Lesson 8: Is intelligence inherited or learned?
Analyse IQ data from an adoption study
SkillsInherited or learnedResearch methods in intelligence
Worked example
Researchers followed 32 children with similar birth backgrounds who were adopted at about age 5. Sixteen were adopted into families with low socioeconomic status (SES) and sixteen into families with high SES. Their IQs were measured at age 14 (illustrative data).
| Adoptive family | n | Mean IQ | SD |
|---|---|---|---|
| Low SES | 16 | 92 | 12 |
| High SES | 16 | 104 | 12 |
(a) Calculate the standard error and the 95% confidence interval for each group's mean. (b) An unpaired t-test gives , . What can be concluded? (c) Discuss what this study shows about whether intelligence is inherited or learned.
Step 1: Standard error
for each group.
Step 2: 95% confidence intervals
Margin . Low SES: , so 86.12 to 97.88. High SES: , so 98.12 to 109.88.
Step 3: Read the t-test
, so reject the null hypothesis: the difference of 12 IQ points between the groups is statistically significant. The confidence intervals do not overlap either.
Step 4: Link to nature and nurture
The children had similar birth backgrounds, so their genetic backgrounds were similar on average; the groups differed mainly in the environment they were adopted into. A 12-point difference therefore points to an environmental effect on IQ.
Step 5: Limit the conclusion
Placement was not random (agencies may place children selectively), the samples are small, and the study cannot say which part of the environment (schooling, resources, expectations) made the difference. It also does not show genes are unimportant: within each group, children's IQs still vary.
Answer
(a) SE = 3 for each; 95% CIs 86.12–97.88 (low SES) and 98.12–109.88 (high SES). (b) The 12-point difference is significant (). (c) With similar birth backgrounds, the difference suggests the adoptive environment raised IQ, so intelligence is partly learned; but placement was not random, the samples are small, and the result does not rule out genetic influence.
Try a similar one
A fresh question of the same kind, checked as you answer, with hints and a full solution.
Method
; 95% CI .
Compare p with 0.05.
Ask what differs between the groups: genes, environment or both.
State the limits of the design.
Watch out for
Using the SD instead of the SE to build the confidence interval.
Concluding intelligence is entirely learned from one adoption study.