When Does an HPI Report Use Percentiles Instead of Raw Counts?

When does an HPI report use percentiles rather than raw subscale counts?

Use a percentile interpretation when an HPI report labels a primary-scale result as a percentile and the report or matching documentation identifies its comparison group. A raw subscale count tallies keyed item scores; alone, it does not show relative standing. Hogan’s 2013 Interpreting HPI Subscales guide says primary scales are often presented as percentiles and subscales as raw scores. It describes an older form, so it does not establish the format of every later report. Hogan’s later reporting history also includes four-band subscale bricks, a distinct display. Check the score label, report legend, and version-specific guide before deciding what a number or bar represents. A percentile addresses relative standing under particular norms; a raw count addresses scoring. Neither label alone establishes whether an interpretation suits a specific decision. The National Council on Measurement in Education’s glossary defines norm-referenced interpretation as comparison with a specified reference population. Thus, a count and a percentile are different units, even when they appear together. Do not compare their printed sizes or infer a conversion from the HPI name. First identify whether each result is a primary scale or subscale, how it is displayed, and which norm and report edition apply.

Sources: Interpreting HPI Subscales; NCME Glossary

What does each format compare, and what does it count?

A percentile, a raw subscale count, and a four-brick display answer different questions. A percentile rank places a score within a specified distribution of other scores. A raw count totals the item points assigned under the scoring key. A brick groups a subscale result into one of four norm-informed bands. Because each uses a different scale, the printed values cannot be compared by size alone. Keep the score label, level, and reference group attached to a result. The National Council on Measurement in Education (NCME) defines a percentile rank as the percentage of scores in a specified distribution that fall below a given score. Its definition makes the reference distribution essential: “percentile” does not name a fixed number of points that means the same thing in every assessment or norm group. NCME defines a norm-referenced interpretation as comparing a test taker’s performance with the score distribution of a specified reference population. Norms are statistics or tables that summarize scores for stated groups. Thus, a percentile is a relative location, not a count of the items a person endorsed. A simple illustration can separate the units without implying any actual HPI result. Imagine a report that shows a percentile for a primary scale and a raw count for one subscale. The percentile answers, in effect, “Where does this score sit in the stated comparison distribution?” The count answers, “How many keyed score points contributed to this subscale total?” One is a position among scores; the other is an accumulation from scored responses. A percentile is not the same as that percentage of items endorsed. Nor does a percentile, by itself, state the probability of an event, how often a behavior will occur, or whether a result is desirable. Those meanings require additional evidence and a defined interpretation. NCME’s glossary defines a raw score more generally as a sum or other combination of item scores; counting correct answers is one common example, not the only possible scoring rule. For HPI interpretation, Hogan’s 2013 *Interpreting HPI Subscales* guide describes raw subscale scores as a way to add detail beneath the primary scales. That wording matters: “raw” does not mean unscored or self-explanatory. Items may have keyed values, and the total has meaning only under the instrument’s scoring rules. A count alone does not disclose those rules. Even if a report presents the count as a fraction of possible item points, that fraction describes the share of available points, not standing among a comparison group. A four-brick display is a third representation. In Hogan’s 2016 *Subscale FAQ*, the company describes four categories based on aligning raw subscale scores with cumulative frequencies in working-population data. The categories summarize relative bands, but do not provide the same detail as a full percentile rank. A brick is not an exact percentile, and it does not reveal the original raw count. Moving from a count to a broad band compresses information; reading the band cannot reverse that compression and recover the underlying tally. These distinctions prevent a common reading error: treating all figures on one page as if they shared a ruler. A larger raw count is not automatically a higher percentile, and a brick number is not a score-point total. To compare two results, first ask whether they concern the same score level, then identify each representation and its relevant reference group. Unless the report’s matching documentation supplies a conversion, arithmetic between a count, percentile, and band has no interpretable meaning.

Sources: NCME Glossary; Interpreting HPI Subscales

What did Hogan mean by percentiles and raw counts in the older HPI format?

Hogan’s 2013 guide describes one older HPI format: a 206-item questionnaire with True/False responses. In that guide, primary-scale results are “often” presented as percentiles relative to the general workforce, while subscales may be reported as raw scores. That distinction answers the question for the format the document describes, but the qualifier matters. It is a description of common reporting practice in that guide, not a rule that identifies every HPI report, including reports generated under later response or scoring systems. The guide’s title, Interpreting HPI Subscales, signals its focus: using narrower score details to add context to the broader primary scales. A primary scale is the broader result in this reporting structure; a subscale is a more specific component associated with that scale. Hogan’s guide presents subscales as interpretive detail that can help a reader consider what contributes to a broader scale result. It does not make each subscale a separate verdict about a person. A percentile expresses relative position against the comparison group used for the report. A raw subscale score is a count of keyed item points under the scoring method described for that form. A count can be a score without being a ranking among people. It is also not a percentage of a workforce or a probability of a future outcome; those interpretations require information beyond the tally itself. To interpret either number, retain its label and the documentation that defines it. The guide also cautions readers about how much meaning to assign to a small difference between subscale scores. Narrower scores contain less information than the broader scale interpretation, so a small shift should not carry disproportionate weight. Two nearby raw counts do not, by themselves, demonstrate a meaningful difference in a person’s everyday behavior. Nor does a subscale count establish why a response pattern occurred. The guide supplies interpretive guidance for its measure; it is not evidence that every small numerical gap maps to a stable or consequential difference. Labels require care, too. The guide notes that some subscale names are negatively worded. The scoring key and the guide’s explanation, rather than the label alone, determine how the result is interpreted. It would be unsafe to infer direction from an ordinary-language reading of a subscale name or to assume that a larger raw count always means “more” of a positively phrased trait. For someone reading a legacy report, this document is useful only if the report matches the described edition and response format. Check the report date, item format, scale labels, and any legend before applying its explanation. If those details are missing, ask the administrator or provider for the guide that matches that report and for clarification of whether a displayed subscale figure is a raw score. The older guide supports a bounded conclusion: in the 206-item True/False format it describes, primary scales were commonly shown as workforce-referenced percentiles and subscales could appear as raw scores for added interpretation. It does not establish that this remains the display in every later HPI report.

Sources: Interpreting HPI Subscales

When did Hogan replace some raw subscale displays with bricks?

Hogan’s 2016 Subscale FAQ says the brick display began on reports generated starting May 2, 2016. For the HPI, it names three report types: Potential, Insight, and Flash. The change replaced raw subscale scores on those reports with a visual bar of up to four bricks. The date and scope matter: this describes a particular update, not a rule inferable from the HPI name alone. If a report has bars, check its title and generation date against the FAQ’s coverage. The FAQ describes each HPI brick as a category tied to a quartile of scores in a working population. Its four labels are Low, Below Average, Above Average, and High, with ranges of 1–25%, 26–50%, 51–75%, and 76–100%. Hogan says it computed cumulative frequencies for raw scores on each subscale, compared them with global working-population data, then aligned the resulting percentiles to one of four brick values. The Subscale Interpretive Guide describes the bars as showing the quartile in which a subscale score falls. The bar retains a relative category while replacing the visible item-score tally. That grouping changes what a reader can know. A raw count or fraction preserves a tally of keyed points; four bricks compress that tally into one of four norm-informed categories. A bar can make the broad comparison easier to see, but it does not identify the exact percentile within the band, the original count, or the distance between people in the same band. The FAQ says raw scores are aligned to a brick but supplies no reader-facing conversion table. The guide says conversions are not exactly linear: the relationship depends on each subscale’s score distribution. Its illustration shows that low raw values can map to one category when uncommon in the global working population. A brick therefore cannot be translated back into one unique count from the public display alone. The four labels are categories of relative position, not quantities of item endorsement. There is a practical distinction between zero bricks and a missing result. The FAQ says that if no items are endorsed on an HPI subscale, the display shows zero bricks. If too few items are answered to generate a score, it shows a dashed line. The guide makes the same distinction between a minimum score and a subscale that cannot be scored. A dashed mark should not be read as low standing, and zero bricks should not be confused with absent data. If unclear, ask the report administrator what the mark means under that report’s documentation. Hogan presents the format as a way to make subscale comparisons faster and more consistent, while saying interpretation should remain primarily at the primary-scale level, with bricks adding context. These are the provider’s rationale and guidance, not an independent evaluation that the format improves accuracy for every reader or use. A brick makes a broad norm-based category more visible than a raw tally, while sacrificing detail about the underlying score. That tradeoff can help orient a reader, but bars alone cannot support fine-grained comparisons. A category is not a precise percentile or evidence of a work outcome. The public documents support applying the 2016 description to the Potential, Insight, and Flash HPI reports named there, generated from the stated date. They do not establish that every later report uses an identical display or unchanged reference data. For a report with bricks, verify its type, date, and edition before applying those labels. If details do not match, ask the administrator for the guide accompanying that report and how its bars are defined. Until then, do not guess the raw count or treat a category as an exact percentile.

Sources: Subscale FAQ; Subscale Interpretive Guide

How should a reader handle the later response and scoring update?

Hogan's “Four-Point Assessments and Norm Upgrades” says HPI responses historically used True/False and that in September 2018 the choices expanded to Strongly Disagree, Disagree, Agree, and Strongly Agree. A guide for the binary form cannot establish how a later form is scored or displayed. The change does not show that every later report replaced raw subscale counts with one particular format. Hogan says that after four years using the expanded choices, it updated its global normative sample and all 39 local norms. The page describes the 2023 global sample as based on Hogan's latest data and intended to better represent work-related demographics, including age, gender, job family, industry, and assessment language. These are provider statements about the update and its aims. They do not identify the norm used for an individual result or show that every report uses the same comparison group. “2023 norms” is a question to ask, not a complete description of a report. The page says Hogan implemented a new scoring method to increase measurement precision in main and subscale scores. Hogan says finer distinctions should make results more “on target,” particularly at the subscale level. The page does not provide report-by-report specifications, a conversion table between old and new results, or a way to calculate how one person's score changed. More fine-grained scoring also does not establish whether a report displays a raw count, percentile, or category. Scoring method and display are related, but not interchangeable. The 2016 brick FAQ names specific report types and a start date for that display; the later update describes response, norm, and scoring changes. Neither settles the display in every subsequent report. The opened pages do not support inferring that the new system preserved earlier displays or replaced them all. Nor does the response shift prove older responses were rescored or legacy reports reissued. The public update leaves these details open. For a report, ask its administrator or provider for its name and generation date, whether the form used two or four response choices, whether the figure is a main-scale or subscale result, and what the score represents. Request the applicable scoring guide and norm edition or reference population. Before comparing with an older HPI report, ask whether response formats, scoring rules, and norms are compatible and whether Hogan documents a valid comparison. Do not reverse-convert a raw count, brick, or percentile from the upgrade summary. Hogan's “Personality Assessment FAQs” says HPI and HDS percentile norms use samples of more than 100,000 working adults. This gives broad context about sample size, but does not identify the edition or comparison group behind one result. The FAQ also says scores should be interpreted in light of the person's current job and future goals, and that there is no ideal HPI score or profile. That supports resisting a universal “good score” reading; it does not answer edition-specific questions. Keep five details attached to any HPI number: report name and date, scale level, score representation, scoring edition, and norm reference. If one is missing, pause the comparison and seek matching documentation. Hogan's update establishes that response options, norms, and scoring changed; older guides remain relevant to the forms and reports they describe. Without report-specific documentation, the format cannot be inferred from the HPI name alone.

Sources: Four-Point Assessments and Norm Upgrades; Personality Assessment FAQs

Why is the reference population part of the score?

A percentile is meaningful only in relation to the group whose score distribution supplies the comparison. The National Council on Measurement in Education’s glossary defines a norm-referenced interpretation as comparing a test taker’s performance with the distribution in a specified reference population; it defines a percentile rank as the percentage of scores in that distribution below a given score. That is why the norm group belongs beside the score when someone explains it. If the reference population changes, the distribution used for comparison changes too. The same underlying result can therefore occupy a different relative position in two groups. This is a general consequence of comparing a score with distributions, not a claim that any particular HPI score would change by a known amount. A percentile from one norm cannot simply be carried over to another, and the percentile alone does not reveal whether its reference group matches the question a reader wants to answer. Hogan’s public “Personality Assessment FAQs” says that the HPI and Hogan Development Survey are normed on samples of more than 100,000 working adults. That gives a broad description of the populations behind the norms, but the FAQ does not identify the norm edition applied to every individual report. The stated size also does not, on its own, show how a sample is distributed across occupations, locations, languages, or other characteristics relevant to a particular comparison. A dated example makes the version issue concrete. In the 2016 “Validity of the Hogan Personality Inventory, Hogan Development Survey, and the Motives, Values, Preferences Inventory for Selecting Sales Representatives at ABC Company” technical report, Hogan describes a historical HPI global norm containing 144,877 working-adult cases from multiple countries, industries, organizations, and jobs. The same report says clients could use a local norm when a single-language norm was available and applicants were likely to come from a concentrated geographic area and be assessed in that language. The FAQ’s figure of more than 100,000 and the technical report’s 144,877 are not necessarily competing counts. They refer to differently dated provider materials, and the FAQ gives a broad threshold while the report records a specific historical dataset. Neither should be silently substituted for the norm named on a reader’s own report. Hogan’s FAQ also acknowledges small differences across languages and cautions against strong assertions when comparing results across translations. That makes the report language relevant metadata, though these public statements do not establish that a particular translation is unfair or unsuitable. For a reader deciding what a percentile means, the useful question is not simply “How many people were in the norm?” Ask which reference population and norm version were used, whether the norm is global or local, and whether its language and context fit the comparison being made. If the report leaves those details unclear, request them from its administrator or provider before comparing the percentile with another report or treating it as a meaningful difference. Sample size is one part of norm documentation; it cannot replace that identification, establish individual precision, or demonstrate that a score supports a separate decision.

Sources: Personality Assessment FAQs; Sample Technical Report, 2016; NCME Glossary

What does the score not establish about a person or a work decision?

A percentile or subscale brick describes a position produced by a scoring and norming system. On its own, it does not establish how someone will behave in a particular event, whether a role suits them, what their future performance will be, or whether they should be hired. Those are separate claims and require evidence relevant to the specific interpretation, population, and decision. A reader may mistake a relative rank for a direct measure of capability, or assume that a detailed subscale makes an interpretation more certain. Neither follows from the display alone. The 2013 Interpreting HPI Subscales guide says primary scales are often presented as percentiles and raw subscale scores can add detail. It cautions against overinterpreting small subscale shifts. Detail can suggest a question, not verify an answer. Validity concerns the evidence supporting a particular interpretation or use of a score. It is not a permanent stamp attached to an instrument, and reliability alone does not prove an inference is accurate for every purpose. Ask whether evidence supports this conclusion for this population and decision. Evidence supporting a broad work-style discussion, for example, would not automatically establish that a score predicts success in a specific role or should determine selection. That additional claim needs evidence matched to it. The American Psychological Association’s Professional Practice Guidelines for Occupationally Mandated Psychological Evaluations address clinical evaluations required for employment, licensure, or related occupational purposes. In that setting, the guidelines say psychologists should choose tools validated for a population appropriate to the evaluation and use multiple sources of relevant, reliable information. This guidance addresses that setting, not HPI specifically or every coaching conversation. Occupational conclusions should fit the evidence, population, and referral question. Hogan’s Personality Assessment FAQs say HPI scores should be interpreted in light of job context and goals, and that there is no ideal profile. The same FAQ describes Hogan assessments as used in selection and development and reports HPI and HDS norms based on samples of over 100,000 working adults. These are provider statements. A large norm sample supports comparison to the population and edition it represents; size alone does not validate every interpretation, language, role, or decision. Hogan’s FAQ also summarizes international reliability and validity work, but does not identify evidence for every possible inference. It cannot by itself answer whether one percentile warrants a hiring recommendation. The absence of an ideal profile also does not establish that every profile fits every role equally well. A subscale result can instead prompt a grounded discussion. If a report describes a tendency related to planning, a coach might ask what happens when priorities change and compare a recent example with the person’s account. The discussion may clarify a pattern or a mismatch between report language and experience. The report alone cannot decide which explanation is right. This is an illustrative use, not a claim about an actual respondent or result. For a consequential decision, ask what exact conclusion is being drawn and what evidence supports it for the relevant population and setting. Combine it with pertinent information and keep its role clear. A percentile is not a hiring score simply because it is comparative; a brick is not a job recommendation because it condenses a subscale. Separate the result’s format from the validity of its proposed use. If use-specific support is unclear, treat the result as a limited prompt for inquiry, not a decision about the person.

Sources: Professional Practice Guidelines for Occupationally Mandated Psychological Evaluations; Personality Assessment FAQs; Interpreting HPI Subscales

What should you ask before comparing or acting on an HPI result?

Before comparing or acting on an HPI result, ask for the report name and generation date, whether the figure is a primary scale or subscale, how it is displayed, which norm and scoring edition apply, and the matching guide. Then ask whether evidence supports the comparison or decision you have in mind. A number without its level, unit, and reference frame is not yet interpretable. The HPI name alone does not settle these details: Hogan’s public materials describe multiple formats and later changes to response options, norms, and scoring.

If a primary-scale result is labeled a percentile, ask which norm group it uses and whether the report or guide identifies it. Hogan’s Personality Assessment FAQs says HPI and HDS are normed on samples of more than 100,000 working adults, but this broad statement does not identify the norm edition behind a particular report. Request the specific norm reference; a large provider-reported sample does not answer every question about comparison. If the number is a subscale count, ask for the scoring key and edition that explain how responses produced it. A raw tally is not a percentile, and its size cannot be compared directly with a percentile rank.

A bar or brick display needs a separate check. Hogan’s 2016 Subscale FAQ says four HPI subscale bricks appeared on Potential, Insight, and Flash reports generated from May 2, 2016, describing them as categories aligned to cumulative frequencies in working-population data. Ask whether the report type and date match that scope and request the guide for its labels. The FAQ says a dashed line means too few items were answered to generate a score, not simply a low category; it gives zero bricks a different meaning. These points clarify that display, but the historical FAQ does not establish how every later report was scored.

For a report generated after later changes, ask which response format and scoring version produced it, whether the 2023 norm and scoring upgrade applies, and whether an earlier result was rescored. Four-Point Assessments and Norm Upgrades dates the change from True/False to four response options to September 2018 and describes a later update to global and local norms and scoring. It confirms change, but provides no universal conversion between editions or display rule for every report. An administrator or provider can identify metadata and supply the guide. A qualified assessment professional or coach can discuss a documented result in context, while separating interpretation from proof about a person or outcome.

For example, someone comparing a percentile in one HPI report with a subscale bar in another could ask: “What are the report names and dates, which scale level does each result describe, what norm and scoring version were used, and which guide explains each display?” If the documents do not resolve this, pause instead of converting a count, guessing the norm, or treating a category as an exact percentile. Once details match, compare like levels and representations, then ask whether evidence supports the intended use. Hogan’s FAQ says there is no ideal HPI score or profile and recommends considering job context and goals. That is context for discussion, not evidence that a score selects someone for a role.

Sources: Subscale FAQ; Four-Point Assessments and Norm Upgrades; Personality Assessment FAQs

What is the next useful step if the report still leaves the question open?

Match the number to the document that produced it. Ask the report administrator or provider for the report name and date, scoring edition, norm group and language, and guide defining the displayed score. If it is a percentile, ask which reference distribution it uses; if it is a raw count or brick, ask for the version-specific explanation. Hogan’s “Four-Point Assessments and Norm Upgrades” says response options expanded in 2018 and describes a 2023 update to norms and scoring. It provides no universal conversion between older counts, bricks, and later results, so do not convert without matching documentation. If the report remains unclear, keep the conclusion narrow: the display supports only the interpretation its label and guide justify. A coach or qualified assessment professional may help connect that interpretation to observed examples, but the score alone cannot settle what a work situation means. A separate question may be how your work patterns combine when you decide, collaborate, or handle change. The publication’s Work Pattern Report offers ten continuums for low-stakes self-reflection. It has no norms, cutoffs, type, or selection score, and does not interpret or convert HPI results. Use it only if that distinct reflection would help; otherwise, ask the administrator your specific documentation question or continue with the guides at /topics.

Sources: Four-Point Assessments and Norm Upgrades

Questions readers ask