Can Two Personality Test Scores Share a Scale?

Can scores from two personality tests be placed on the same scale?

Usually not from the reports alone. Two personality tests may print scores on the same 1–5 range, use the same trait label, or place a person at the same percentile, yet those similarities do not show that the numbers share units or meaning. A numerical comparison is defensible only when evidence for those specific instruments supports the intended link for the relevant population and use, with uncertainty made clear. Without that evidence, keep the scores on their original scales and compare what each report says it measures. This question concerns a reader looking at two results from different personality instruments and judging apparent agreement or conflict. Distinguish reading reports side by side from converting one result into the other's terms. In a side-by-side reading, you can notice that both reports discuss a similar tendency, then inspect how each defines it. A conversion makes a stronger claim: that a value on one instrument corresponds to a value on the other while preserving the interpretation you want. Matching formats do not establish that relationship. The chapter on score interpretation in *Educational Measurement* (5th edition) describes reported scores as products of a testing program, shaped by how a test is developed, administered, scored, and intended to be used. A printed number is not a free-standing measure whose meaning travels automatically to another report. *ETS R&D Connections No. 6* distinguishes forms of score linking, including concordance and equating; equating aims at interchangeable scores, a more demanding claim than showing that results are related. These are general measurement distinctions, not validation of a personality-test pair. The rule is simple: do not average, subtract, or normalize two results because their labels or ranges look alike. Ask whether the reports provide evidence for a shared interpretation. If they do not, retain each score in its own context and compare the constructs and descriptions in words. The useful question is not whether arithmetic can align the numbers, but whether evidence supports treating the aligned values as comparable.

Sources: Scaling, Equating, and Linking; The Practice of Comparing Scores on Different Tests

What does ‘the same scale’ mean here?

A shared scale means more than putting two results on a line with matching numbers. It means there is a defined relationship between the scores that preserves the interpretation a reader wants to make. Two reports may use the same range or label without showing that a value represents the same standing on both instruments. Before comparing the numbers, ask what kind of sameness is being claimed. A raw score is the result of applying an instrument’s scoring rule to a person’s responses, often by summing or weighting item responses. Its size depends on that rule: two totals can differ because the tests use different numbers of items, response options, or weights. A standardized score is a transformed score, often expressed relative to a distribution or reference group. The label alone does not identify the transformation or reference. A percentile rank gives a person’s position within a specified group: the 70th percentile means the score is at or above roughly 70 percent of that group’s scores, not that the person has “70 percent” of a trait. These quantities answer different questions. A scale score is the testing program’s reported value, produced through its choices about the test, scoring, administration and intended interpretation. In “Scaling, Equating, and Linking,” the Oxford Academic chapter describes scale scores as products of a testing program and separates establishing a scale from maintaining or relating it to another program’s scale. A scale built for one instrument does not itself bridge to another publisher’s scores. A crosswalk is a documented mapping between scores on two scales. It might estimate which score on one test tends to correspond to a score or range on another in a studied group. But “corresponds” is not automatically “means the same thing.” The ETS explainer “The Practice of Comparing Scores on Different Tests” distinguishes several kinds of linking, including concordance for related but distinct tests and equating, whose stronger goal is interchangeable scores. The explainer concerns educational tests, so its categories clarify the strength of claims; they do not validate any personality-test conversion. Think of comparability as a ladder. At the first rung, two reports offer related descriptions. At the second, scores may be associated or tend to rank people similarly. At the third, an empirical concordance maps scores for a defined set of tests and population. At the fourth, evidence supports treating the scores as interchangeable for the stated purpose. A relationship or mapping can be useful without making scores interchangeable; arithmetic alone cannot establish the top rung. When a report says two scores are “aligned,” ask which rung that word means, for which instruments and people, and what interpretation the link supports.

Sources: Scaling, Equating, and Linking; The Practice of Comparing Scores on Different Tests

Why the same trait name may hide different measurements

A shared trait label can locate related patterns, but it does not show that two scores measure the same quantity. Tests turn abstract ideas into questions, response options, and scoring rules. Two reports may use a broad term such as conscientiousness while emphasizing different parts of it. The label starts a comparison; it does not establish that one number can be translated into the other. The meta-analytic investigation “How similar are personality scales of the ‘same’ construct?” examined commonly used personality measures in organizational contexts. The study reports that scales treated as measuring the same construct were only moderately correlated in many cases. The discussion identifies differences in definitions and measurement, including broad factors versus narrower facets, as possible reasons. It offers no universal estimate or conversion table for an unnamed pair. A broad domain can contain narrower facets, meaning distinguishable aspects of a larger trait. Tests may select different facets, use different item wording, or give those facets different weight in a composite score. Item wording also shapes which situations a respondent considers, while response options and scoring rules determine how answers combine. A report may present the resulting total as one broad trait even though its components differ from another report’s. Unless the reports explain those choices, the label alone cannot reveal whether their totals emphasize the same pattern. The same label can therefore conceal differences in the questions asked and the response patterns that contributed to the result. The study discusses this issue even within familiar broad trait groupings. Accounts of openness, for example, have emphasized different facets, such as aesthetic or creative interests on one hand and ideas, curiosity, or intellectual complexity on the other. Taxonomies differ, and some facets may overlap with other broad traits. It means the operational definition, or the way a particular measure represents a construct, affects what its score can mean. A score that gives more weight to one facet may not stand in for a score that emphasizes another. Consider an illustrative, invented case. Two reports both describe “conscientiousness.” One description might emphasize orderly routines and attention to detail; another might place more emphasis on follow-through and advance planning. Both may sound related to conscientious behavior while pointing toward different situations. The first might prompt a question about keeping materials or steps organized. The second might prompt a question about carrying tasks through after priorities change. These are illustrative prompts, not test findings. Without instrument documentation, the reader cannot know whether either report uses these exact facets or how much each affects its score. This distinction matters when results conflict. Opposing results could reflect different item coverage, facet weights, scoring rules, reference groups, or uncertainty.It cannot identify which explanation applies to a person or named pair. It does show why a shared construct label alone cannot settle the question. Calling one test wrong would require evidence about those instruments and the intended interpretation; forcing the numbers together would add an unsupported assumption. Moderate correlation needs careful interpretation. It does not mean either scale is useless, and it does not establish that the measures are unrelated. Correlation describes how scores vary together in the studied data; it does not show matching units or allow substitution of one person’s result.The study’s organizational focus also limits its reach: its findings concern the measures, samples, and evidence included in that research project, not every current personality assessment or every way a report might define a trait. Its practical contribution is that a shared name may not justify conversion. When two results differ, compare their definitions and facet coverage before choosing a winner. Check the construct description, facets, scoring explanation, and stated population or use. Ask whether they point toward different observations.If the documentation does not answer these questions, leave the relationship uncertain. The reports may concern related tendencies, while their scores remain on separate scales unless evidence supports a link between those exact measures.

Sources: How similar are personality scales of the ‘same’ construct? A meta-analytic investigation

Why percentiles and standardized scores can mislead

A percentile rank describes a score’s position in a particular reference distribution:. The same percentile on two personality reports therefore means a similar relative position within each report’s own reference group and scoring method. It does not mean the person has the same amount of a trait on both instruments. The Standards for Quality and Fairness defines percentile rank in relation to a specified score distribution. The reference group is part of the result, not small print. “Above most people in this sample” and “high on the same underlying quantity measured by another test” are different statements. One report might compare a respondent with a broad adult sample; another might use a group defined by age, language, occupation, or testing program. Unless the reports identify those groups and their scoring rules, equal percentile labels cannot show whether the comparison populations are alike. The Standards for Quality and Fairness calls for documenting who was sampled, how the sample was drawn, when norming occurred, and any limits of program-based norms. This is testing guidance, not a personality-specific study of norming practices. Educational Testing Service’s “The Practice of Comparing Scores on Different Tests” offers a bounded illustration. Discussing SAT and ACT comparisons, it explains that using each test’s percentiles as if they directly aligned assumes equivalent test-taking groups; in its example, the groups were not equivalent. This is an educational-testing example, not evidence about personality-test populations. The narrower lesson is that a percentile comparison carries an assumption about reference groups. It cannot establish that two personality reports use equivalent norms, much less that their constructs or score units match. Standardized scores also require context. “Standardized” means a score has been transformed according to a stated system; it does not name one universal unit. Instruments can use different centers, spreads, transformations, or reference samples. Two scores with familiar labels or identical values may therefore follow different rules. A raw total depends on which items are included and how responses are combined. A percentage of possible points is a fraction of a test’s available total unless the report defines a percentile rank; it is not automatically a comparison with other people. None of these transformations alone establishes that tests measure the same construct or that one score can be translated into the other. For each report, note the displayed score type, stated reference group, norming date or version if provided, and description of what is measured. Write “not stated” when documentation is missing. If a report gives a percentile without identifying the comparison group, the label alone does not reveal who the rank refers to. If two reports show standardized numbers without explaining their scales, matching values do not establish shared units. ETS’s admissions-test example makes this a measurement principle, not a direct personality finding. The Standards for Quality and Fairness and the Standards for Educational and Psychological Testing treat norm-group description and population relevance as necessary to score interpretation. Neither source demonstrates that any particular pair of personality tests can be linked. Together they support a practical sequence: establish what each number represents and whom it compares the reader with, then look for separate evidence linking the instruments. Until that evidence exists, keep each result on its original scale and compare the stated meanings in words.

Sources: The Practice of Comparing Scores on Different Tests; Standards for Quality and Fairness; Standards for Educational and Psychological Testing

What can a correlation or crosswalk actually claim?

Correlation can show that people who score higher on one measure tend to score higher on another in a studied sample. By itself, it does not show that the measures use matching units or that one person's score can be substituted for the other. A prediction equation, concordance table, and equating result make different claims, each with its own uncertainty. The useful question is not simply whether two scores are related, but what the evidence permits someone to do with that relationship.

Think of the claims as a ladder. An association describes how scores vary together: higher results on one measure may tend to accompany higher results on another. A prediction equation goes further by estimating one score from another, but the estimate has residual error. A person's observed result can differ from the predicted value, and the estimate is meaningful only under the conditions and population for which it was developed.

A concordance, sometimes presented as a crosswalk, reports corresponding score ranges for specified tests and groups. It does not mean the tests measure the construct in precisely the same way. “The Practice of Comparing Scores on Different Tests” describes concordance as a linkage for related but distinct tests, typically with similar constructs, examinees, and purposes, while specifications or perspectives differ. Such a study can support a limited comparison, distinct from claiming that either score can stand in for the other in every interpretation or setting.

Equating is the stronger claim: scores are intended to be interchangeable. ETS places equating at the top of its linking categories and says the tests must measure the same construct; similar numbers or a calculation connecting them are not enough. The article concerns educational testing, so its examples do not validate a conversion between personality measures. The distinction still clarifies the claim: a relationship, a range correspondence, and score substitution are separate conclusions, each requiring evidence suited to its strength.

Reliability does not close this gap. It concerns the consistency of scores ; it does not establish that two tests share a unit or agree for individuals. Two measures can each produce consistent scores yet emphasize different parts of a broad trait. Evidence for a between-test claim must address the relationship between those measures, not only the repeatability of each considered alone. A correlation is not agreement: scores can rise together while differing systematically in level or carrying substantial uncertainty around an individual's predicted counterpart.

“Quality Testing Standards and Criteria for Comparability Claims” asks programs to supply evidence for the particular comparison they intend to make. ETS's “Standards for Quality and Fairness” call for a defined population, study design, analysis, results, and estimated error in linking work. The AERA, APA, and NCME testing standards distinguish direct evidence for comparability from a stronger equating claim. These documents guide testing practice, ; they do not prove that an unnamed personality-test pair has been linked successfully.

The strongest alternative to keeping scores separate is a carefully designed study supporting a modest, bounded comparison. If its sample fits the intended readers, methods suit the relationship being estimated, and error is reported, a crosswalk may help interpret correspondence without establishing interchangeability. A study of one pair of versions does not automatically cover another version, group, administration, or purpose. A responsible report states the inference its evidence supports and where it stops.

Ask what a coefficient or conversion table licenses: an association in a sample, a predicted value with error, correspondence between ranges, or score substitution. Check which instruments and versions were studied, for whom, and how uncertainty is described. If the evidence supports only association or concordance, do not upgrade that wording to interchangeability. Keep results on their original scales unless evidence supports the intended conversion: a relationship is informative, but it is not a common ruler.

Sources: The Practice of Comparing Scores on Different Tests; Quality Testing Standards and Criteria for Comparability Claims; Standards for Quality and Fairness

When is a crosswalk credible enough to use?

A crosswalk deserves consideration when its documentation identifies the exact test versions, target population, construct relationship, sample and administration design, conversion method, intended meaning, and uncertainty. Without those details, a reader cannot tell which people or which interpretation the mapping covers. The practical test is not whether the report displays a conversion, but whether the evidence answers the question the conversion invites: what may be inferred from this converted score, for whom, and with what expected imprecision? Start by checking that the document names both instruments and the versions actually taken. Then look for the authors’ definition of comparability. Does the table intend to place people in similar relative positions in a specified group, estimate one score from another, or support substituting one score for the other? Those are different claims with different and practical consequences. “Linked” is not itself a promise of interchangeability. The AERA, APA, and NCME Standards for Educational and Psychological Testing distinguish the claims: when tests cannot be equated, they call for direct evidence of comparability, a defined population, and a clear account of the link’s construction, interpretation, and limits. The population statement should be specific enough to judge fit. A study of one language group, age range, country, or occupational setting cannot automatically establish the same mapping for a different set of readers. Look for who took both instruments, how they were recruited, and whether the sample resembles the people for whom the conversion will be used. ETS Standards for Quality and Fairness call for the relevant population and definition of comparability, as well as information about the samples and similarity of forms. These are standards for an educational testing program, not validation of a personality-test crosswalk; here they provide a useful documentation benchmark, not a personality-specific rule. Next, inspect how the data were collected. Some linking designs have the same people take both measures; others compare groups intended to be equivalent or use shared anchor material. Each design relies on assumptions. If different groups were used, the technical report should explain why their results can be compared. If anchors were used, it should describe what they contain and how they relate to each test. ETS Standards for Quality and Fairness specifically ask for the data-collection design and its assumptions, including how groups were judged equivalent or how anchors were characterized. The conversion itself also needs a traceable method. A technical report should identify the statistical procedure, its assumptions, and enough detail or references for qualified readers to evaluate it. The Joint Standards’ score-linking guidance likewise calls for technical information on how a link was constructed, its accuracy, and evidence relevant to the population and intended use. Finally, match the strength of the conclusion to the strength of the study. A normative table describes positions relative to a norm group; that alone does not show that two tests measure the same construct or that their scores can be substituted. A direct cross-test study may support a limited correspondence for its sampled population without establishing universal equivalence. To claim that scores are interchangeable, ETS Standards for Quality and Fairness call for a rationale for interchangeability, a clear description of the method and its accuracy, and evidence that the forms measure the same construct with documented similarities. So request a technical report that answers five questions: Which versions? Which population and intended interpretation? What sample and design? What conversion method and assumptions? What uncertainty and limits? If one answer is missing, treat the mapping as unestablished for your purpose, rather than filling the gap with the provider’s terminology. Even a well-documented link applies only within the conditions studied; it does not automatically transfer to another version, population, or decision. These educational measurement standards are analogues for judging documentation quality. They cannot establish that any particular pair of personality tests has a valid crosswalk. For that conclusion, the evidence must concern those tests and the claim being made.

Sources: Standards for Quality and Fairness; Standards for Educational and Psychological Testing; Quality Testing Standards and Criteria for Comparability Claims

How to compare two reports without converting them

When a verified crosswalk is unavailable, compare the reports' construct statements, facets, score types, norms, intended uses, and behavioral descriptions side by side. You can note overlap and disagreement without claiming that the numbers measure the same quantity. The aim is to preserve each report’s meaning and identify a question to check in experience. Copy each report’s wording before replacing it with a shared label. Record its instrument name and version, construct definition, and narrower facets when available. “Detail focus” might sit beneath a broad domain in one report but refer to a specific preference in another. The meta-analytic investigation, How Similar Are Personality Scales of the “Same” Construct?, examined commonly used personality measures in organizational contexts and found that scales grouped under the same construct were only moderately correlated in many cases. It points to differences in construct definitions and facet emphasis. This bounded finding does not establish how any particular pair relates; it supports comparing what was measured before comparing outputs. Use a compact worksheet: | Question | Report A | Report B | |---|---|---| | Instrument and version | Copy report name; mark missing details unknown. | Copy report name; mark missing details unknown. | | Construct and facets | Preserve wording and scope. | Preserve wording and scope. | | Score type | Note raw total, scale score, percentile, band, or stated form. | Note raw total, scale score, percentile, band, or stated form. | | Norm or reference | Record the group, if given. | Record the group, if given. | | Intended use | Copy the stated use. | Copy the stated use. | | Link evidence | Record a source, or “not found.” | Record a source, or “not found.” | | Uncertainty | Note error, limits, or absent information. | Note error, limits, or absent information. | | Behavior to notice | Name an action in context. | Name an action in context. | “Unknown” is a useful entry, not an invitation to guess. If the report gives no version, norm group, or cross-test evidence, the worksheet should show that gap plainly. The ETS Standards for Quality and Fairness says users need information about score meaning and limits, and cautions that raw or percent-correct scores have limits for comparisons across tests. These standards concern ETS practice; they offer a documentation prompt, not proof that a personality crosswalk exists. Consider an explicitly illustrative comparison. One report’s description of detail focus might emphasize checking work for omissions; another might emphasize wanting instructions and procedures specified. Both descriptions could matter to the same reader, yet they point toward different observations. The reader could notice whether they reread a finished document to catch missing pieces, seek clarification when instructions are broad, both, or neither. This illustration makes no claim about either instrument; it shows how exact report language can prompt a situational question. Keep the observation bounded. One instance does not verify a tendency. Ask whether behavior recurs across situations, whether the setting permits it, and what else could explain it. A checklist may be required, or instructions may be incomplete. Context prevents a report description from becoming a verdict about character. Summarize in plain language: “Both reports draw attention to…”; “They differ in…”; and “I would want to notice…” followed by one low-stakes situation. For example, before a project handoff, the question might be whether the person independently checks for missing details, asks for clearer steps, or responds differently depending on the task. This is an editorial application of construct differences, not a meta-analysis finding. Do not average, subtract, or rank the scores merely because the reports use similar labels or familiar-looking numbers. A difference may reflect facets, scoring, reference groups, or other design choices; without link evidence, its source is unresolved. If the provider supplies a study for these exact versions, check its population, comparison method, intended interpretation, and reported uncertainty before using its mapping. Otherwise, preserve both original results and use their descriptions to frame a reflection or coaching conversation. The useful result is a precise question about each report’s scope and what contextual evidence could clarify the overlap.

Sources: How similar are personality scales of the ‘same’ construct? A meta-analytic investigation; Standards for Quality and Fairness

What changes when a decision has consequences?

The more a decision affects someone's options or treatment, the less acceptable an informal score conversion becomes. For private reflection, compare reports qualitatively, label it tentative, and tie it to what each instrument describes. In coaching or workplace development, ask what each report supports and whether work context offers another explanation. Use comparison to frame a question, not turn different numbers into a verdict.

For selection, promotion, pay, or diagnosis, a casual crosswalk is not an adequate basis for interpretation. Reliability describes consistency under set conditions; it does not establish that an interpretation is valid for a use or that two scores can be substituted. The ETS Standards for Quality and Fairness say evidence should fit the intended purpose and population, with stronger evidence needed as possible negative consequences increase. The Standards for Educational and Psychological Testing call for support and limitations when scores are linked. These standards do not validate any personality-test pair.

If a consequential decision depends on the comparison, ask the publisher for the linking study: which versions and population it covers, what interpretation it supports, and what uncertainty remains. A qualified assessor can evaluate that evidence. Personality Report's live 100-item Work Pattern Report offers reflection across ten decision-and-collaboration continuums, but has no norms, cutoff, type, or cross-test conversion; it is not an employment score. Suitable evidence could change the answer for a defined test pair, population, and claim. Otherwise, keep each score on its original scale.

Sources: Standards for Quality and Fairness; Standards for Educational and Psychological Testing

Questions readers ask