SCATTERPLOT
What it does:
It allows us to visually examine the relationship between two interval/ratio variables.
The dependent variable is plotted along the vertical and the independent variable along the horizontal axis.
EXAMPLE:
The following is a list of 7 students. We know how much time they spent studying and watching TV per day, how many sport competitions they won last year and their scores on two exams.
|
Name |
# of hours studied per day |
# hours of TV watched |
Points on final exam |
# of sport competition won |
Points on 2nd final exam |
|
|
1. |
John |
1 |
5 |
20 |
3 |
100 |
|
2. |
Jane |
2 |
4 |
60 |
1 |
50 |
|
3. |
Joe |
3 |
3 |
40 |
1 |
10 |
|
4. |
Mike |
4 |
2 |
80 |
3 |
50 |
|
5. |
Ann |
5 |
1 |
80 |
2 |
100 |
|
6. |
Mary |
2 |
4 |
60 |
5 |
40 |
|
7. |
Juan |
4 |
2 |
100 |
5 |
40 |
PLOT 1.
Correlation: +.788
The more you study the better your grades are,
Note: Jane and Mary share the same exact location on this plot.
PLOT 2.
Correlation: -.788
The less TV you watch the better your grades are.
Note: Jane and Mary share the same exact location on this plot.
PLOT 3.
Correlation: .000
Your success in sport is unrelated to the number of hours you study.
PLOT 4.
Correlation: .000
If you study a little or a lot you will get good grades. If you study a medium amount you will get bad grades. (This would be the case if those who are especially talented would need no study and would study little, and the rest would have to study a lot to get a good grade. )
This last plot shows that a zero correlation coefficient does not necessarily mean no relationship. It only means that there is no linear relationship.