Should You Retake a Personality Assessment You Dislike?

What does ‘I dislike it’ tell you—and what does it not tell you?

Dislike alone is a reaction to a report, not evidence that its score or method is wrong. It can still point to a useful question. Before repeating the assessment, locate the exact part you reject: the description of your answers, the comparison group, the interpretation, or a conclusion someone drew from it. These are different claims. A report might summarize a measured tendency too broadly; its norm group might be unclear; or someone might add an implication the instrument never established. A retake addresses none of these automatically. First restate the disputed claim in narrower, observable terms. Consider an illustrative case: a report describes someone as reluctant to take charge. That wording might feel wrong because the person has led a project. The scale might concern a narrower pattern, such as whether someone seeks decision authority when responsibility is unclear; the actual items and documentation would need to support that reading. One instance of leadership neither proves nor disproves a tendency across situations. Ask what the items measured and what time frame respondents were told to consider. Do not assume an assessment measures this pattern unless its documentation says so. Separate disagreement with your responses from disagreement with the label attached to their combined score, and from a conclusion about what that score means for work or relationships. A response-frame question differs from a norm-reference question: perhaps you answered about an unusual week rather than your usual behavior, or the report may not identify its comparison group. An inference question is different again: a score about one tendency cannot by itself establish a broad judgment about character or predict a specific choice. The American Psychological Association’s *Guidelines for Psychological Assessment and Evaluation* says assessment interpretation should attend to purpose, technical score meaning, context, and relevant information. The guidelines are professional guidance for psychologists; they do not study consumers’ reactions to online reports or prescribe a personal retest checklist. Applied cautiously, they support asking what evidence bears on the disputed claim instead of letting a pleasing or upsetting label decide the matter. Confirmation bias, the tendency to give greater weight to evidence that supports an existing view, is one possible reason someone may favor the result that feels better. But discomfort does not prove bias, just as a preferred result does not prove accuracy. Write down the sentence you dispute and convert it into a question the report could answer: What tendency was measured? Which reference group and time frame apply? Does the interpretation follow from the score, or has someone added a larger conclusion? Until you can name a measurement problem or a specific decision affected by uncertainty, a second attempt would mainly produce another result to interpret.

Sources: Guidelines for Psychological Assessment and Evaluation

Which kind of ‘reliability’ would a retake actually test?

A repeat of the same measure speaks most directly to test-retest reliability: how similar scores are across separate administrations under specified conditions. It does not, by itself, show that the measure captures the intended personality construct, predicts an outcome, or makes either of your two results the true one. Internal consistency asks a different question: whether items within a scale tend to cohere. A scale can have internally consistent items and still produce less similar scores weeks later. So before treating a retake as a fact-check, identify which kind of evidence the report offers and which question that evidence can answer.

The distinction matters because “reliability” is not a single property. In “Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity,” the authors analyze NEO Inventory facet data and explain that internal consistency concerns coherence among items, while retest reliability concerns scores on different occasions. In the study’s analyses, two retest estimates predicted three validity criteria, whereas none of three internal-consistency estimates did. The result is specific to the NEO facet scales and the criteria examined; it does not rank every instrument or establish that a particular person’s score is correct. It does show why an internal-consistency figure cannot stand in for evidence about repeat scores.

A comparison across different methods makes the point concrete. “Examination of the Test–Retest Reliability of a Forced‐Choice Personality Measure” studied 743 participants who completed multiple measures at two time points. For its forced-choice computerized adaptive measure, reported retest estimates averaged .63 across dimensions and ranged from .45 to .77; the Big Five-related composite estimates ranged from .63 to .81. For the Likert-type BFI-2 measure in that sample, dimension estimates ranged from .74 to .88, with a mean of .83, while Big Five composite estimates ranged from .87 to .92. These are correlations for those instruments, scores, participants, and study interval, not predictions for an unnamed online quiz. They also show why a report’s dimension and composite results should not be treated as interchangeable: combining related scores can yield a different consistency estimate than examining narrower scales.

A reliability coefficient is usually a correlation summarizing consistency in a group. “Test Reliability—Basic Concepts,” an ETS research memorandum, stresses that the coefficient refers to a group’s relative score positions and can differ with the group and the sources of error included. It cannot tell you the exact score you personally should expect on your next attempt. The standard error of measurement (SEM) is another group-based summary, expressed in score units; it describes typical measurement error under its stated assumptions, not a personalized guarantee. Read the instrument’s documentation for the population, interval, scale level, method, and uncertainty measure behind any reported figure.

For your decision, separate three questions. Would the same result tend to recur under comparable conditions? That is the retest question. Does the score support the report’s intended interpretation? That is a validity question. Would another administration resolve a specific uncertainty that matters to your next step? That is a practical-use question. A high internal-consistency estimate answers none of these alone, and a repeat result answers only part of the first. Seek evidence matched to the actual claim; do not use a reliability label as a shortcut from repeatability to accuracy.

Sources: Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity; Examination of the Test–Retest Reliability of a Forced‐Choice Personality Measure; Test Reliability—Basic Concepts

Could uncertainty change a decision enough to justify repeating the assessment?

A retake may be worth discussing when uncertainty around a score could change a specific decision. If the report places someone near a category boundary, a small difference might change the band label even while leaving the broader interpretation uncertain. If no practical decision depends on which side of that boundary the score falls, however, a new label may not add useful information. The first question is therefore not “Could I get a different result?” It is “What would I do differently if this score were interpreted another way?” Start by identifying what the report actually gives you. An observed score is the score recorded from this administration. A standard error of measurement (SEM) is an estimate, in the score’s units, of how much scores vary because of specified sources of inconsistency. In Test Reliability—Basic Concepts, the Educational Testing Service explains that an SEM describes consistency for a group and depends on which sources of inconsistency the estimate includes. A confidence interval, when the report provides one, expresses a range around an estimate under stated assumptions. It should be interpreted using the instrument’s documentation, including the confidence level and method, rather than treated as a guaranteed boundary around a person’s “real” personality. The APA Guidelines for Psychological Assessment and Evaluation call for understanding the technical meaning and limits of scores in relation to the purpose of an evaluation. Consider an illustrative report that assigns a person to one of two adjacent descriptive bands. The person’s score sits close to the dividing point. If the score were slightly different, the band might change, but that switch would not by itself show a meaningful change in the underlying tendency. Nor would it establish that the first band was an error: the boundary may divide a continuous measure for convenient reporting. The useful question is whether the distinction changes a real action, not whether the report can supply a more satisfying category. For example, if either band leads to the same next step in self-reflection, the label difference may have little practical value. If the bands trigger different treatment by an organization, the administrator should explain the decision rule and uncertainty before the score is used. Use a practical decision rule. First, name the decision that depends on the result. Next, ask whether the report provides an SEM, confidence interval, or classification information for the relevant score, and what population, scale, and conditions those figures describe. Then ask the provider whether a repeat is an accepted way to address that uncertainty, and how a second score would be interpreted. Confirm that any precision estimate applies to the named scale and score, rather than a different measure or population. If plausible uncertainty would not change the action, or if the provider offers no supported method for resolving it, retaking may simply replace one uncertain label with another. A second observed score does not automatically narrow uncertainty; its meaning depends on the measure, the repeat protocol, and how both administrations are combined or considered. If the result could alter a consequential interpretation and the administrator says a repeat is appropriate, discuss that procedure before acting.

Sources: Test Reliability—Basic Concepts; Guidelines for Psychological Assessment and Evaluation

What can a changed score mean besides ‘the first test was wrong’?

A changed score can reflect measurement error, different responding, a changed situation or reference frame, or genuine change over time. The comparison alone cannot identify which explanation is correct. Both results may accurately record responses made under different conditions without either serving as a complete account of the person.

A retest compares scores across occasions; it does not explain why they differ. “Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity,” a review of NEO Inventory research, distinguishes temporary state variation from other sources of unreliability. The authors note that an item asking how someone feels “nowadays” may draw a different response from one asking about childhood, because the first time frame is more open to change. This helps explain response variation; it does not prove that a particular person’s trait changed between attempts.

The same score movement can have different explanations. Someone may misunderstand an item, answer carelessly, or interpret it differently on the second occasion. Alternatively, an item may ask about a current pattern, and circumstances or self-observation may have changed. Stress, a new responsibility, or a different routine are context to record, not evidence that a trait shifted predictably. A repeat gives another observation under another set of conditions; it does not isolate the cause.

The instrument and score level matter. In “Examination of the Test–Retest Reliability of a Forced-Choice Personality Measure,” 743 adults recruited through Amazon Mechanical Turk completed a computerized adaptive forced-choice measure and the Big Five Inventory Two (BFI-2), a Likert-type questionnaire, twice. The interval averaged about 23 days and ranged from 19 to 45 days. Retest correlations for the forced-choice measure ranged from .45 to .77 across 13 narrower dimensions, averaging .63; its broader Big Five-related composites ranged from .63 to .81. BFI-2 dimension estimates ranged from .74 to .88, averaging .83, while its broad composites ranged from .87 to .92. These sample-level associations describe those measures, participants, interval, and scoring methods. They are not an individual score forecast or a prediction for an unnamed online report.

A correlation summarizes how scores co-varied across study participants; it does not reveal why one person’s result differed. The study reported different estimates for narrower dimensions and broader composites, so those levels should not be treated as interchangeable. The NEO review also explains that item wording and time frame can affect retest consistency. Together, these findings support a limited inference: interpret change at the level the report provides, rather than turning one label or scale into a conclusion about the whole person.

Labels can conceal change or exaggerate it. If a report groups scores into bands, two different scores may receive the same label, while modest movement near a boundary may produce a new one. These are general effects of categorizing scores, not claims about any report’s cutoffs. Without its scoring rules and uncertainty information, a changed label does not establish a large shift; an unchanged label does not show every response stayed the same.

To compare results, record the dates, measure and version, instructions’ time frame, conditions, and which score or label moved. Ask whether the response frame stayed consistent and what else differed. Distinguish a narrow scale from a broader composite, and note life context as a question to investigate. If the administrator provides a retest procedure, use it to learn whether the scores are intended for comparison. The difference is evidence to interpret, not proof that the first result was wrong or that the later one is the person’s “real” profile.

Sources: Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity; Examination of the Test–Retest Reliability of a Forced‐Choice Personality Measure

When is there a concrete reason to question the first administration?

A retake is more defensible when you can name a problem with the first administration, explain how it could have affected your answers, and confirm that the instrument permits another attempt. The question is whether something made the recorded responses a poor basis for the intended interpretation. Disliking the result, without a specific administration concern, does not establish that repeating it will clarify anything. Begin with the instructions and session record. If you misunderstood an item, answered for a different time frame than it specified, or completed only part of the assessment, describe exactly what happened. Ask the publisher or administrator whether this makes the attempt unusable and whether a repeat is the prescribed remedy. Do not assume it invalidates the result: the instrument’s documentation and use determine what can be concluded. Treat a technical interruption with the same care. Check whether answers were saved, items were skipped, or the session resumed at the right point. The Standards for Educational and Psychological Testing research memorandum on reliability explains that score consistency is evaluated under specified conditions and procedures. It does not establish that every interruption changes a score or requires a retake. Ask the administrator how the event is handled before starting again. Separate what happened from what you infer. “The session ended before I could answer the remaining items” is a concrete event. “I was tired, so the result must be wrong” is a possible explanation, not proof of invalidity. If you were rushed, distracted, or distressed, report that plainly and ask whether the provider’s rules consider it relevant. The American Psychological Association’s Guidelines for Psychological Assessment and Evaluation call for attention to purpose, administration context, and possible error when scores are interpreted. They support checking the procedure; they do not supply a universal personal retest checklist. Then ask whether a repeat is allowed in this setting. The provider may specify a waiting interval, an authorized alternate form, a correction for an incomplete administration, or an accommodation process. In a workplace or coaching process, the administrator may also have rules about timing and use. Confirm these before acting if another attempt could become part of an ongoing decision. This procedural check protects the comparison. If you repeat the assessment after deciding which answers would produce a more appealing description, the second attempt reflects responding after seeing the first report. It cannot cleanly show whether the original session was compromised. The research memorandum describes reliability as consistency across defined conditions, not a contest in which the preferred result wins. Disclose changed instructions, format, accommodations, or response frame rather than treating unlike attempts as equivalent. Before repeating, ask: What specifically went wrong? Which answers or interpretation could it affect? What does the instrument’s documentation or administrator recommend? Could a repeat affect an ongoing decision? If a specific problem is documented and the responsible source authorizes a remedy, follow that procedure and retain the relevant details. If no concrete administration issue emerges, pause. A more comfortable label is not evidence that the first attempt failed; examine the report’s claim and compare it with observable examples.

Sources: Guidelines for Psychological Assessment and Evaluation; Test Reliability—Basic Concepts

How long should you wait, and what makes two attempts comparable?

There is no universal waiting period established for every personality assessment. The useful interval depends on the instrument’s directions, whether item memory matters, how stable the measured construct is expected to be over that span, and whether meaningful change is plausible. Follow the manual or administrator’s procedure and keep the response frame consistent where the purpose requires comparison.

A fixed interval can sound more precise than the evidence allows. The Defaults article, “How Often Should You Retake a Personality Test? The Test-Retest Answer,” recommends waiting about six months in some situations. That is secondary editorial advice, not a rule established for every measure. Its cited studies address personality development and change, but do not establish six months as the right threshold for every instrument, population, or purpose. Treat it as one publisher’s recommendation, not a psychometric standard.

The interval changes what a comparison can mean. A short gap may reduce the chance that a broad tendency has changed, but answers may be influenced by remembering items or the previous response frame. With a longer gap, item memory may matter less, while circumstances and the measured tendency have had more time to change. Neither possibility makes one interval automatically best. The review “Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity” explains that state variation can affect scores across occasions and that an item’s time frame matters: a question about current behavior can reasonably invite different answers from one about a distant period. Its discussion concerns personality scales and research designs, including NEO measures; it does not tell an individual when to retake an unnamed online assessment.

Check the instructions for the exact assessment and version. A publisher may specify an interval or restrict repeat attempts; an administrator may control whether a repeat is permitted. Ask whether an alternate form or another run supports a valid comparison. If no schedule is documented, ask what interval fits the measure and your question. A research study’s interval describes its design; it is not automatically advice to consumers.

For a useful comparison, preserve the response frame. If items refer to a stated period, use that period on both occasions rather than answering from this week first and from a longer stretch next time. Keep the version and administration mode the same when directions call for it; record any authorized alternate form. Note dates and major contextual changes. These details do not explain a score change alone, but help distinguish a comparable repeat from a different question about another period or context.

Do not deliberately adjust answers to produce a more appealing profile. That changes the response process, so the second output cannot show whether the first result would repeat under comparable conditions. If your perspective has genuinely changed, answer from that perspective and label the comparison accordingly. A difference may reflect a changed frame, circumstances, measurement variation, or a combination; the score alone cannot sort these explanations.

The practical rule is to request an instrument-specific interval, then record the version, item time frame, administration conditions, date, and relevant context for any authorized repeat. If conditions differ, interpret the second attempt as evidence from another occasion, not a correction that automatically replaces the first. A retake is informative only when its procedure and purpose make the comparison interpretable.

Sources: The Stability of Personality Traits From Childhood to Adolescence: Links to Parent and Child Reports; Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity; How Often Should You Retake a Personality Test? The Test-Retest Answer

Does the reason for taking the assessment change the answer?

Yes. The purpose of the assessment changes what a useful next step looks like. For private self-reflection, you can often leave the score alone and check whether its description fits specific behavior across situations. In coaching or development, the report may prompt questions, but discussion should stay tied to the stated goal and the instrument’s limits. In selection or another consequential process, the assessment sits inside a procedure with a defined purpose and rules. Ask the responsible administrator how results are interpreted and whether a repeat is allowed; do not use another attempt to search for a more favorable profile. The American Psychological Association’s *Guidelines for Psychological Assessment and Evaluation* describe assessment as a decision process that can integrate test scores with relevant collateral information, such as observations or reports, and emphasize the purpose for which an instrument is used. The guidelines address psychologists’ professional practice, not consumer retest decisions. Their practical lesson is that a score should be interpreted in relation to the question it was meant to inform. If you took a voluntary questionnaire to reflect on work habits, ask what examples support or challenge the description. Disagreement may point to an overly broad interpretation or a context the report misses. Repeating the questionnaire alone will not settle that. Coaching adds another person and purpose to the interpretation. Before repeating a measure, clarify what decision the coaching is meant to support and whether a new score would change the conversation. The APA guidelines call for understanding score limits and integrating relevant information. A coach can translate a report statement into an observable question: when does a preference for planning help, and when might it slow a task? These examples prompt discussion; they do not diagnose a problem or establish character. If the concern is a particular work difficulty, consider the task, expectations, and context alongside the report rather than asking one score to explain everything. Formal evaluations require care because others may use the result. The APA’s *Professional Practice Guidelines for Occupationally Mandated Psychological Evaluations* say evaluators should understand the referral question and use methods and information relevant to it. They also recommend multiple relevant sources rather than relying on a single source. These guidelines concern occupationally mandated evaluations conducted by psychologists; they do not establish that every workplace questionnaire is such an evaluation. Ask who commissioned the assessment, what use was disclosed, who receives the result, and what rules govern a repeat. If an employer or evaluator arranged the assessment, ask those questions before repeating it. A second attempt may join the record, not replace the first. Follow the administrator’s directions and answer honestly; steering the profile makes comparison less useful. A general personality result alone is not a diagnosis, complete account of character, or job recommendation. The decision rule is simple: for reflection, check the claim against behavior; for coaching, connect it to the agreed goal; for selection, clarify procedure and intended use with the administrator. Retake only if that process identifies a legitimate, permitted reason that could change the interpretation.

Sources: Guidelines for Psychological Assessment and Evaluation; Professional Practice Guidelines for Occupationally Mandated Psychological Evaluations

What should you do instead when a retake has no clear purpose?

When a retake has no clear purpose, keep the first report as limited evidence and test its wording against observable examples. Translate one disputed phrase into behavior you can describe without the report’s label. Note one example that fits and one that does not. This tests how well the interpretation travels across situations; it does not create a new score or prove the assessment valid. The APA’s 2020 professional guidelines support interpreting results in light of purpose, possible error or bias, and other relevant information. They do not prescribe this exercise as validated. Ask whether the disagreement depends on context or the time frame in the questions. Record that difference instead of treating answers as comparable. For unresolved questions about scoring, repeat rules, or an interrupted attempt, ask the publisher or responsible administrator. The NEO measurement review describes how transient variation and time frame can affect retest consistency, but cannot identify the cause of a personal difference. Before repeating, state what the second attempt should answer and whether an authorized, comparable retest could change a real decision. If yes, follow the documented procedure, retain both reports and dates, and interpret movement cautiously. If there is no administration problem, repeat protocol, or decision that depends on resolving uncertainty, stop. Retake for a defined measurement or use question, not to vote a disliked result away. A documented problem or consequential uncertainty the administrator says a repeat can address would change that judgment.

Sources: Guidelines for Psychological Assessment and Evaluation; Internal Consistency, Retest Reliability, and their Implications For Personality Scale Validity

Questions readers ask