Self-report vs observer report: who knows your personality better?

What is the difference between self-report and observer report?

Self-report means you rate yourself, usually on statements such as "I get anxious easily" or "I enjoy meeting new people." Observer report, also called informant report, means someone who knows you rates you on the same or a matched set of statements. Both are structured measurement methods, not shorthand for accurate versus inaccurate. Each rater has a different slice of the truth: you have your own inner states, and other people have your behaviour as they have seen it.

Personality research has used both since the Big Five model took shape, partly to check whether the same five factors show up regardless of who is rating. They generally do, which is one reason the model is treated as a structure of the person rather than an artefact of self-description (McCrae & Costa, 1987).

Who knows you better, you or someone who knows you well?

It depends on the trait, and the pattern is called the self-other knowledge asymmetry. In a study of 165 people who rated themselves, were rated by up to four friends and up to four strangers, and completed behavioural tasks that supplied outside criteria, the self turned out to be the best judge of neuroticism-related traits, such as how anxious or self-doubting someone is. Friends were the best judges of intellect-related traits, which carry a clear value judgement. For extraversion-related traits, such as sociability, every perspective did about equally well (Vazire, 2010).

The logic is straightforward. Anxious feelings and private self-doubt rarely show on the outside, so you have information nobody else does. Intelligence and similar evaluative traits are easier to judge from outside, partly because your own view is coloured by wanting to think well of yourself. Sociability and similar visible traits are on display to everyone, so nobody has a special advantage.

Do observer reports predict real outcomes as well as self-reports?

Across three meta-analyses covering 263 independent samples and 44,178 rated people, observer ratings predicted behaviour about as well as self-ratings, and better for some outcomes: for academic achievement and job performance, observer reports added real predictive validity on top of self-ratings (Connelly & Ones, 2010). Asking someone else is not a weaker substitute for asking the person themselves, and for some purposes it is the stronger source.

What made an observer accurate was not how often they saw you but how well they knew you. Intimacy mattered more than frequency of contact, and helped most for traits low in visibility, such as emotional stability, with only a small difference for evaluative traits such as agreeableness (Connelly & Ones, 2010). A friend of ten years tends to beat a colleague you see every day.

Why doesn't introspecting harder make you more accurate?

Because the limits are not about effort. Reviewing the evidence, Wilson and Dunn describe two sources of limited self-knowledge: personal motives, since there are things about ourselves we would rather not know, and the basic architecture of the mind, since much of what drives perception, personality and attitudes runs outside conscious awareness. Introspection is not a direct line into those processes, though some kinds of reflection do help people build workable accounts of themselves (Wilson & Dunn, 2004).

The same review points to two routes that outperform looking inward: seeing yourself through other people's eyes, and watching your own behaviour rather than your intentions. It also notes that a perfectly accurate self-view is not automatically an advantage. The useful account of who you are is one you can act on, built from more than one source. Self-awareness tests covers this ground in more depth.

Do self-report and peer-report measures agree with each other?

Substantially, though not identically. In a study of 738 peer ratings of 275 adults, using both adjective ratings and questionnaire scales, correlations between the mean peer rating and the person's own self-report ranged from .25 to .62 across the five factors, and the same five-factor structure was recognisable in both self-report and peer-report data (McCrae & Costa, 1987). That range is real agreement, not the near-unity you would see if the two methods measured identical things, and not the near-zero you would see if they measured unrelated things.

The honest reading is that self-report and peer-report each carry information the other is missing, which is what makes it worth collecting both when the stakes are worth the extra step.

How do you use self-report and observer report together?

• Ask something specific. A narrow question about a real situation beats a general one, which tends to get polite reassurance.
• Pick a close informant over several distant ones. Closeness predicted accuracy more than frequency of contact did (Connelly & Ones, 2010).
• Ask about what you cannot see. Let others judge your evaluative traits; keep your own report for the internal ground only you have access to (Vazire, 2010).
• Treat disagreement as data. A self-report and an informant's report landing in different places usually means each is picking up something real.
• Compare more than once. Two readings a year apart, from either source, show a direction one score cannot.

How does Wellington measure this?

Wellington is a science-based personality assessment from Therabot Labs LLC. It measures personality as continuous traits, built on the Big Five and HEXACO models, and writes the results back as a personal report rather than a type.

The free Snapshot is a self-report and nothing more: 60 questions, about 7 minutes, scoring 6 broad dimensions, Extraversion, Agreeableness, Conscientiousness, Emotional Stability, Openness to Experience and Honesty-Humility. Every trait is reported as a percentile against the reference sample. The three bands (Quiet below the 30th percentile, Balanced to the 70th, High above) and the Portrait names exist to make the result easier to talk about; the percentile is the measurement, and it is always shown alongside. Because it is a self-report, it carries the same blind spots any self-report does, so once you have read your bands, a reasonable next step is to ask someone who knows you well to read the same pages.

The Wellington Membership, $4.99 a month, extends the same self-report method to more traits rather than adding an outside rater. Wellington is a wellness tool for reflection, not a medical or psychological diagnosis, and it does not replace care from a qualified professional. You can export or delete your data at any time. The methods page sets out how the items are built and why the norms stay provisional until the norming sample is complete.

Questions people ask

Do other people know my personality better than I do?

For some traits, yes. Friends judge evaluative traits such as intellect more accurately than the person themselves does, while the self is the better judge of internal, hard-to-observe traits such as anxiety and self-doubt. For visible traits such as sociability, self and others do about equally well (Vazire, 2010). Neither side wins outright.

Should I trust a friend's opinion of my personality more than my own?

Alongside your own, not instead of it. Observer ratings predict real outcomes about as well as self-ratings, and better for some, including academic achievement and job performance (Connelly & Ones, 2010). A close friend's view is most useful for traits you struggle to see in yourself, and adds less for traits already visible to everyone.

Is an observer report more objective than a self-report?

No, it carries its own limits rather than none. An observer only sees your behaviour in the settings you share, and their judgement is coloured by how much they like you and how evaluative the trait is. Observers add real information, particularly for low-visibility traits, when they know you well rather than merely often (Connelly & Ones, 2010).

Does Wellington ask people who know me to rate me too?

No. Wellington's Snapshot and Wellington Membership are both self-report, like most personality assessments people take on their own. That is worth knowing plainly, because self-report and observer report each catch things the other misses. Asking someone who knows you well whether your bands match what they would say is a reasonable way to close some of that gap.