In: Statistics and Probability
Twenty-five volunteer athletes participated in a study of cross-disciplinary athletic abilities. The group comprised athletes from football, baseball, water polo, volleyball, and soccer. None had ever played organized basketball, but did acknowledge interest and some social participation in the game. Height (inches), weight (pounds), and speed in the 100-yard dash (seconds) were recorded for each subject. The basketball test consisted of the number of field goals that could be made in a 60-min. period. The data are given in Athlete.jmp on Blackboard. We are interested in predicting GOALMADE using some combination of WEIGHT, HEIGHT, DASH100.
3) Using JMP output from above, do any of the input variables (HEIGHT, WEIGHT, DASH100) exhibit collinearity?
4) Now, we want to determine if the variables WEIGHT and/or HEIGHT should be added to the model that already contains DASH100 to explain GOALMADE. Write the full and reduced models; then perform the appropriate test. (Please show your expanded ANOVA table, but you do not need to state your hypotheses explicitly.)
5) Finally (without running any more JMP output), determine if the variable DASH100 is useful in explaining GOALMADE when HEIGHT and WEIGHT are already in the model. State the test-statistic, p-value, and conclusion only
JMP DATA:
W H D G
130 71 11.5 15
149 74 12.23 19
170 70 12.26 11
177 71 12.65 15
188 69 10.26 12
210 73 12.76 17
223 72 11.89 15
170 75 12.32 19
145 72 10.77 16
132 74 11.31 18
211 71 12.91 13
212 72 12.55 15
193 73 11.72 17
146 72 12.94 16
158 71 12.21 15
154 75 11.81 20
193 71 11.9 15
228 75 11.22 19
217 78 10.89 22
172 79 12.84 23
188 72 11.01 16
144 75 12.18 20
164 76 12.37 21
188 74 11.98 19
231 70 12.23 13
3. Coefficients
Term | Coef | SE Coef | T-Value | P-Value | VIF |
Constant | -67.36 | 5.22 | -12.91 | 0.000 | |
Weight | -0.01057 | 0.00505 | -2.09 | 0.049 | 1.01 |
Height | 1.2027 | 0.0627 | 19.19 | 0.000 | 1.01 |
Dash 100 | -0.142 | 0.216 | -0.66 | 0.519 | 1.01 |
Since the VIF value is 1.01, which is very close to 1, there is no collinearity between the variables
If the value of VIF is between 5 to 10, we can say the the variables are highly collinear
4.& 5. Full model -
Coefficients
Term | Coef | SE Coef | T-Value | P-Value | VIF |
Constant | -67.36 | 5.22 | -12.91 | 0.000 | |
Weight | -0.01057 | 0.00505 | -2.09 | 0.049 | 1.01 |
Height | 1.2027 | 0.0627 | 19.19 | 0.000 | 1.01 |
Dash 100 | -0.142 | 0.216 | -0.66 | 0.519 | 1.01 |
Reduced Model -
Analysis of Variance
Source | DF | Adj SS | Adj MS | F-Value | P-Value |
Dash 100 | 23 | 213.36 | 9.277 | 0.52 | 0.823 |
Error | 1 | 18.00 | 18.000 | ||
Total | 24 | 231.36 |
Since the p-value is more than 0.05 in the case of Dash100 only and we are using 95% confidence interval, Dash 100 only is enough to explain Goal Made