3. Partial correlation is an expression of analyses of
co-variance (ANCOVA) applied to questions of
prediction and relationship
4. Partial correlation is an expression of analyses of
co-variance (ANCOVA) applied to questions of
prediction and relationship
Prediction
&
Relationship
5. . . . rather than questions of differences
addressed by ANCOVA.
6. . . . rather than questions of differences
addressed by ANCOVA.
7. Partial correlation estimates the relationship
between two variables while removing the
influence of a third variable from the
relationship.
8. Partial correlation estimates the relationship
between two variables while removing the
influence of a third variable from the
relationship.
Relationship between
a guy and a girl
9. Partial correlation estimates the relationship
between two variables while removing the
influence of a third variable from the
relationship.
Taking out the effect
of video games on
that relationship
10. For example, a researcher might want to know
the relationship between height and weight.
11. For example, a researcher might want to know
the relationship between height and weight.
&
12. However, she is aware that people’s bone and
muscle structures vary according to gender.
13. However, she is aware that people’s bone and
muscle structures vary according to gender.
14. She can calculate a partial correlation between
height and weight while removing (holding
constant, eliminating) the effect of gender on
the correlation.
15. She can calculate a partial correlation between
height and weight while removing (holding
constant, eliminating) the effect of gender on
the correlation.
&
16. She can calculate a partial correlation between
height and weight while removing (holding
constant, eliminating) the effect of gender on
the correlation.
&
17. She can calculate a partial correlation between
height and weight while removing (holding
constant, eliminating) the effect of gender on
the correlation.
&
The Effect of Gender
19. Here’s the data set:
Individual Height
(inches)
Weight
(pounds)
Gender (1 – male, 2 – female)
A 73 240 1
B 70 210 1
C 69 180 1
D 68 160 1
E 70 150 2
F 68 140 2
G 67 135 2
H 62 120 2
20. Here’s the data set:
Individual Height
(inches)
Weight
(pounds)
Gender (1 – male, 2 – female)
A 73 240 1
B 70 210 1
C 69 180 1
D 68 160 1
E 70 150 2
F 68 140 2
G 67 135 2
H 62 120 2
21. First let’s see what the correlation between
height and weight is and then we will see what
the correlation would be if we controlled for
gender
23. Correlation between Height and Weight
Individual Height
(inches)
Weight
(pounds)
Gender (1 – male, 2 – female)
A 73 240 1
B 70 210 1
C 69 180 1
D 68 160 1
E 70 150 2
F 68 140 2
G 67 135 2
H 62 120 2