Marx: Lab9
Today just get the r, intercept and b (slope) out from the software programs. You should be able calculate and interpret the "r" today.
Table 9-1 shows the results based on a classic piece of small group research in which the size of groups (X) is related to the tendency of group members to perceive closeness (Y). Using the data in Table 5-1 do the problems by hand and by computer. DUE: OCT 24 (beginning of class).
Only new command needed is:
Traditional SAS:
Proc Reg Corr Simple;
Model dependent var = predictor var/P;
Run;
SPSS:
Click>Analyze, then Correlate, then Bivariate. Move the variable over (make sure
that the Pearson box under correlations is checked)--okay to run
Click>Analyze, the Regression, then Linear-be sure to drag the dependent (Y) and
Independent into the appropriate box--okay to run;
Enterprise (ignore)
Select Linear Regression (from Taskbar)
Drag dependent variable (y) and predictor (x) to appropriate slots.
Select Multivariate Correlation.
Drag Analysis variable (which is X or size here)
Drag correlated variable (which is Y or closeness)
All work must be done by hand and computer (traditional SAS and SPSS)
Table 9-1
Perception of Group Closeness (Y) by Size of Group (X), for Ten Groups
X Y
Percent Perceiving
Group Size
Group Closeness
2
100
2
65
3
30
3
40
4
25
4
8
5
2
5
15
6
0
7
10
1) Compute the correlation (r) between x and y.
* B) Put the results in a sentence that describes the relationship between
group size and
member’s
perceptions.
2) Compute the slope (b).
member’s perceptions (interprete slope)
3) Compute the intercept (a)
B) Write out the line equation as a whole.
4) Compute the coefficient of determination (r square).
B) Put the result in a sentence that describes the relationship between
group size and
member’s perceptions (interprete coefficient of determination)