In: Statistics and Probability
The data from data47.dat contains information on 78
seventh-grade students. We want to know how well each of IQ score
and self-concept score predicts GPA using least-squares regression.
We also want to know which of these explanatory variables predicts
GPA better. Give numerical measures that answer these questions.
(Round your answers to three decimal places.)
(Regressor: IQ) R 2
(Regressor: Self-Concept) R 2
obs gpa iq gender concept 1 7.94 111 2 68 2 8.292 115 2 65 3 4.643 106 2 61 4 7.47 112 2 70 5 8.882 117 1 65 6 7.585 105 2 70 7 7.65 97 2 66 8 2.412 90 2 39 9 6 105 1 49 10 8.833 138 2 57 11 7.47 101 1 71 12 5.528 107 1 61 13 7.167 114 2 46 14 7.571 112 1 68 15 4.7 98 1 46 16 8.167 103 1 81 17 7.822 121 1 65 18 7.598 106 1 67 19 4 114 2 57 20 6.231 108 1 71 21 7.643 117 2 76 22 1.76 81 2 55 24 6.419 103 1 39 26 9.648 112 2 50 27 10.7 130 1 67 28 10.58 116 2 59 29 9.429 123 2 44 30 8 94 2 78 31 9.585 116 2 54 32 9.571 126 1 64 33 8.998 97 1 63 34 8.333 106 1 60 35 8.175 110 2 74 36 8 123 2 62 37 9.333 91 1 91 38 9.5 111 2 54 39 9.167 103 2 52 40 10.14 110 1 67 41 9.999 124 1 98 43 10.76 116 2 79 44 9.763 125 2 71 45 9.41 109 2 71 46 9.167 103 2 70 47 9.348 116 2 80 48 8.167 121 2 71 50 3.647 95 2 50 51 3.408 98 1 46 52 3.936 97 2 48 53 7.167 87 2 71 54 7.647 119 2 71 55 .53 60 2 44 56 6.173 107 2 71 57 7.295 107 2 55 58 7.295 114 1 71 59 8.938 121 1 59 60 7.882 112 1 60 61 8.353 114 2 56 62 5.062 116 2 67 63 8.175 116 2 65 64 8.235 114 2 74 65 7.588 110 2 50 68 7.647 100 2 85 69 5.237 107 1 52 71 7.825 114 2 47 72 7.333 86 1 58 74 9.167 109 2 36 76 7.996 98 2 72 77 8.714 116 1 60 78 7.833 103 1 62 79 4.885 101 2 39 80 7.998 105 1 46 83 3.82 79 2 36 84 5.936 108 1 33 85 9 99 1 78 86 9.5 106 1 47 87 6.057 107 2 60 88 6.057 94 1 64 89 6.938 115 2 44
R2 is highlihted in yellow in the pics: The linear regression linear for best ft is run and results are below:
a) Regression with Self concept
SUMMARY OUTPUT | |||||
Regression Statistics | |||||
Multiple R | 0.438798403 | ||||
R Square | 0.192544038 | ||||
Adjusted R Square | 0.181919617 | ||||
Standard Error | 1.899003434 | ||||
Observations | 78 | ||||
ANOVA | |||||
df | SS | MS | F | Significance F | |
Regression | 1 | 65.35462429 | 65.35462429 | 18.12278016 | 5.85433E-05 |
Residual | 76 | 274.0722671 | 3.606214041 | ||
Total | 77 | 339.4268914 | |||
Coefficients | Standard Error | t Stat | P-value | Lower 95% | |
Intercept | 3.179726112 | 1.0250894 | 3.10190127 | 0.002698498 | 1.138083631 |
Self concept | 0.069786405 | 0.016392999 | 4.257085877 | 5.85433E-05 | 0.037136917 |
b)Regression with IQ:
SUMMARY OUTPUT | |||||
Regression Statistics | |||||
Multiple R | 0.618195085 | ||||
R Square | 0.382165163 | ||||
Adjusted R Square | 0.374035757 | ||||
Standard Error | 1.6611258 | ||||
Observations | 78 | ||||
ANOVA | |||||
df | SS | MS | F | Significance F | |
Regression | 1 | 129.7171332 | 129.7171332 | 47.0102212 | 1.63509E-09 |
Residual | 76 | 209.7097582 | 2.759338924 | ||
Total | 77 | 339.4268914 | |||
Coefficients | Standard Error | t Stat | P-value | Lower 95% | |
Intercept | -4.040199423 | 1.685855661 | -2.396527482 | 0.019009996 | -7.397871967 |
IQ | 0.106700673 | 0.015562201 | 6.856400018 | 1.63509E-09 | 0.075705864 |
c) The regression with IQ yeild a R2 of 0.3821 which is more than 0.1925 for Self concept.
The R2 determines which is a better predictor. The higher it is the more predictive power the regressor has.
Hence, IQ is a better predictor.