How to Compare Adult IQ Scores After a Test Edition Changes

What conclusion can you draw from two scores after an edition change?

When an adult IQ test has been revised, the safest first conclusion is modest: the two reports describe performance on two assessment occasions under their own measurement conditions. Their numbers are not automatically a before-and-after measure of change. A later score may be higher or lower, but arithmetic subtraction alone does not show that the person’s underlying ability changed by that amount. The key question is whether technical evidence supports comparing these particular scores. The 2014 Standards for Educational and Psychological Testing treats score linking as a family of procedures serving different purposes; a link may not make scores interchangeable, and evidence may not extend to populations or uses beyond those for which it was developed. That distinction matters when an edition changes. Matching labels can make results look directly comparable, but a label does not establish that the score scales, definitions, or interpretations stayed the same. A stronger conclusion therefore needs evidence suited to the exact editions, score types, population, testing conditions, and decision at hand. An examiner should be able to explain what relation the evidence supports and where it stops. Evidence supporting a limited comparison does not necessarily support a claim about one individual’s change; that depends on the comparison method and its intended interpretation. If no suitable evidence is available, keep both scores with their dates and contexts, and leave the cause of any difference unresolved. This caution does not erase the older report: it remains a record of the earlier assessment and may still inform the question it originally addressed. The next step is to ask what, if anything, the documentation permits you to conclude from the pair.

Sources: Standards for Educational and Psychological Testing

What exactly are you comparing in the two reports?

Before comparing numbers, make a record of what each report actually describes. Write down the instrument and edition, date of testing, language or national adaptation, score name, stated purpose, and administration conditions. These details identify the conditions under which the reported performance was obtained. The Standards for Educational and Psychological Testing ties interpretation to intended use and examinee population, so a number separated from context carries less meaning than it appears to. Then distinguish the kinds of numbers on the page. A raw score is the number of responses credited under the instrument's scoring rules before it is converted to a derived scale. A norm-referenced score locates performance on a defined scale by comparing it with a specified reference group. The reference group and scoring system matter: the same raw total can map differently if the test edition, conversion rules, or norms differ. Therefore, a raw total from one edition is not automatically comparable with a standard score from another, and matching score labels do not prove that the underlying scales were constructed in the same way. A percentile is a relative position in a reference distribution, usually expressed as the percentage of the reference group at or below a given score under the test’s scoring conventions. It does not mean the person answered that percentage of questions correctly. A response total and percentile answer different questions: credited responses under scoring rules, versus relative position in the reference group. Neither number, by itself, states how much a person's underlying ability changed between dates. The Standards discusses score scales, norms, and linking separately, so read labels and reference context from each report. For each assessment, copy the score name exactly as printed and note whether the report calls it a raw, scaled, standard, composite, index, or percentile score. Record the comparison group or norm date and any reported uncertainty information. Do not calculate a percentile from an IQ score, convert one edition's standard score into another, or treat a confidence interval as an edition bridge unless technical documentation for those exact instruments supports that operation. This comparison record prevents a common mistake: reading two familiar labels as if the measurement behind them remained fixed. A revised edition may change norms, scoring rules, or which components contribute to a composite. Those are reasons to inspect the documentation, not grounds to guess which result is more accurate or to declare a gain or loss. Describe each result as performance under its own defined administration and scoring conditions. A pair of numbers does not establish identity, worth, diagnosis, or fixed potential. First establish what each number means; only then ask whether the technical evidence permits a direct comparison.

Sources: Standards for Educational and Psychological Testing

When are scores comparable, linked, or interchangeable?

A score link is a documented relation between scores on different forms or editions. It can make a limited comparison possible without showing that the scores are interchangeable. Equating is a narrower kind of linking: it is used when a testing program expects scores from successive editions to have the same meaning over time. Interchangeability is the practical claim that scores can stand in for one another for a particular interpretation or decision. A common scale is useful because it puts results into a shared frame, but shared units alone do not prove that two tests measure performance in the same way or with comparable precision. The intended inference controls the standard of evidence. A program reporting scores across successive forms may need evidence that supports consistent interpretation across those forms. Someone asking whether their own performance changed over time needs evidence suited to repeated individual measurement as well. Treating these as separate questions prevents a technically legitimate score conversion from being stretched beyond its purpose. The distinction matters because a conversion can answer a group-level scoring question while leaving an individual change question unanswered. A linking study may estimate how scores on one scale correspond to scores on another for a defined population. It does not, by itself, show that a particular person’s underlying performance changed by the amount suggested by subtracting two results. A table that translates one reported number into another is therefore not a personal change analysis unless its evidence and method support that use. The Standards for Educational and Psychological Testing treat linking as a set of procedures that can serve different purposes, with different assumptions and limits. It also says equating evidence should address whether conversions remain stable across relevant subgroups of the population for which a test is designed. The Standards provide professional guidance for evaluating evidence; they do not supply a conversion for any particular pair of intelligence-test editions or calculate an individual’s change. Principles and Practices of Test Score Equating, an ETS research report, describes equating as important when a program continually issues new test editions and expects scores to retain the same meaning. It supports a useful question: was the conversion designed to preserve meaning for this pair and use, and how was its accuracy evaluated? When reviewing two assessment reports, ask the examiner to identify the exact editions and score types, then name the technical evidence connecting them. Ask which population and purpose that evidence covers, whether it supports a common-scale comparison or actual interchangeability, and what precision or uncertainty remains. If documentation establishes a suitable link for a limited comparison, use it only within that scope. If it does not establish the interpretation you need, keep both results as separate records of performance under their respective conditions. A stronger pair-specific study could support a stronger conclusion; a generic conversion chart alone cannot.

Sources: Standards for Educational and Psychological Testing; Principles and Practices of Test Score Equating

What changed in the WAIS-IV and WAIS-5 example?

The U.S. English WAIS-IV and WAIS-5 both report a Full Scale IQ (FSIQ), but the shared label does not mean that the score is built in exactly the same way. Pearson’s WAIS-5 Comparison Flyer identifies WAIS-IV norm and validity studies from 2007–2008 and WAIS-5 studies from 2023–2024. It also lists ten subtests for WAIS-IV FSIQ and seven for WAIS-5 FSIQ. The flyer describes four index scores and one optional index for WAIS-IV, compared with five primary and fifteen ancillary index scores for WAIS-5. These are edition differences shown in publisher documentation. They explain why the edition name and score definition belong beside any numerical comparison. They do not, on their own, tell us whether an individual’s later score should be higher, lower, or unchanged. The subtest lists show what changed in construction. The flyer names ten WAIS-IV core subtests, including Information and Arithmetic, and seven WAIS-5 primary subtests, including Figure Weights, Digit Sequencing, and the new Running Digits. WAIS-5 also lists secondary subtests, several described as new, and expands the index structure. The point is not to inspect or reconstruct test content; these are publisher-reported structural facts. A composite condenses performance on selected tasks into a reported score. When the selected tasks or the organization of indices changes, two composites with the same label may summarize performances assembled differently. The label is useful for orientation, but it cannot establish that every component, weighting, or interpretation stayed constant. Pearson’s WAIS-5 product page describes the edition as published in 2024, gives an age range of 16 years through 90 years and 11 months, and lists 45 minutes for the seven-subtest FSIQ and 60 minutes for the ten primary index subtests. Its FAQ says the FSIQ can be obtained about 20 minutes faster than WAIS-IV, while noting that total administration time depends on which composites the examiner chooses. These details describe the newer assessment’s structure and administration. They should not be read as a head-to-head finding about an individual’s scores. Product descriptions explain what the publisher offers; they do not calculate a person’s change across editions. The WAIS-5 Comparison Flyer reports that the editions’ norm-study periods differ, but that fact alone cannot establish the direction or size of a score difference for a particular examinee. A norm update describes a reference basis; individual interpretation also depends on the assessment’s construction, scoring, examinee, and conditions. Neither the flyer nor the product page provides a WAIS-IV-to-WAIS-5 conversion for an individual or shows how much a specific person’s result should move. Edition changes make a direct comparison question worth asking, but the public feature comparison does not answer it. Ask the examiner whether technical evidence for these exact editions supports the intended comparison and what that evidence can establish. If it does not address individual change, retain both scores with their edition and date rather than turning their arithmetic difference into a conclusion.

Sources: WAIS-5 Comparison Flyer; Wechsler Adult Intelligence Scale | Fifth Edition

Why can a familiar composite label hide a different comparison?

A composite name identifies the kind of summary being reported, but it does not guarantee that two editions combine the same tasks or organize results in the same way. In the WAIS example, Pearson’s *WAIS-IV vs. WAIS-5: What’s New* comparison flyer says that Full Scale IQ (FSIQ) draws on ten subtests in WAIS-IV and seven in WAIS-5. It also lists four index scores and one optional index for WAIS-IV, compared with five primary and 15 ancillary index scores for WAIS-5. Those are changes to the reported structure, not evidence about how much a particular person’s score should change. A composite compresses information from component performances into one summary. When the selected components or their organization change, the shared label may describe a related broad purpose while concealing a different recipe for producing the result. This is why a reader should inspect the score definition and component structure before treating two composites as direct repeats of one measure. Different construction warrants caution; it does not establish the direction or meaning of an individual difference. The flyer supplies concrete reasons not to rely on the label alone. Alongside the FSIQ subtest count, it lists different primary and secondary subtest groupings, new subtests, and an expanded set of index scores for WAIS-5. Pearson’s current *Wechsler Adult Intelligence Scale | Fifth Edition* product page likewise describes the WAIS-5 FSIQ as based on seven subtests and notes changes to items and scoring in particular subtests. Together, these publisher materials document features of the U.S. English editions. They do not provide a personal WAIS-IV-to-WAIS-5 conversion, show that every score relation is unsuitable, or establish the size or direction of an individual difference. That boundary matters because evidence about how scores relate across a group is not automatically an account of one person’s change. Even if technical materials report that two scores are strongly associated across examinees, association alone would not tell a reader whether a particular person retained the same relative profile, whether a numerical gap exceeds the uncertainty relevant to that comparison, or why the gap occurred. Those questions require evidence matched to the editions, score definition, examinee population, and intended use. A shared label cannot supply the missing evidence, and a structural difference cannot substitute for it either. For a practical report comparison, look for the technical description of what contributes to each named composite and whether the index structure changed. Then ask the examiner whether documentation directly addresses the relation between these editions for the proposed interpretation. The comparison flyer is useful for identifying what changed in the instrument’s published structure; it is not itself a study of individual change. Treat matching names as a starting point for checking definitions, not as proof that the results are interchangeable. If the available evidence does not support the stronger interpretation, keep both scores attached to their own edition and assessment context.

Sources: WAIS-5 Comparison Flyer; Wechsler Adult Intelligence Scale | Fifth Edition

How can retesting affect the comparison?

A second assessment can differ because the person has seen the same or similar tasks before, even when the test edition has not changed. Some of the observed gap may reflect familiarity with testing or items rather than a change in the ability being assessed. The size and direction of that influence cannot be assumed for one person from a general retest study, and it cannot serve as a ready-made correction for a comparison across editions. “Effects of Practice on the Wechsler Adult Intelligence Scale-IV Across 3- and 6-Month Intervals” followed 54 participants who took the WAIS-IV at baseline and again after either three or six months. Their mean age was 20.9 years, mean education was 14.9 years, and mean initial Full Scale IQ was 111.6. At retest, average scores increased approximately seven points on Full Scale IQ, five on Verbal Comprehension, four on Working Memory, five on Perceptual Reasoning, nine on Processing Speed, and six on the General Ability Index. Increases were similar across the two intervals. These are group averages from one sample under a same-edition retest design. They do not predict another examinee’s score, establish a universal adjustment, or show what happens when the edition changes. The researchers also compared ways to estimate whether a person's score change exceeded what might be expected from retesting and measurement variation. A reliable-change index is a method for judging whether a difference is larger than expected under a specified model of score stability and error; its meaning depends on the method and relevant data. In this study, regression-based estimates were more accurate than simply subtracting baseline from follow-up. That finding supports asking how a change estimate was calculated. It does not authorize applying the study’s method or average to a different edition pair, and the abstract supplies no personal threshold transferable to an individual report. “Implications of Short-Term Retest Effects for the Interpretation of Longitudinal Change” examines three short-interval studies of adults aged 18 to over 80 who completed cognitive test batteries. It explains that differences across occasions can include prior-test experience as well as other influences during the interval. The studies used self-selected volunteers and were not a WAIS-IV-to-WAIS-5 comparison. This article supports a general caution about repeated performance; it does not quantify the practice component in any particular pair of reports. For interpretation, tell the examiner about earlier testing, including the test or edition if known, approximate date, and any practice with similar tasks. Ask whether the interpretation method accounts for prior exposure and what evidence supports that method for these scores. Retest evidence and edition-linking evidence answer different questions. Without data suited to the tests, examinee, and purpose, neither a group-average gain nor subtraction between report numbers can identify how much of the gap reflects practice, edition changes, other testing conditions, or individual change. Leave those possibilities unresolved until the examiner can explain which the evidence supports.

Sources: Effects of Practice on the Wechsler Adult Intelligence Scale-IV Across 3- and 6-Month Intervals; Implications of Short-Term Retest Effects for the Interpretation of Longitudinal Change

Which other differences in testing conditions matter?

When two assessment reports differ, record the conditions of each session before trying to explain the gap. Note the interval between dates, the language and edition used, any accommodations, whether administration was paper or digital, and whether the examiner or setting changed. Record any relevant health or practical context as well. They do not, by themselves, show that any one condition caused a score difference. A pair of scores cannot separate the influence of changed conditions from edition changes, prior exposure, ordinary measurement uncertainty, or actual change in the ability being assessed. The point is to preserve context, not build a list of explanations after seeing the result. A language change may affect interpretation of verbal performance; a mode change raises the question of whether that format has been studied for that instrument. That is a prompt to check relevant technical evidence, not a basis for adding or subtracting points. Record what accommodation was provided and whether it changed the task or response method. Without technical information tied to the exact test and use, do not assign a numerical effect to any of these circumstances. Ask the examiner whether a condition affects the score’s interpretation for the intended decision. The WAIS-5 technical report, “Equivalence of Q-interactive and Paper Administration of Cognitive Tasks: WAIS-5,” illustrates why the scope of evidence matters. Pearson compared digital and paper administration within WAIS-5 using separately assigned groups in a U.S. English-speaking standardization context. The report describes a digital group of 498 examinees ages 16–90 and says the formats met its equivalence criterion for examined scores after an adjustment to digital Spatial Addition norms. Its results concern that study context, not every language version, accommodation, examiner, or setting. This is evidence about two administration formats of one edition under that study’s conditions. It is not evidence that WAIS-IV and WAIS-5 scores are interchangeable, or that a particular adult’s scores across editions represent change. Its result should not be carried over to a different, unrelated comparison merely because both involve the same test family. The Standards for Educational and Psychological Testing also make population and purpose part of the question: evidence supporting a score interpretation for one group or use may not support it for another. Ask whether the technical evidence matches both reports’ language versions, relevant population, administration conditions, score, and intended use. If not, preserve the reports as separate records and describe the circumstances without claiming a cause.

Sources: Equivalence of Q-interactive and Paper Administration of Cognitive Tasks: WAIS-5; Standards for Educational and Psychological Testing

What can an older report still tell you?

An older report can remain useful as a dated record of performance and as evidence for the question it originally addressed. Its age alone does not make it meaningless. But a report suitable for one purpose at one time is not automatically suitable for a later decision. Whether it can inform a new decision depends on purpose, conditions, time between administrations, and evidence supporting use of that edition. In its WAIS-5 product information, Pearson says that when a newer edition is available and a user chooses an older version, the user is responsible for providing evidence that the older version is as appropriate for that particular use. That statement concerns test selection and justification. It does not say every older WAIS-IV report has become worthless, establish an expiration date, or show that two editions produce interchangeable scores. Read it as a reason to ask whether the earlier instrument fits the decision, not as a rule to erase the earlier result. The *Standards for Educational and Psychological Testing* frame interpretation around intended use and supporting evidence. Technical support for a test does not answer every question. The relevant question is whether evidence supports this interpretation for this purpose and these examinees. An older assessment may contribute to a history or comparison while requiring care about the claim attached to it. Preserving a result as a record is a smaller claim than treating it as a current estimate or combining it numerically with a revised edition. An older report may document what was assessed, which scores were reported, and the conditions in which the assessment took place. Those details can help an examiner understand the sequence of assessments. They do not, by themselves, show that a later score rose or fell, or tell why two numbers differ. A report may remain informative even when evidence does not support direct score comparison. Keep its date and context attached, and state what role it can play in the current question. There is a reasonable counterpoint: longitudinal questions can require earlier records. If the goal is to understand performance across time, discarding an older assessment would remove information that may matter. But the need for a record does not settle how its score relates to a revised instrument. It supports retaining the report for qualified interpretation, while comparison still depends on technical evidence matched to editions, score definitions, population, conditions, and intended use. Pearson’s guidance does not make the old result irrelevant; the Standards do not turn its presence into proof of individual change. Keep both reports, note each edition and date, and bring them to the professional interpreting the current decision. Ask whether the older assessment remains appropriate for that use and what the evidence allows them to say about the pair. If no suitable cross-edition evidence applies, both records can still be described on their own terms, with the numerical difference left unexplained.

Sources: Wechsler Adult Intelligence Scale | Fifth Edition; Standards for Educational and Psychological Testing

What should you ask the examiner before interpreting the gap?

Bring both reports to the conversation, Start with the record: What are the exact test names and editions, dates, language versions, score labels, and administration formats? The Standards for Educational and Psychological Testing say that a claim of score comparability should identify the population for which it applies and should fit the intended use. Their guidance is general; it does not itself calculate a relationship between two particular editions. Then ask the examiner to name the evidence behind the comparison and explain its scope in plain language. A “reliable change” analysis is a method for judging whether an observed difference is larger than expected measurement error under stated assumptions; it is not a synonym for any numerical increase. Ask which error estimate and retest evidence the examiner used, whether the method accounts for the interval, and whether its assumptions fit the two editions and this purpose. Do not request a universal threshold when the method and score pair have not been identified. Is there a study or manual section relating these two editions? Does it support using the scores interchangeably, a narrower comparison, or an estimate of reliable change? The Standards call for a clear rationale and supporting evidence when scores are said to be interchangeable; for linked tests that are not closely parallel, they also call for the linking method, intended interpretation, and limitations to be described. If the examiner refers to retesting, ask whether the evidence matches the actual pair. A PubMed-indexed study listed in this article’s evidence plan examined WAIS-IV retesting after three or six months, not a WAIS-IV-to-WAIS-5 edition change. Its planned use here is limited: it flags prior exposure as a factor to consider, not a correction that can be applied to a different person or cross-edition pair. In that study, the test edition remained WAIS-IV; therefore it cannot isolate the effect of changing to a later edition. Ask whether the examiner has pair-specific evidence in addition to same-edition retest information. For example, imagine the newer report shows a score several points higher than the older one. Subtracting the earlier number from the later one gives an arithmetic difference. That calculation alone does not show a meaningful change: the editions may use different score construction or reference data, and ordinary measurement uncertainty and prior exposure may also matter. Ask instead: “What conclusion does the technical documentation permit for these two scores, and what remains unknown?” For a consequential education, accommodation, or clinical decision, ask the qualified examiner to interpret both reports for that specific purpose and consider other relevant evidence. The Standards emphasize matching score interpretation to the intended use and examinee population. If no direct comparison evidence applies, request a clear statement that the reports are separate records under their own conditions, rather than a quantified gain or loss. The useful outcome of the conversation may be a narrower conclusion, not a single combined number.

Sources: Standards for Educational and Psychological Testing; Effects of Practice on the Wechsler Adult Intelligence Scale-IV Across 3- and 6-Month Intervals

What is a responsible next step for a low-stakes reader?

For personal curiosity, keep each professional result attached to its edition, date, language version and testing context. Describe the numerical difference as a difference between two reported results; call it a gain or loss only if evidence for these exact scores supports that interpretation. The Standards for Educational and Psychological Testing distinguish links designed for different purposes and caution that a relation between scales does not automatically make scores interchangeable.

If you want open-ended reasoning practice, Test IQ Free offers an original 50-question browser test with worked answers. Its displayed IQ range is a provisional estimate from an unvalidated conversion with no validated adult reference population. It is not a substitute for professional assessment or equivalent to WAIS. The optional $9 report adds explanation and practice structure, not accuracy or validity. Use it to work through reasoning questions, not to settle whether your performance changed between professional assessments.

The useful next step is one conversation with the examiner who can see both reports: “Which comparison does the evidence support for these editions and this purpose, and what remains unknown?” If the answer is that no direct comparison applies, keep both scores as separate records and describe each in its own context. For a consequential decision, ask what additional evidence it requires.

Sources: Standards for Educational and Psychological Testing

Questions readers ask