Correlation and linear regression
Practice section 1 of 6
Questions 1–6 in this collection
Practice questions
Worked example · 1 of 6
Identify explanatory and response variables
Identify the independent and dependent variables in each source situation:
- Investigate whether increased calcium intake would help prevent osteoporosis.
- Investigate how the concentration of an injected drug in the bloodstream varies with time.
- Investigate the time taken to complete a chemical reaction when different amounts of catalyst are used.
- Record air temperature and altitude at noon on a particular day at eight mountainous locations.
- Tabulate the distances of seven cities from the capital and the corresponding railway fares.
CJC · Lecture · Variable-identification examples
Solution
Worked example · 2 of 6
Sketch the eleven insulin and time observations
The source records the time for blood sugar to return to normal after insulin was injected into a diabetic patient.
| Insulin (milli-units) | Time (minutes) |
|---|---|
| 40 | 4.10 |
| 60 | 4.07 |
| 95 | 3.98 |
| 100 | 3.90 |
| 125 | 3.81 |
| 160 | 3.75 |
| 180 | 3.72 |
| 210 | 3.77 |
| 240 | 3.63 |
| 260 | 3.57 |
| 500 | 3.30 |
Sketch a scatter diagram with insulin injected as the independent variable. Give a brief description of any observed pattern.
CJC · Lecture · Example 1
Solution
Worked example · 3 of 6
Calculate correlation and investigate a radiation outlier
The data show the intensity of radiation from a substance after time .
| 0.2 | 3 |
| 0.4 | 3.5 |
| 0.6 | 1.3 |
| 0.8 | 0.6 |
| 1.0 | 0.4 |
| 1.2 | 0.2 |
(i) Use a suitable diagram to explain why the product moment correlation coefficient between and is negative and not very close to .
(ii) By using the formula for , calculate the product moment correlation coefficient between and .
(iii) State the coordinates of a point in the diagram in (i) which does not seem to follow the linear trend. If it is removed, without any calculations, explain why is closer to .
CJC · Lecture · Example 2
Solution
Worked example · 4 of 6
Fit both regression directions and recover the means
Eight pairs of observations are given below.
| 16 | 7 |
| 13 | 17 |
| 16 | 12 |
| 15 | 11 |
| 16 | 18 |
| 12 | 10 |
| 17 | 17 |
| 15 | 12 |
(i) Find the regression line of on .
(ii) Find the regression line of on .
(iii) On a scatter diagram, sketch the two regression lines.
Use your answers in (i) and (ii) to find and .
CJC · Lecture · Example 3
Solution
Worked example · 5 of 6
Recover two unknown observations from a fitted line
When and are known, the following data give the regression line of on as .
| 3 | |
| 2 | 2 |
| 3 | 6 |
| 6 | |
| 1 | 0 |
(i) Find .
(ii) Find .
(iii) Find and .
CJC · Lecture · Example 4
Solution
Worked example · 6 of 6
Correct recorded marks and compare the scatter
The average weekly study hours and corresponding test marks of ten students are recorded below.
| Student | ||
|---|---|---|
| A | 24 | 6 |
| B | 17 | 30 |
| C | 20 | 40 |
| D | 41 | 76 |
| E | 52 | 96 |
| F | 23 | 46 |
| G | 46 | 90 |
| H | 17 | 35 |
| I | 15 | 26 |
| J | 29 | 57 |
(i) Calculate the product moment correlation coefficient between and .
(ii) Determine the least-squares regression line of on .
(iii) Three marks were wrongly recorded: A should receive 35, I should receive 6, and E should receive 99. Using appropriate diagrams, comment on how these changes affect (a) the gradient of the regression of on and (b) the correlation coefficient.
CJC · Lecture · Example 5