AP Psychology · Lesson 3

Intelligence, creativity and testing

Know Know it

Psychologists disagree on what intelligence is. Charles Spearman proposed a general intelligence (g) that underlies many abilities. Howard Gardner proposed multiple intelligences, such as linguistic, spatial and interpersonal. Robert Sternberg proposed three kinds: analytical, creative and practical. Fluid intelligence is reasoning quickly about new problems; crystallized intelligence is stored knowledge, such as vocabulary.

Creativity depends on divergent thinking, which produces many possible answers, while convergent thinking narrows to the one best answer.

Aptitude tests predict what a person can learn, and achievement tests measure what a person has already learned. A test is standardized when it is given and scored the same way and compared with a large norm group. On many intelligence tests, scores follow a normal curve with a mean of 100 and a standard deviation of 15, so about 68 percent of people score between 85 and 115, and about 95 percent between 70 and 130.

A good test is reliable: it gives consistent results, as in test-retest reliability. It is also valid: it measures what it claims to, as in construct validity, or predicts what it should, as in predictive validity.

Scores can shift for reasons other than ability. In the Flynn effect, average scores on intelligence tests rose over generations. Stereotype threat can lower performance when people fear confirming a negative stereotype about their group. A growth mindset, the belief that ability can improve with effort, is linked to persistence. Comparing different age groups at one time is a cross-sectional design; following the same people over years is a longitudinal design.

Trap Reliable is not the same as valid

A scale that always reads five pounds too heavy is perfectly consistent but wrong. In the same way, a test can give the same scores every time and still fail to measure what it claims to. Reliability is needed for validity, but it does not guarantee it.

Worked example Easy

A new reasoning test is given twice, two weeks apart, to the same 200 students. Their scores on the two dates are very similar. What does this show about the test?

  1. Construct validity
  2. ReliabilityAnswer
  3. Standardization against a norm group
  4. The Flynn effect

How to solve it

  1. Same people, same test, similar scores: the results are consistent.
  2. Consistency of results is reliability.

Why each choice is right or wrong

  • A. Incorrect. Construct validity is about whether the test measures the trait it claims to, which matching scores over time cannot show.
  • B. Correct. Similar scores when the same people retake the test show consistency, which is test-retest reliability.
  • C. Incorrect. Standardization against a norm group means giving the test the same way and comparing with others, not retesting the same people.
  • D. Incorrect. The Flynn effect is a rise in average scores across generations, not two weeks of stable scores.

Explain Explain it

Task verbs: explain, compare

When you compare two test ideas, define each in one clause and then show where the scenario fits one and not the other. For reliability and validity, ask two questions: are the scores consistent, and do they match the thing the test is meant to measure?

Worked example Medium

On an intelligence test, scores follow a normal curve with a mean of 100 and a standard deviation of 15. About what share of people score between 85 and 115?

  1. About 34 percent
  2. About 50 percent
  3. About 68 percentAnswer
  4. About 95 percent

How to solve it

  1. Find the distance from the mean: 115 minus 100 is 15, one standard deviation, and 100 minus 85 is also 15.
  2. Within one standard deviation on both sides of the mean lies about 68 percent of a normal curve.

Why each choice is right or wrong

  • A. Incorrect. About 34 percent is the share between the mean and one standard deviation on just one side, such as 100 to 115.
  • B. Incorrect. About 50 percent is the share above or below the mean, not the band from 85 to 115.
  • C. Correct. The range 85 to 115 is one standard deviation on each side of the mean, which holds about 68 percent of a normal curve.
  • D. Incorrect. About 95 percent falls within two standard deviations, from 70 to 130, a wider band than 85 to 115.

Explain Your turn

A school's new reading test gives nearly the same score each time a student retakes it, but the scores do not match teachers' ratings or students' later reading grades. Explain why this test can be reliable but still not valid.

Write it, then compare. Nothing you type here is saved or graded.

Compare with a model answer

A full answer does this

  • Says the test is reliable because its scores are consistent on retakes
  • Says the test lacks validity because its scores do not match what it should measure or predict

Model answer

The test is reliable because a student gets nearly the same score each time, so its results are consistent. It is not valid because a reading test should match other signs of reading skill, yet its scores do not line up with teachers' ratings or later reading grades, so it is not measuring reading ability well.

Apply Apply it

Practice these topics. AP Practice questions on exactly what this lesson covers, with an explanation for every answer.