Guide
Personality Test Exercises for the Classroom
Put “Conscientiousness: 65%” on the board. Ask what it tells the class. Does it mean that someone is 65% conscientious? That they scored higher than 65% of other people? Or that their answers produced a score that the website converted to a percentage?
The number alone cannot settle those questions. In this lesson, learners read an invented report, calculate two comparisons, and see how category labels can hide differences between scores. The aim is to explain what a result means before deciding what to do with it.
All the values are fictional. Learners can discuss the reasoning without taking a personality test or sharing anything about their own traits.
Set up the lesson
Plan for about 15–20 minutes, with more time if percentages are new to the group. Students need the handout or a shared screen, paper, and optionally a calculator. They should be familiar with averages and percentages; the lesson explains the difference between a numerical score and a category label.
By the end, learners should be able to explain why a scale value differs from a percentile, why a different reference group can change the percentile of the same score, and why two category labels do not reveal the distance between scores.
- Read the report together, about 3 minutes. Explain that the values are invented. Ask what the scale tells learners and what they would still need to know about a real test.
- Work through the reference groups, about 5 minutes. Learners count the values below and equal to 3.6, then compare their reasoning with a partner.
- Compare the labels and rewrite the report summary, about 5 minutes. Ask for an explanation, not just the correct number.
- Discuss the answers and try the new case, about 4 minutes. Use the answer key to compare explanations, then ask learners to apply the tie rule to the final set of scores.
The student handout comes first. The answer key follows in a separate section so you can give out the questions before discussing the answers. You can also use this page on your own; cover the key until you have tried the questions.
Student handout
1. Find out what the number means
Example report: invented scores
These values were invented, not calculated from questionnaire answers. The scale runs from 1 to 5, with higher scores indicating more of each trait. Conscientiousness, for example, concerns tendencies such as being organized and dependable. No measurement-error estimates are available for these invented scores.
| Trait | Score |
|---|---|
| Openness | 3.2 |
| Conscientiousness | 3.6 |
| Extraversion | 2.7 |
| Agreeableness | 3.8 |
| Neuroticism | 2.5 |
Question 1. Someone replaces “Conscientiousness: 3.6 on a 1–5 scale” with “72% conscientious,” because 3.6 ÷ 5 × 100 = 72. Does dividing by the top of a scale tell you how much of a personality trait someone has? Explain what the original value does tell you.
In a real report, the same number could be a total, an average, or a comparison with other people's scores. The words beside the number should tell you which it is. You also need the questionnaire's name: “Big Five” names a broad model of personality, not one particular set of questions and scoring instructions.
Question 2. A different report simply says “65%.” What would you need to know before interpreting it? If a website adds that it uses a Big Five questionnaire, is that enough to know what the percentage means or whether it was calculated correctly?
2. Change the comparison, not the score
A reference group provides scores to compare with your result. A percentile rank describes where your score stands among those scores. To see how the group affects the rank, keep the score at 3.6 and compare it with each invented group below.
| Group | Scores, lowest to highest |
|---|---|
| A | 1.1, 1.8, 2.2, 2.8, 3.0, 3.2, 3.6, 4.0, 4.2, 4.8 |
| B | 2.9, 3.6, 3.8, 4.0, 4.1, 4.3, 4.4, 4.6, 4.8, 4.9 |
Each group contains a score equal to 3.6, so we need a rule for ties. For this exercise, use the midrank rule: 100 × (number below + 0.5 × number equal) ÷ group size. Count each value below 3.6 once and each value exactly equal to 3.6 as half. Divide that count by ten, the group size, then multiply by 100. Other tests may handle ties differently.
Question 3. For each group, write down the number below 3.6, the number equal to 3.6, and the resulting percentile rank. Did the score of 3.6 change? What caused the difference between the two percentile ranks?
Question 4. Could these invented ten-value groups tell you how a real person compares with all adults? For a real comparison, what would you want to know about the people whose scores make up the reference group?
3. Compare the distance and the label
For this next example, use a scale from 0 to 100. Values below 50 receive the label lower side; values of 50 or more receive the label upper side. It is an invented reporting rule, not official MBTI scoring.
| Score | Label |
|---|---|
| 49 | Lower side |
| 51 | Upper side |
| 90 | Upper side |
Question 5. Which pair is closer on the scale: 49 and 51, or 51 and 90? What information would you lose if you saw only the labels? The example gives one score for each case. Is that enough to tell you how likely a person's label is to change if they take a test again?
4. Rewrite the claim
Question 6. An invented report summary says: “The score is at the 65th percentile in Group A, so we know which partner this person should choose.” Rewrite the sentence to keep what you learned from the calculation. Then explain why the calculation cannot tell someone which partner to choose.
Question 7. Suppose a real report has good evidence that a person generally prefers preparation. Which is the more reasonable next step: recommend the same detailed schedule every day, or ask when preparation helps and compare a few ways to prepare? What do you still need to learn before recommending a particular routine?
5. Try a new case
A report states: “Score: 4.0. Comparison scores: 2.0, 3.0, 4.0, 4.0, 5.0.” Use the same midrank rule as above.
Question 8. What percentile rank does 4.0 receive? If the two values of 4.0 were mistakenly counted as lower values rather than equal values, what went wrong? Explain without describing the result as a percentage of personality.
Answer key and discussion notes
Ask learners to explain how they reached each answer. Someone can calculate 65 correctly and still mistake it for a percentage of personality. The explanation tells you whether they understand what they have calculated.
1. Why 3.6 does not mean “72% conscientious”
Dividing 3.6 by 5 and multiplying by 100 gives 72. But 5 is the top of the scale, not “all the conscientiousness a person can possess,” so the calculation does not make the person 72% conscientious. We know that 3.6 is above the scale midpoint of 3. To know whether it is above a group's average, or to calculate a percentile rank, we would need that group's scores.
2. Find out how the score was calculated
First, look for the exact questionnaire and version. Then find out what the score represents and whether larger numbers mean more or less of the named trait. If the percentage compares the score with a group, you need to know which group and how the ranking was calculated. Without those details, “65%” leaves several possible interpretations open.
Calling a questionnaire “Big Five” does not answer these questions. Even naming the exact questionnaire is only a starting point: a website could use the right questions but apply the wrong scoring instructions or explain the scores incorrectly. You need both evidence about the questionnaire and a clear explanation of how the website uses it.
3. The score stays fixed
- Group A: six values below, one equal. 100 × (6 + 0.5 × 1) ÷ 10 = 65.
- Group B: one value below, one equal. 100 × (1 + 0.5 × 1) ÷ 10 = 15.
The score remains 3.6. Six values in Group A are below it, compared with only one in Group B. It therefore stands lower in Group B. The percentile changed because the comparison changed, not because anything happened to the score or to a person.
4. Find out who the comparison group represents
No. These values were invented to teach the calculation. For a real group, ask who the people were, how they were selected, and when they took the test. Those details tell you what the comparison can show. A class can tell you how a score compares with that class; it does not automatically tell you how it compares with all adults. Nor does a large group of volunteers automatically represent everyone.
5. Labels hide how far apart scores are
The first pair is two points apart: 51 − 49 = 2. The second is 39 points apart: 90 − 51 = 39. The differently labeled pair is closer on this scale. Seeing only “lower side” and “upper side” would hide that fact, while seeing “upper side” twice could make the more distant pair look alike.
The example shows how a category can hide numerical distance. It does not tell us how often a real person's label would change on another occasion, because we have no repeated measurements from a real test. For the official MBTI, the publisher reports that about half of people get the same four-letter type on retest, and that a preference that changes is usually the one that was least clear, where the answers leaned least strongly to either side. The invented example shows why: a score near the dividing line needs only a small shift to receive a different label.
6. A rank cannot choose a partner
A suitable rewrite is: “Using our midrank rule, the score of 3.6 is at the 65th percentile of invented Group A.” The calculation locates a number among other numbers. It gives us no information about compatibility, relationship outcomes, or a particular partner. Moving from that rank to a partner recommendation adds a claim the example gives no evidence for. The example also never assessed a real person.
If a learner writes “65% of the group scored lower,” return to the tie rule. Six out of ten values are strictly lower, which is 60%. Giving half weight to the one equal value adds five percentage points and produces the percentile rank of 65. The rank includes that half-weighted tie; the count of lower scores does not.
7. A description can suggest a question
Ask when preparation helps and compare possible ways to prepare. A preference for preparation gives you a reason to consider these options; it does not tell you which one will help with the person's goal and circumstances. Writing the first task on a card and scheduling every hour both involve preparation, but they ask different things of the person. A detailed schedule might fit a predictable day better than one full of interruptions. You still need to find out what the person wants to accomplish and which approach helps them do it.
8. Count ties as ties
Two values are below 4.0, and two equal it. The group contains five values, so the answer is 100 × (2 + 0.5 × 2) ÷ 5 = 60. Each equal value counts as half, so the two ties together add one to the count. Counting both ties as lower would count each one fully and give 80 instead. That would answer a different question: what percentage of the reference scores are at or below 4.0? Using our midrank rule, 4.0 is at the 60th percentile. It is not “60% of a personality trait.”
For a final check, return to “Conscientiousness: 65%” on the board. Ask learners to complete this sentence: “Before I interpret this number, I need to know…” They should be able to explain why the questionnaire, the score definition, and any reference group matter, rather than simply recall the answers 65 and 15.
Sources and classroom copies
Packet edition: October 2, 2026. Use the learner handout on screen or print it with the separate answer key. Keep the title, author attribution, source notes, and fictional-data labels with classroom copies. The links below provide further explanation and lead to the original sources; those sources remain subject to their own terms of use.
- How to Read Personality Test Results explains each step in reading the example report.
- Read more about comparison groups in Test norms and about labels in the MBTI explanation.
- The 2014 Standards for Educational and Psychological Testing support the main points of this lesson. Standard 5.1 (printed page 102) asks for clear explanations of what scores mean. Standards 5.8 and 5.9 (printed page 104) ask for a clear description of the reference group. Chapter 1 (from printed page 11) ties validity to each proposed use of a score, which is why a percentile cannot choose a partner.
- SciPy's percentile documentation distinguishes strict, inclusive, and mean conventions. This packet uses the mean of the strict and inclusive proportions, expressed as half weight for ties.
The report, groups, numerical examples, and questions are original teaching material, not a personality questionnaire. To continue from reading a result to choosing an activity, use the Behavior Menu. Learners can consider different ways to reach a goal using what they know about their interests and circumstances; they do not need a personality score.