequences. Sociability offers a concrete illustration. Hogan’s “Hogan Personality Inventory” page describes one end with phrases such as “works independently” and “quiet,” and the other with “outgoing” and “talkative.” These are descriptions of tendencies, not a transcript of what someone does in every meeting. A lower relative standing does not establish that a person avoids colleagues, just as a higher standing does not show they always seek interaction or perform better in a role requiring frequent conversation. The sample report, “HPI Career Report UK English Sample,” says characteristics important for success in one job may interfere with performance in another. This example is dated 2013, so its score bands and wording should not be treated as current universal guidance. A pattern may have different implications depending on what work asks someone to do. To use a scale description responsibly, turn it into a question that can be checked. If a report suggests a preference for less social stimulation, ask when focused independent work helps and when a task calls for initiating contact or speaking up. Look across repeated situations, not one memorable meeting. Compare examples that fit the description with those that complicate it, and consider feedback from people who have seen the relevant work. The report can prompt observation; it cannot settle the interpretation by itself. This avoids treating a scale as identity or diagnosis. Someone may be reserved in a large group and active in a familiar project team; neither example alone cancels the other. What matters is the recurring pattern across settings. Translate “high” or “low” into a tentative description, then test it against behavior and context. Keep supporting and contrary examples. That gives a clearer account of when a tendency helps, creates friction, and what evidence would change the view.

Sources: Personality Assessment FAQs | Hogan Assessments; Hogan Personality Inventory | Hogan Assessments; HPI Career Report UK English Sample

What is a responsible next step with my result?

Record the scale name, percentile, assessment date, language or form, and norm identified in your report. If the comparison group or scoring version is missing, ask the issuer for those details and any documentation explaining the score. Hogan's “Personality Assessment FAQs” describes norms broadly, but cannot identify the reference distribution used for your individual result. Until that is clear, avoid comparing reports as though their conditions were identical. Turn one report sentence into a question you can check. If it describes a tendency around interaction, ask when that pattern appears in your work and what examples or feedback support or complicate it. Treat the interpretation as a reflection prompt, not a fixed identity. For hiring or another consequential employment outcome, ask what evidence connects this assessment, the role, and the outcome being judged. A percentile alone does not answer that question. If you want to consider how several work tendencies fit together, the site's Work Pattern Report at /assessment offers low-stakes self-reflection across ten work continuums. It has no norms, cutoff, hiring score, or job recommendation, and does not validate or interpret a Hogan result. Decide whether your report's norm and scoring version are known and whether evidence fits the intended use. If either is unclear, seek that information before drawing a stronger conclusion.

Sources: Personality Assessment FAQs | Hogan Assessments; Work Pattern Report | Personality Report

Questions readers ask