What Do AERA, APA, and NCME Standards Say About a New IQ-Score Use?

What do the AERA, APA, and NCME Standards say in one sentence?

The 2014 Standards for Educational and Psychological Testing require evidence for a particular interpretation of IQ scores and its proposed use. For a new use, the user must explain why existing evidence supports the inference and obtain more evidence if needed. A fresh validation study is not automatic, but the unchanged test or score label does not establish that the new interpretation is justified. Here, validity means the support that evidence and theory provide for interpreting scores for a proposed use. The Standards frame validity around that interpretation and use, not as a permanent badge attached to a test. The AERA, APA, and NCME Standards direct attention to what a score is asked to mean, for whom, and toward what decision. A result may describe performance under one set of conditions without supporting a broader claim or consequential decision. The practical question is whether evidence fits the proposed claim and context. If it does not, the user should narrow the interpretation or gather relevant additional evidence before relying on it. The Standards provide a professional framework; they do not validate a particular IQ test or establish that its score can serve every purpose.

Sources: Standards for Educational and Psychological Testing (2014); Standards for Educational and Psychological Testing (New Edition—2014) — AERA publisher page

Which Standards are we applying, and what do they govern?

The assigned document is the 2014 joint AERA, APA, and NCME Standards. It provides professional guidance for developing, evaluating, administering, interpreting, and using tests across education, psychology, and employment. It does not validate an individual IQ test or tell a reader what a score means. A score interpretation still needs evidence tied to the instrument, people tested, conditions, and proposed use. The Standards offer a framework for judging whether evidence fits, but their title alone cannot answer that question.

AERA’s publisher page, “Standards for Educational and Psychological Testing (New Edition—2014),” says the text was developed jointly by the American Educational Research Association, the American Psychological Association, and the National Council on Measurement in Education. It addresses professional and technical issues in test development and use in education, psychology, and employment. The official Standards site also identifies these organizations as joint publishers and labels its edition information as covering the 2014 Standards. These records establish which document is under discussion, not whether a particular IQ-score interpretation is supported.

The official site links to information about the 2014 edition and displays announcements about a joint revision committee, dated February and June 2024. Those announcements show that revision work was announced; they do not establish that a replacement edition has been issued. The careful description is “the 2014 Standards.” A committee announcement is not a new set of requirements when the official pages continue to identify the 2014 text.

The Standards state broad expectations for test development and use. A publisher’s technical documentation describes a specific instrument and may describe its scoring and studies. Test-specific studies provide evidence about specific interpretations and populations. The person proposing a new use must explain why that evidence supports the inference. The Standards guide this review, but are not a test manual, independent validation study, or endorsement of a named instrument. They do not determine the meaning of a reader’s result.

NCME’s “Code of Professional Responsibilities in Educational Measurement” is a useful supplementary source. The opened document identifies itself as the second edition, revised in 2023, and says it guides professional conduct in educational assessment. It discusses responsibilities for people who develop, use, evaluate, and market assessments. The code adds an ethical lens when considering assessment claims, but it is not the joint Standards and does not replace evidence about a particular test or proposed IQ-score use. Readers can use the Standards as a framework while still asking for evidence about the score and decision at hand.

Sources: Standards for Educational and Psychological Testing (New Edition—2014) — AERA publisher page; Standards for Educational and Psychological Testing — official Standards site; NCME Code of Professional Responsibilities in Educational Measurement (2023)

What exactly changes when a score gets a new use?

An IQ score does not carry one meaning into every decision unchanged. Describing performance on reasoning items is narrower than interpreting that performance as an estimate of general ability, and narrower still than using that interpretation to decide eligibility for an opportunity. The 2014 Standards for Educational and Psychological Testing treat these as claims needing support for their particular use. Imagine a reader completes a reasoning test and says, “I answered these administered items this way under these conditions.” That is a statement about observed performance. Raw performance means what the examinee did on the administered items, such as the responses recorded or the number answered correctly. It does not, by itself, say how performance compares with other people or what broader ability it represents. A normed comparison interprets a score against a specified reference group; moving from responses to that comparison requires a scoring method and evidence that make it meaningful for the relevant examinees. Now imagine the same reader proposes that an IQ interpretation should determine eligibility for an educational opportunity. The test, person, and displayed number may be unchanged, but the claim has expanded. It assumes that the items represent the ability relevant to the opportunity, that the score estimates that ability well enough for this decision, and that the ability is a suitable basis for eligibility. The inferential chain runs from responses to a scored result, from that result to an ability interpretation, and from the interpretation to action. Each step adds a claim between testing and the proposed action. The Standards make that shift consequential. Standard 1.4 says that when a score is interpreted for a use in a way that has not been validated, the user must justify the new interpretation, provide a rationale, and collect new evidence if necessary. Its commentary says professional judgment must determine how much existing evidence applies to the new situation or group and what more may be needed, considering similar prior uses and the amount, quality, and relevance of existing data. Novelty alone neither disproves nor establishes support. The 2013 article “Validating the Interpretations and Uses of Test Scores” explains the distinction: validation evaluates the plausibility of score-based claims by laying out the inferences and assumptions from responses to interpretations and uses. It also separates an interpretation from the action taken on its basis. Evidence supporting an ability interpretation does not, by itself, establish that it is appropriate for an eligibility decision. That decision needs a reasoned account of why the construct matters to the opportunity and evidence for any further claim about the decision. A reader assessing a proposed new use can write: “This score means ___ for these people under these conditions, and I want to use it to ___.” Identify which part describes performance, interprets it, and introduces a decision. This separation exposes the new inference that needs justification. The same number may support a modest description while remaining insufficient for a consequential classification.

Sources: Standards for Educational and Psychological Testing (2014); Validating the Interpretations and Uses of Test Scores — Kane (2013), Journal of Educational Measurement

Where is the inferential bridge, and what does it need to carry?

A defensible argument for a new use makes visible the steps between what a person did and what someone proposes to conclude. The 2014 Standards for Educational and Psychological Testing organize this as evidence and reasoning for a specified interpretation and use, not as a certificate attached to a test name. Kane’s article “Validating the Interpretations and Uses of Test Scores” offers a way to inspect the chain: state the claim, identify its assumptions, and ask whether evidence makes them plausible. This is a map for judgment, not a form whose boxes prove validity.

Start with observed responses: what tasks were presented, and what did the examinee do under the stated conditions? Then ask how those responses became a score. The scoring rule, any transformation, and the precision of the result matter because the next claim depends on them. A score can summarize performance without yet establishing what broader ability it represents. Moving from the score to an interpretation therefore requires a reasoned account of why the tasks and scoring support that meaning for the people being assessed. The 2014 Standards emphasize the quality and relevance of evidence to the intended interpretation; a familiar label such as “IQ” cannot supply the missing link by itself.

Different evidence can illuminate different links. Test-content evidence asks whether tasks represent the intended domain. Evidence about response processes considers whether examinees engage with tasks as the interpretation assumes. Internal structure can show whether score relationships fit the proposed account, while relationships with other variables can test expected connections. Evidence about consequences may reveal an irrelevant barrier or an important part of the construct being left out. These sources are not interchangeable votes. Their value depends on how they were obtained and whether they bear on this interpretation, population, and use. The Standards call for a coherent body of evidence, not a universal shortcut.

A reliability estimate concerns consistency under specified conditions; by itself it does not show that a score means a particular ability or matters to a decision. A correlation with an outcome matters when the measure, outcome, population, setting, and inference fit the claim. If a decision predicts later performance, the bridge needs evidence connecting the score interpretation to a suitable criterion in context. Even then, a relationship does not settle whether using the score is fair or worthwhile. Kane distinguishes support for interpreting scores from the further argument for using that interpretation; a plausible first step does not establish every later step.

The argument should also face contrary findings and alternative explanations. An apparent relationship might reflect another feature of the sample or setting; a result that differs across groups or conditions may limit how far a conclusion travels. These possibilities do not automatically defeat an interpretation, but matter when they bear on its assumptions. The Standards’ relevance test is not satisfied by collecting evidence without asking what it shows. A favorable statistic cannot settle a chain whose other links remain unsupported. Ask which assumption carries the conclusion, and what evidence addresses it. If that link is weak, narrow the claim to what observations support, or seek evidence suited to the broader inference. This lets a reader distinguish demonstrated performance from an interpretation or action that still needs justification.

Sources: Standards for Educational and Psychological Testing (2014); Validating the Interpretations and Uses of Test Scores — Kane (2013), Journal of Educational Measurement; Validity Evidence of the Use of Quantitative Measures of Students in Elementary Mathematics Education

How much does the original population resemble the new one?

Evidence transfers more plausibly when the people represented in the research resemble those in the proposed use for the score interpretation. A difference matters if it could change task performance or how scores relate to the decision. The 2014 Standards call for describing the composition of samples used for validity evidence, including relevant sociodemographic and developmental characteristics, so users can judge fit. The Standards identify possible limits: self-selection, attrition, linguistic ability, disability status, and exclusion criteria. Education, developmental level, and community context may also matter, depending on the intended use. These factors prompt inquiry into whether evidence covers the people and conditions in the new claim. If a study excludes people who need an accommodation, for example, uncertainty remains about applying its interpretation when that accommodation is used. This calls for relevant evidence or a narrower claim, not an automatic rejection. The Standards also call for reporting missing data and how it was handled, because who remains in an analysis can affect what its findings support. ([Standards for Educational and Psychological Testing](https://www.testingstandards.net/uploads/7/6/6/4/76643089/9780935302356.pdf)) Keep two population questions distinct. A norm-reference population is the group against which a score is compared. A validation sample is the group from whom evidence was collected to support an interpretation or use. They may overlap, but answer different questions: “Compared with whom?” and “Evidence for whom?” A norm group alone does not show that a score predicts a new outcome or supports an eligibility decision. Validation evidence does not itself establish the comparison group. Ask both what population underlies the comparison, if any, and whether evidence for the new inference includes sufficiently relevant people. The Standards explain that sample characteristics can influence statistical findings. If evidence comes from a narrow group but the proposed use concerns a broader one, the user needs an account of why the findings apply, and more evidence if the gap matters. When the people, conditions, and intended inference closely match, prior evidence may still be informative. Similarity is not a mechanical checklist: the key is whether a difference bears on the conclusion being drawn. An example from another field illustrates why documentation matters, without establishing anything about IQ tests. “Validity Evidence of the Use of Quantitative Measures of Students in Elementary Mathematics Education” reports a systematic search of 1,582 articles that identified 192 measures used in elementary mathematics research. The review found explicit interpretation and use statements were infrequent in the material examined. Its coding recorded whether evidence was reported, not whether analyses were correct or high quality. It cannot establish how often IQ measures lack population-matched evidence. It illustrates why readers should check the people and intended use behind a claim. For a proposed new IQ-score use, write down who the original evidence covered and who will be affected. Compare relevant age or developmental stage, language demands, disability and access conditions, education or selection context, and exclusions. If the documentation cannot answer that, treat the fit as unresolved: seek evidence from a closer population, qualify the interpretation, or avoid using the score for that decision. A mismatch is a reason to investigate, not a verdict on an individual or group.

Sources: Standards for Educational and Psychological Testing (2014); Validity Evidence of the Use of Quantitative Measures of Students in Elementary Mathematics Education

What if the new use changes administration or scoring?

A change in testing conditions matters when it changes what responses can reasonably be taken to show. The 2014 *Standards for Educational and Psychological Testing* tie interpretations to the conditions under which evidence was collected, including administration and scoring. They do not say every difference between settings invalidates a score. Users should check whether new conditions still support the same interpretation. Consider moving an assessment to another room versus changing the task itself. A quieter room may be an operational difference with no obvious effect on the intended reasoning operation, though that depends on the test and evidence. By contrast, adding a time limit to a task previously administered without one could make speed part of performance. If the intended interpretation concerns reasoning without a speed demand, the change may introduce a barrier unrelated to the intended construct. The score label cannot resolve that problem. The Standards’ fairness guidance connects access with interpretation. Conditions should generally be comparable to those used when norms and interpretive data were obtained, and procedures should reduce barriers irrelevant to the construct. This does not mean every examinee must receive identical conditions. An accommodation may remove an irrelevant barrier and make performance more interpretable; a modification that changes the operation required may alter what the score represents. Ask what the change does to the inference, rather than relying on whether it is called an accommodation or modification. “Online” and “supervised” are not enough to decide whether evidence transfers. Delivery mode can bundle differences in instructions, timing, assistance, interruptions, equipment, or answer entry. Identify the actual conditions and ask which could affect responses, scoring, or interpretation. The Standards establish neither automatic equivalence nor automatic inferiority for browser testing. Evidence from one set of conditions informs another only to the extent that conditions and proposed interpretation are sufficiently alike. Scoring changes also deserve scrutiny. A conversion from responses to a reported score, a rule for incomplete responses, or an adaptive procedure that selects later items based on earlier answers can affect what a summary represents. Ask whether the rules are documented and evidence supports the interpretation for the intended population and conditions. If a report turns a score into a further claim, such as an ability label or recommendation, that claim needs support too. The 2023 *NCME Code of Professional Responsibilities in Educational Measurement* calls for clear statements about purpose and cautions, appropriate reporting, and claims grounded in available evidence. It is professional guidance, not proof that a particular scoring method works. Compare the conditions behind the old evidence with the proposed use: task demands, timing, instructions, access supports, scoring rules, and reporting meaning. A difference that could change performance or interpretation creates a question to resolve. Existing evidence and a clear rationale may answer it; sometimes the interpretation must be narrowed or more evidence gathered. A changed setting alone does not settle the matter, but a familiar score name cannot carry evidence across changed conditions without explanation.

Sources: Standards for Educational and Psychological Testing (2014); NCME Code of Professional Responsibilities in Educational Measurement (2023)

How do stakes and consequences affect the decision?

A consequential decision raises two questions: does evidence support the proposed score interpretation, and is using it for this action justified? The 2014 Standards for Educational and Psychological Testing distinguish consequences that expose a problem in interpretation from consequences that matter to the broader decision to use a test. Greater stakes call for scrutiny; they do not strengthen weak evidence.

A consequence can reveal a measurement problem if a test omits important parts of the intended ability or irrelevant features distort performance. It may then point to a flaw in what the score is said to mean. But an unwanted effect can arise even when interpretation is supported. The Standards discuss applicants forming a negative view of an organization because it uses an employment test. That reaction may matter to whether the test should be used, but it does not by itself show that the score fails to predict intended performance. Consequences matter to use decisions without every consequence testing score meaning.

Some accounts give consequences a broader role in validation. The Standards distinguish consequences tied to interpretation errors from other outcomes and value judgments; they do not dismiss consequences. Kane’s “Validating the Interpretations and Uses of Test Scores” likewise treats validation as an argument whose claims and assumptions need support. Evidence for interpreting a score does not settle every question about whether to use it.

Standard 1.5 addresses promises about what testing will accomplish. When a recommendation states or implies a specific outcome, its basis and relevant evidence should be presented. The Standards illustrate this with a claim that employee selection will reduce errors or training costs. Evidence that a score represents a construct does not alone show that acting on it will produce an advertised benefit. That outcome claim needs support suited to its context.

Scrutiny should rise with possible impact, but the Standards give no numerical threshold for how much evidence is enough. Their test-user guidance says greater potential impact on test takers brings greater need to identify and satisfy relevant standards. Using results to choose a practice exercise differs from using an IQ interpretation to allocate access, place a student, or make an employment decision. Such uses can affect opportunities, so users need support for the interpretation among relevant people and conditions, a reason the construct bears on the decision, and attention to likely benefits and harms. A desirable outcome does not make a browser quiz suitable for that decision.

Ing and colleagues’ “Validity evidence of the use of quantitative measures of students in elementary mathematics education” offers a bounded illustration: the review found rigorous evidence for intended uses in one in ten elementary mathematics measures examined. It illustrates why users should inspect intended-use documentation rather than assume having a measure settles the question.

Keep interpretation, claimed outcome, and decision distinct. Ask whether a consequence signals that the score captures the wrong thing or includes irrelevant barriers; then ask what evidence supports the promised result, including findings that complicate it. A supported interpretation is necessary for consequential use, but does not compel it. Users must separately weigh whether the action is fair, useful, lawful, and proportionate in context.

Sources: Standards for Educational and Psychological Testing (2014); Validating the Interpretations and Uses of Test Scores — Kane (2013), Journal of Educational Measurement; Validity Evidence of the Use of Quantitative Measures of Students in Elementary Mathematics Education

Does a new use always require a new validation study?

No. Standard 1.4 of the 2014 Standards for Educational and Psychological Testing makes additional evidence conditional: the user must explain why existing evidence applies to the proposed use and obtain further evidence when support is insufficient. A new purpose does not erase relevant evidence, but an unchanged test or IQ label does not carry evidence automatically into a different decision. Compare the old and proposed arguments. Write what the existing evidence supports: what score is interpreted, for which people, under what administration and scoring conditions, and for what purpose. Then write the proposed claim with the same details. Ask which links remain supported and which have changed. A familiar score name can conceal a new inference about the people assessed, the conditions under which they respond, or the outcome the score is expected to inform. Consider an example. A test has evidence supporting a description of reasoning performance among people tested under conditions similar to those studied. A reader then proposes using its IQ interpretation to predict eligibility for a separate opportunity. The second claim adds a prediction and a decision. Evidence about the first interpretation may still matter, but does not by itself show that the score predicts this outcome or that the prediction applies to the new group. The user must identify and support those added links. Now consider a closer change: the same score gives descriptive feedback about a related reasoning task, for a similar group, with comparable instructions and scoring. Existing evidence may address much of this interpretation. The user still needs to explain why the task, people, conditions, and meaning match. Similarity is a judgment about relevance, not a checklist pass. The Standards specify no formula, minimum study count, or rule that every new purpose requires a new validation project. The argument-based approach in “Validating the Interpretations and Uses of Test Scores” evaluates the claims and assumptions connecting observations to an interpretation and proposed use; more ambitious claims require more support. This explains why evidence can travel partway. Evidence about score behavior may support an interpretation, while a new prediction or decision may need evidence for an additional step. Kane's article is conceptual guidance, not an IQ-test study or prescribed quantitative procedure. Standard 1.4's commentary makes the judgment depend on professional evaluation of fit, including the amount, quality, and relevance of existing evidence and the similarity of prior uses. What could change the conclusion? Documentation closely matching the proposed claim, population, administration, scoring, and purpose could show that existing support is adequate. A material mismatch, an unsupported new outcome, or contradictory findings could show that more evidence is needed, or that the claim should be narrowed. The purpose change alone gives no automatic verdict. Retain only the interpretation whose bridge is supported, and seek evidence for a missing link before relying on a broader use.

Sources: Standards for Educational and Psychological Testing (2014); Validating the Interpretations and Uses of Test Scores — Kane (2013), Journal of Educational Measurement

What does the transfer rule show when applied to Test IQ Free?

Applied to Test IQ Free, the Standards’ transfer rule separates what the test records from what its displayed estimate might imply. The adaptive test presents 50 original reasoning questions across pattern, number, and spatial reasoning. A person’s responses and domain feedback describe performance on those questions under the conditions in which they were answered. That is narrower than saying the result establishes the person’s position in an adult population or qualifies them for a consequential decision.

A normed interpretation compares a score with performance in a defined reference population using a supported scoring conversion. The product brief states that Test IQ Free’s provisional estimated IQ range uses placeholder conversion parameters and has no validated adult reference population. These are limits of this product, not findings made by the AERA, APA, and NCME Standards. The 2014 Standards say an interpretation for a specified use needs appropriate support; they do not supply missing norms or validate this conversion. The range is therefore not a normed population comparison, established IQ score, or confidence interval. It cannot answer a new question about eligibility, diagnosis, gifted identification, education, or employment.

The transfer comparison asks whether evidence for an earlier claim supports the new claim for the relevant people, conditions, scoring, and decision. Here, the modest claim that someone answered reasoning items in a particular way is supported by the recorded responses. Moving from that observation to a population-level IQ interpretation requires a reference group and a justified conversion. Moving from that interpretation to a consequential decision requires evidence that it is relevant to that decision. The provisional range supplies neither bridge. The AERA publisher description confirms the Standards’ broad guidance on test development and use; it is not an evaluation of this test.

For low-stakes curiosity, the test can offer a structured reasoning exercise and feedback across its three domains. A reader can review the worked answers and notice which operations were difficult or familiar, keeping that observation local to the exercise. The optional $9 report adds explanation and worked examples but uses the same provisional conversion; payment adds no validity or accuracy. Neither result supports conclusions about worth, fixed potential, or clinical status. For an educational, clinical, or employment decision, use an assessment supported for that purpose and population and seek qualified interpretation.

Sources: Standards for Educational and Psychological Testing (2014); Standards for Educational and Psychological Testing (New Edition—2014) — AERA publisher page

What can you responsibly do next?

Write the proposed interpretation as a claim: what does this score mean, for whom, under which conditions, and for what decision? Then compare that claim with the evidence collected for the test and its earlier use. The Standards for Educational and Psychological Testing (2014) say that when an interpretation is proposed for a use without adequate support, the user must justify it and gather additional evidence if needed. It means the reason prior evidence applies must be clear, and any important gap should limit the conclusion. If the evidence does not support the new claim, narrow it to what was observed or set the score aside for that decision.

For a consequential educational, clinical, or employment decision, use an assessment with evidence suited to that purpose and population, and seek interpretation from a qualified professional. For curiosity, Test IQ Free can offer low-stakes practice feedback: its adaptive 50-question test reports domain performance and a provisional estimated IQ range. The conversion uses placeholder parameters and has no validated adult reference population, so that range is not a normed comparison, confidence interval, diagnosis, or established IQ score. The free test and worked answers are available at [/test](/test); treat the result as a starting point for reviewing reasoning operations, not as a decision about who you are.

Sources: Standards for Educational and Psychological Testing (2014)

Questions readers ask