Skip to main content

Verified by Psychology Today


How a Test Labels You as Introvert, Extravert, or In-Between

How do psychologists interpret personality test scores?

How does a psychologist determine from your answers to personality questions whether you are an introvert, extravert, or somewhere in between? (The same question can be asked about other personality types, but we will use the familiar Introvert-Extravert categories to provide a general answer that can be applied to other types.)

First, it is important to realize that most personality psychologists prefer to describe personality in terms of traits rather than types. For our example, psychologists see introversion-extraversion as a matter of degree on a continuous scale rather than an either/or placement into the Introvert and Extravert type categories. One widely-used measure of introversion-extraversion is a set of 10 items from the International Personality Item Pool (IPIP). If you browse to the page with those items, you will see the following:

+ keyed Am the life of the party.
Feel comfortable around people.
Start conversations.
Talk to a lot of different people at parties.
Don't mind being the center of attention.

– keyed Don't talk a lot.
Keep in the background.
Have little to say.
Don't like to draw attention to myself.
Am quiet around strangers.

Respondents are asked to check a box to indicate how accurately each item describes them. Response options are:
1 very inaccurate
2 inaccurate
3 neither accurate nor inaccurate
4 accurate
5 very accurate

You do not need a Ph.D. to see that describing the +keyed items as accurate and the –keyed items as inaccurate would indicate extraversion. (The reverse would be true for introversion. You could score these 10 items either in the direction of extraversion or introversion—that is totally arbitrary. Let's score them in the direction of extraversion, which is the usual convention.)

A person's score for extraversion is literally the sum of the numbers assigned to his or her responses about the accuracy of each item. For +keyed items, a "very inaccurate" response is scored "1" while a "very accurate" response is scored "5." The reverse is true for –keyed items (sometimes called "reverse-scored" items); a "very inaccurate" response is scored "5" while a "very accurate" response is scored "1."

If the items are scored this way (which is the usual method for IPIP scales), possible scores can range from 10 to 50. (There are other ways of scoring, such as assigning the numbers 0 through 4 to responses, which gives a range from 0 to 40, or computing the average response to items, which gives a range from 1 to 5. This does not affect score interpretation because the different scoring methods are just linear transformations of each other.)

If scores range from 10 to 50, it might seem that a score of 30, the exact midpoint, would be the line for separating Introverts from Extraverts. Scores lower than 30 could be used to assign you to the Introvert category, and scores greater than 30, to the Extravert category. But virtually no psychologist interprets scores this way. (One exception is a Dutch psychologist, Wim Hofstee, who has presented some very good arguments for taking the midpoint of a scale at face value.)

Instead, what most psychologists do is to compute the average score on a personality scale for a very large sample of respondents and compare individuals to that average score to describe the person as relatively introverted or extraverted. As a hypothetical example, let's say that we have responses from a million people who completed this measure of introversion-extraversion, and the average score was actually 35. This would suggest that scores lower than 35 indicate introversion, and higher than 35, extraversion.

But there is another consideration here. Is 34 really significantly lower than 35, low enough to call someone an Introvert? Is 36 really significantly higher than 35, high enough to consider a person an Extravert? Most psychologists say no. Scores must differ by a considerable amount to describe someone as an Introvert or Extravert. If a score is too close to average, psychologists refrain from using the Introvert or Extravert label, perhaps describing the person as an "Ambivert." But what is a "considerable amount?"

This is where things get interesting. For reasons that have never been fully explained, there is a convention in psychometrics (the measurement of psychological characteristics) that scores that deviate by one-half of a standard deviation from average can be interpreted as greater than average (in this case, extraverted) or less than average (in this case, introverted). (The standard deviation is an index of how spread-out or bunched-up scores are. It can easily be computed with a formula if you are interested.)

Even though I have used this convention all my life, I have always wondered whether this convention is a valid way to categorize scores that are significantly higher or lower than the average score on a personality scale. Because there is no absolute method for determining whether one is an Introvert or Extravert, we can't answer the question definitively. But it has been suggested that the best, most objective way to describe personality is to have a group of people who are well-acquainted with a person provide judgments of personality. Those judgments can then be averaged, which hopefully cancels out any idiosyncratic biases. Averaged acquaintance ratings can then be compared to scores on personality scales such as our introversion-extraversion scale to see how far a score on the self-reported personality scale must differ from average before knowledgeable acquaintances are willing to describe the person as introverted or extraverted.

I recently published such a study, comparing scores on the major personality dimensions from the Five-Factor Model (introversion-extraversion, disagreeableness-agreeableness, unconscientiousness-conscientiousness, neuroticism-emotional stability, and openness-closedness to experience) from an IPIP inventory of those traits to personality judgments made by knowledgeable acquaintances. This article, listed in the Reference section below, indicates that one-half of a standard deviation is not enough distance from average to define someone by one of the labels of each personality trait dimension.

Wim Hofstee/Researchgate
Wim Hofstee
Source: Wim Hofstee/Researchgate

This is something that Wim Hofstee predicted. Professor Hofstee seems to be right about a lot of things, so maybe people should start listening to him.

The data also indicate that a vast majority of us are close enough to the average value on personality scales such that applying one of the labels is inappropriate. This is a problem for inventories such as the Myers-Briggs, which attempts to categorize absolutely everyone with one of two terms from each of its personality dimensions. It might be wiser to simply talk about the way that people lean toward one or the other end of a personality dimension, rather than sort people into either-or categories.

For up to 50 days after the publication of this post, you can access a copy of the article for no charge by following this link. After that, you'll have to get it from a library that carries the journal.


Johnson, J. A. (2021). Calibrating personality self-report scores to acquaintance ratings. Personality and Individual Differences, 169, 109734.

More from John A. Johnson Ph.D.
More from Psychology Today
More from John A. Johnson Ph.D.
More from Psychology Today