Should You Choose a Personality Report With Observer Forms?

Should you choose both forms when the key behavior is unseen?

Usually, no. If the behavior you most want to understand is private or happens where a proposed observer is absent, that person’s form cannot directly verify it. A self report is the closer source for private experience. Choose both forms when you also want to examine a recurring outward pattern in settings the observer shares, or when the difference between your view and another person’s view is itself useful.

Consider the decision in concrete terms. Someone wants to understand how often they rehearse a contribution silently before speaking in a meeting. A colleague may see whether the person speaks, interrupts, asks questions, or waits. The colleague does not see the rehearsal unless it is described afterward. A self form can ask about that internal process; an observer form can speak to the visible meeting behavior. These are related but distinct pieces of information.

An observer or informant report is a questionnaire in which another person rates the target using items designed to describe that person. Parallel forms may use similar questions, but matching wording does not make the observer’s access equal to the target’s. Nor does a second score operate as a correction key. The choice depends on what you are trying to learn: private experience, visible conduct, or how the two accounts compare.

Research supports a conditional answer. In a study of 165 participants rated by themselves, friends, and strangers, accuracy varied by trait: self ratings were strongest for neuroticism-related traits, friends for intellect-related traits, and perspectives were similarly good for extraversion-related traits. A recent review reports substantial convergence between acquainted people on full-length Big Five and HEXACO questionnaires, but that does not mean a familiar person can directly observe a concealed event. The useful question is not whether observers are generally good or bad. It is whether this observer had evidence relevant to this behavior.

So before paying for a paired report, name one behavior and one setting. If the important part happens only in your thoughts, an outside rating may add little about that part. If the question includes how you act in shared situations, a well-chosen observer may contribute a second view. Keep those purposes separate as you compare forms.

Sources: Who knows what about a person? The self-other knowledge asymmetry (SOKA) model; Self- and Observer Reports of Personality

When can an observer know the behavior you want to understand?

An observer can report what they have had repeated opportunities to notice. Their access is usually more direct for outward actions in shared settings than for private thoughts, feelings, or reactions that occur elsewhere. Familiarity matters, but the kind of familiarity matters more: knowing someone for years does not guarantee seeing the particular behavior at issue.

In personality research, observability describes how available a tendency’s signs are to another person. It does not mean that a trait is always obvious or that a private state has no consequences. Self other agreement means that two reports correspond; agreement is not by itself proof that either account captures an objective essence. These distinctions prevent a questionnaire from turning an inference into an observation.

The Self Other Knowledge Asymmetry study tested who judged which aspects of personality more accurately. The 165 participants supplied self ratings and were rated by four friends and up to four strangers in a round robin design. The participants also completed behavioral tests that provided criterion measures. The reported pattern matched the model’s predictions for the traits examined: the self was the best judge of neuroticism related traits, friends of intellect related traits, and self, friend, and stranger perspectives were similarly accurate for extraversion related traits. This is useful evidence that information advantages can vary with what is being judged.

Its scope matters. It was one study with particular traits, informants, and behavioral criteria, not a validation of every observer form or every commercial report. “Neuroticism related” does not mean all private behavior is unknowable to others, and “friends best judged intellect” does not mean friends can assess every skill accurately. The finding supports a question about access, not a universal ranking of raters.

A 2024 study of 101 married dyads provides a different, more specialized illustration. It examined self other agreement and relationship adjustment among nonclinical, transplant, and divorcing groups using psychological problem measures. Agreement varied with relationship group and visibility. Because the sample included specific marital and clinical or custody related contexts, it cannot set expectations for ordinary personality reports. It does, however, show why relationship context can matter when interpreting two accounts.

A practical access test is simple: could this person name more than one occasion when the behavior appeared, describe the setting, and recall what you did? If they would need to guess from your general reputation, their answer is an impression rather than a direct account of that behavior. That impression may still matter if you want to know how you come across. It should not be mistaken for evidence about an unobserved inner process.

Sources: Who knows what about a person? The self-other knowledge asymmetry (SOKA) model; Assessing self-other agreement and dyadic adjustment in marital dyads

What does a second form add, and what does agreement mean?

A second form adds another person’s sampled perspective. That perspective can complement self knowledge when the observer sees the target behavior, and it can answer a separate question about reputation or interpersonal impact. Agreement means that two reports converge to some degree. It does not automatically establish criterion validity, while disagreement alone does not identify which account is wrong.

The 2025 Annual Review synthesis describes findings for close acquaintances, including spouses, relatives, or close friends, using full length Big Five or HEXACO factor measures. It reports that self and observer mean scores tend to be comparable, with self reports tending higher on Openness to Experience. Self observer agreement tends to be rather high, though lower for some cooperation related traits. The review also notes some similarity and assumed similarity between acquainted people on particular traits. This is a useful baseline: familiar informants often broadly converge on established measures, but convergence varies and is not the same as independent verification of a hidden event.

Accuracy can refer to different outcomes. Agreement among observers, correlation between self and other ratings, and prediction of behavior are separate questions. Multiple observers may agree while sharing the same limited view. A self other correlation may show rank ordering across people, not whether a particular person’s private memory is correct. Prediction of an external behavior asks something else again. A paired report should say which kind of evidence supports its interpretation.

These findings make paired forms potentially useful without making them mandatory. Suppose the target is whether a person tends to leave decisions open in team discussions. If a teammate regularly attends those discussions, their account could add visible examples and indicate how the pattern lands. If the target is how uncertain the person feels while weighing options alone, the teammate’s rating is indirect. A useful report keeps those two claims apart rather than averaging them into a single stronger sounding score.

Agreement can strengthen confidence only in a bounded sense. If two people independently report similar patterns across relevant occasions, the reader has reason to treat that pattern as worth checking against examples. But reports can share a source of bias: the same relationship expectations, a shared interpretation of an item, or repeated discussion of the result. An observer may also influence the person’s behavior, so the reports may concern a pattern produced partly by their interaction.

Therefore ask what the second form is meant to add. If you want to compare self view with how a familiar person experiences your visible conduct, paired reporting can make that contrast explicit. If the only question concerns a private event the observer cannot see, agreement on a broad trait label does not settle it. The additional score is useful only when its information source is relevant to the decision.

Sources: Self- and Observer Reports of Personality; Self/observer agreement in personality assessment by observers’ relationship types

Are observer ratings a more objective correction?

There is no general rule that an observer rating is a more objective correction to self report. Each source has different access and possible distortions. The relevant evidence depends on the behavior, the observer, the instrument, the comparison being made, and the report’s intended use.

A meta analysis published in 2019 compared self and informant report means for Big Five traits across 152 samples and 33,033 targets. Its average standardized difference was −.038, and the authors reported that self report means generally did not differ from informant means. Moderate mean differences appeared when self reports were compared with stranger reports. This challenges the broad claim that people typically describe themselves more positively and therefore need an observer to correct them. The analysis compared averages across studies; it cannot tell whether a specific informant knows one person’s private experience.

A 2016 study offers a counterpoint of a different kind. It examined informant reports on adult personality pathology inventories and the five general NEO personality factors in a community sample of 1,449 adults. Informant forms showed modestly higher internal consistency in the studied measures. Internal consistency asks whether items intended to measure a scale tend to move together. It does not show whether a rater has seen a particular event, whether the scale is valid for a particular decision, or whether an observer score better represents a person’s lived experience. The authors also cautioned that informants with less knowledge may answer in generic ways, losing nuance or accuracy. The community sample and general NEO factors mean the study is not confined to pathology inventories, but its results still do not establish a universal advantage for observer reports or access to private behavior.

This is why “objective” should be used carefully. A person outside the target can be less influenced by the target’s private interpretation, but may also lack relevant evidence. The target can report internal experience, but may overlook how a repeated action affects others. Neither perspective is free from context. Memory is selective; item words can mean different things to different raters; a person may be seen in only one role; and a relationship can shape what each notices. These are ordinary limits of measurement, not signs of deception.

The strongest case for observer forms is that behavior is often social and that outside reports can contribute information unavailable through self report. The strongest exception is when the central criterion is inaccessible to that observer. Both can be true. A report can benefit from an observer for visible collaboration while remaining unable to independently verify silent worry before the meeting. A higher reliability statistic for one scale, or close agreement in a meta analysis, does not override that basic mismatch in evidence.

Do not treat the ratings as competing witnesses in a trial, or as numbers that must be averaged. First examine what each person had a chance to know. Then look for the instrument’s evidence for the specific form and use. If the provider offers only a paired package but no description of how observer ratings are interpreted, that missing explanation is a reason to ask questions before relying on the result.

Sources: Self-Other Agreement in Personality Reports: A Meta-Analytic Comparison of Self- and Informant-Report Means; Are Informant Reports of Personality More Internally Consistent Than Self Reports of Personality?; Self- and Observer Reports of Personality

What counts as a useful observer?

A useful observer is someone who has repeated, reasonably recent access to the settings and behaviors covered by the items and can answer them as written. Closeness can help, but it is not a substitute for access. A close friend may know what you disclose privately yet rarely see your work routines; a colleague may see meetings but not solitary preparation or recovery afterward.

Check whether the form specifies a time window and context. “How often does this person change plans?” may invite different answers if one rater thinks of work and another of weekends, or if one considers the last month while another draws on years of memory. If the self form says “I usually…” and the observer form asks “this person usually…,” the parallel phrasing helps, but the forms still need clear referents and evidence that they are intended to be compared.

Research summarized in the 2025 review concerns close acquaintances and established full length trait measures. It reports broad convergence on average, while also noting some assumed similarity between acquainted people for particular traits. That is a reminder to distinguish a second view from a fully independent one. A spouse or friend may share experiences and expectations with the target; that can make the observer informed, and can also make the two reports less independent.

The marital dyad study is a warning against turning “close” into “unbiased.” In 101 married pairs drawn from nonclinical, transplant, and divorcing groups, self other agreement varied by relationship group and visibility. Because this was a specialized sample examining psychological problems and relationship adjustment, it should not be treated as a personality norm. It supports a narrower point: the relationship setting can affect how ratings relate, so the observer’s stake and the nature of contact deserve attention.

Before asking someone to complete a form, state the behavior and setting aloud. Ask yourself whether they have seen several relevant occasions, whether they can distinguish the action from its presumed motive, and whether they can respond without being pressured to confirm a desired result. The point is not to recruit a perfect judge. It is to avoid giving a score the appearance of direct evidence when the observer would have to infer most of it.

If no one has the needed access, do not select a nominal observer merely to complete the package. For personal reflection, you can start with your own report and keep a short record of relevant situations: what happened, what you did, and what you were thinking or feeling. If you want feedback on visible behavior, you might invite a trusted person to discuss specific examples, with their consent and without demanding a score. In work settings, use proportionate, transparent feedback practices rather than covert monitoring. This is a practical application of the access principle, not a validated assessment procedure.

Sources: Self- and Observer Reports of Personality; Assessing self-other agreement and dyadic adjustment in marital dyads

How should you compare forms before choosing one?

Compare forms by fit to the question, not by counting how many perspectives they include. Start with the instrument’s documentation. The International Test Commission’s Guidelines on Test Use say that reliability and validity evidence should support a test’s intended purpose, that evidence for score inferences should be accessible for scrutiny, and that test use should be appropriate to the context. The guidelines are professional guidance, not a trial comparing self only with paired reports, but they provide a sound standard for what a provider should explain.

First, check what each form measures and how its items are worded. Are self and observer items parallel? Do they name the same time period and setting? Does the report distinguish private experience from observable actions, or does it use one broad label for both? A form can be well presented and still leave the reader unclear about whether a score describes what the person reports, what an observer sees, or an interpretation built from both.

Second, ask whether the provider reports evidence for each form separately and for the combined interpretation. Reliability concerns the consistency or precision of scores under specified conditions. Validity concerns whether evidence supports the interpretation and use being made. Neither can be inferred simply because two forms exist. Evidence for a self questionnaire does not automatically validate its observer adaptation, and a validation study of one language or population may not settle use with another group. The article’s scope is form choice for low stakes self understanding, not selection or diagnosis.

Third, inspect how the report handles differences. Does it show the two perspectives separately before offering a synthesis? Does it explain what can and cannot be inferred from a gap? Does it invite checking concrete examples, or does it label one rater accurate and the other biased without a suitable criterion? Ask how missing responses, unfamiliar items, or insufficient observer contact are handled. If the report does not say, treat its interpretation as less transparent.

A self only report is often the better fit when the target is private experience, the observer has no relevant access, or the second view would add cost and burden without advancing the decision. Paired forms are a better fit when visible behavior in shared contexts matters and the observer can speak to it, or when your explicit question is how your self view compares with another person’s experience. They are a poor fit when the observer is chosen only because the package includes a second form.

This is not a contest where one form wins in every circumstance. The 165 person SOKA study suggests that the comparative value of perspectives differs by trait and criterion. The 2025 review and large meta analysis show that close informants and self reports often converge on broad trait measures. The 2016 internal consistency finding shows that one form may score better on one psychometric property in a bounded study. Taken together, these findings argue for matching the form to the question, then checking form-specific evidence. None establishes the quality of an unnamed product.

Before buying, a reader can ask the provider four plain questions: What evidence supports the observer form? Who should complete it, and what contact should they have with the relevant situations? How does the report interpret differences? Is the result intended only for reflection, or does the provider claim another use? A provider’s clear answer helps distinguish a complementary perspective from an extra score with uncertain meaning. If claims extend to hiring, promotion, diagnosis, or another consequential decision, the evidence and safeguards required are substantially different; an informal paired report should not be treated as sufficient on its own.

Sources: ITC Guidelines on Test Use; Who knows what about a person? The self-other knowledge asymmetry (SOKA) model; Self- and Observer Reports of Personality

How can you read a self other gap without forcing a verdict?

Treat a difference between forms as a question about perspective, context, and item meaning. Do not read it as proof that someone is lying, lacks insight, or has a fixed flaw. A score difference is an observation about reports. Its cause remains open until the underlying behavior and conditions are examined.

The 2019 meta analysis is useful here because it tested a popular explanation: that self reports are generally inflated by self enhancement. Across 152 samples and 33,033 targets, average self and informant Big Five means were nearly alike; the reported average standardized difference was −.038. Stranger comparisons showed moderate differences. This does not mean every person agrees with every observer. Mean similarity across many targets can coexist with individual differences, and it does not identify who is right in a specific disagreement.

The 2025 review adds that agreement among close acquaintances is generally substantial on full length Big Five and HEXACO factor measures, though some traits show lower convergence and some similarity between acquainted people. A report reader should therefore avoid both extremes: assuming disagreement is inevitable and assuming agreement settles the truth. Similarity may reflect accurate shared observation, shared assumptions, or both. The abstract-level synthesis does not establish which explanation applies to a particular pair.

A careful reading begins with the item. Does “I plan ahead” refer to private preparation, meeting deadlines, or telling others about a plan? Then consider exposure. One person may see only the last stage of a task, while another sees the early planning. Next identify the reference period. Different memories can produce different ratings without either person deliberately distorting anything. These checks convert an abstract gap into a narrower question that can be tested against examples.

Keep the levels of claim separate. A self form might support “I often experience uncertainty before a decision.” An observer form might support “My colleague sees me ask for more time in meetings.” A report that infers “I am indecisive” has moved from two reports to a broader interpretation. That label should be evaluated against repeated situations, not accepted simply because a score graph places the profiles side by side.

Use an open question if the other perspective matters: “Can you recall a recent time you noticed this?” or “When I ask for time, what do you see me doing?” Compare the answer with your own account of the same occasion. One concrete exchange may reveal that both accounts are accurate at different stages: the target feels uncertain, while the observer sees a deliberate pause. This is an illustrative way to reason about a gap, not evidence that any particular assessment result means this.

If disagreement persists, leave the disagreement visible. Look for repeated examples across settings, and ask whether the report’s scale is broad enough to combine behaviors that differ by context. The specialist should be the instrument manual or a qualified practitioner when the score is being used in a formal assessment; for ordinary reflection, a short observation log or a respectful conversation may be enough. Avoid averaging scores unless the instrument’s documented scoring procedure calls for it. Averaging can hide the very context difference the two forms reveal.

There are limits to this approach. A private account can also be incomplete: people may not recall every occasion, and a self report captures reported experience rather than direct access to all behavior. An observer can notice indirect signs, but those signs do not establish an internal state or its cause. The practical conclusion is modest: use each form for the kind of evidence it can provide, then check important interpretations against specific situations. That is more informative than appointing either reporter the winner.

Sources: Self-Other Agreement in Personality Reports: A Meta-Analytic Comparison of Self- and Informant-Report Means; Self- and Observer Reports of Personality; Assessing self-other agreement and dyadic adjustment in marital dyads

What is the proportionate next step?

Return to the original decision by writing one sentence: “I want to understand whether I ___ when ___.” Then mark whether the behavior is private, visible, or mixed. Name the people who actually see the relevant setting. If no one does, a paired report cannot supply a direct observer account of that target; use self reflection or a method that collects relevant observations transparently. If someone does, paired forms may help you compare their view with yours, provided the instrument documents how both are interpreted.

The verdict is conditional but clear. Choose a self and observer report when both perspectives serve the question and the observer has repeated access to the behavior in context. Choose self report alone when the main question concerns an unshared internal experience or when no proposed observer has relevant exposure. If you want to learn how your actions affect another person, the observer form may still be useful even though it cannot verify what you privately felt.

The practical comparison is not “one form is accurate, two forms are more accurate.” It is whether the added form supplies evidence for the intended interpretation. Check the forms’ wording, time frame, populations and purposes documented by the provider, and how disagreement is handled. Research on Big Five and HEXACO measures supports using multiple perspectives thoughtfully, but it does not validate every paired report or turn it into a hiring score, diagnosis, or job recommendation.

For a low stakes work question, a self report can be a starting map rather than a verdict. The live Work Pattern Report at /assessment asks about ten decision and collaboration continuums and provides an individual self report for reflection. It has no observer form, norms, cutoff, or validation for employment decisions. It can help turn a vague question such as “Why do handoffs feel difficult?” into specific areas to notice, but the reader must compare those prompts with actual examples and workplace conditions. If you want to explore that kind of self observation, you can [take the Work Pattern Report](/assessment); for more report literacy, see [/topics](/topics).

A person choosing a paired report should not have to guess whether another person can see the behavior that matters. Start with the behavior and context, then choose the form that can speak to them. When no observer can see the central behavior, the honest next step is not to add a second score. It is to decide what evidence would make the private question clearer, and gather only what is relevant and appropriate.

Sources: ITC Guidelines on Test Use; Who knows what about a person? The self-other knowledge asymmetry (SOKA) model

Questions readers ask