Data displays, normal distributions and association
Introduced| English | Meaning |
|---|---|
| skewed/skjuːd/ | Having a longer tail on one side of the distribution. |
| correlation/ˌkɒrɪˈleɪʃn/ | A statistical association between variables, which does not by itself establish causation. |
Test scores on a bell curve
- A test's scores are approximately normal with mean 70 and standard deviation 8. About what percentage of students scored between 62 and 78?
- ACT statistics questions ask you to read displays, interpret a normal model and judge association between two variables.
Reading data displays
- Histograms: shape (symmetric, skewed 偏斜的 left or right), centre and spread. A long right tail pulls the mean above the median.
- Box plots show the median, quartiles and range; the box covers the middle 50%.

About 68% of scores lie within one standard deviation, between 62 and 78
A histogram of household incomes has a long tail to the right. Which is most likely?
Difficulty 1/5 · 100 coins + a small bonus on first completion today
The high values in the tail pull the mean up.
The normal model
- About 68% of values lie within 1 SD of the mean, 95% within 2 SD, 99.7% within 3 SD.
- 62 to 78 is $70 \pm 8$ (1 SD): about 68%. Scores above 86 (2 SD above): about 2.5%.
Heights: mean 165 cm, SD 6 cm. 153–177 is $\pm 2$ SD, so about 95% of people fall in that range.
How much of the data?
Apply the 68–95–99.7 rule.
Weights are normal with mean 50 kg and SD 4 kg. About what percentage lie between 46 and 54 kg?
Difficulty 2/5 · 140 coins + a small bonus on first completion today
That is within 1 SD of the mean.
With mean 70 and SD 8 (normal), about what percentage of scores are above 86?
Difficulty 2/5 · 140 coins + a small bonus on first completion today
86 is 2 SD above; 5% lie outside ±2 SD, half of it above.
Association between two variables
- A scatterplot shows direction (positive or negative), form (linear or curved) and strength. A correlation 相关性 coefficient $r$ near 1 or −1 is strong; near 0 is weak.
- Association is not causation; a third variable may explain both.
Which correlation coefficient shows the strongest linear association?
Difficulty 1/5 · 100 coins + a small bonus on first completion today
Strength is the size of $r$; −0.92 is closest to ±1.
Traps
- Skew and the mean: in right-skewed data, mean > median.
- Strong correlation as proof of cause.
A correlation of 0.9 between ice-cream sales and drowning proves ice cream causes drowning.
Difficulty 1/5 · 100 coins + a small bonus on first completion today
A third variable (hot weather) explains both.
Recap
- Read shape, centre and spread; skew pulls the mean toward the tail.
- Normal model: 68–95–99.7 within 1, 2, 3 SD; correlation measures linear association, not cause.
- skewed: Having a longer tail on one side of the distribution.