In: Statistics and Probability
17.3
Consider a binary response variable y and an explanatory variable x. The following table contains the parameter estimates of the linear probability model (LPM) and the logit model, with the associated p-values shown in parentheses. |
Variable | LPM | Logit |
Constant | −0.69 | −6.30 |
(0.04) | (0.06) | |
x | 0.05 | 0.15 |
(0.07) | (0.06) | |
a. |
Test for the significance of the intercept and the slope coefficients at a 5% level in both models. |
Coefficients | LPM | Logit |
Intercept | (Click to select)Not significantSignificant | (Click to select)Not significantSignificant |
Slope | (Click to select)SignificantNot significant | (Click to select)Not significantSignificant |
b. |
What is the predicted probability implied by the linear probability model for x = 21 and x = 37? (Round intermediate calculations to 4 decimal places and final answers to 2 decimal places.) |
y-hat | |
x = 21 | |
x = 37 | |
c. |
What is the predicted probability implied by the logit model for x = 21 and x = 37? (Round intermediate calculations to 4 decimal places and final answers to 2 decimal places.) |
y-hat | |
x = 21 | |
x = 37 | |
a.)
siginificance level, = 0.05
Null Hypotheses, Ho: Not significant
Alternate Hypotheses, Ha: Significant
For LPM Model:
P(Intercept) = 0.04 which is less than 0.05
Hence Null Hypotheses is rejected which means Intercept is Significant
P(Slope) = 0.07 which is not less than 0.05
Hence Null Hypotheses is not rejected which means Slope is not Significant
For LOGIT Model:
P(Intercept) = 0.06 which is more than 0.05
Hence Null Hypotheses is not rejected which means Intercept is not Significant
P(Slope) = 0.06 which is not less than 0.05
Hence Null Hypotheses is not rejected which means Slope is not Significant
b.)
LPM is given as
p = -0.69 + 0.05*x
-> For x = 21
p = -0.69 + 1.05
p =0.36
-> For x = 37
p = -0.69 + 37 * 0.05
p = 1.16
c.)
In case of logit model, we get
ln (p / (1-p)) = -6.30 + 0.15 * x
So, For x = 21
ln (p / (1-p)) = -6.30 + 0.15 * 21
ln (p / (1-p)) = 0.36
=> p / (1-p) = exp(0.36)
=> p / (1-p) = 1.433
=> p = 1.433 - 1.433*p
=>p = 1.433 / 2.433
=> p = 0.5889 = 0.59
For x = 37
ln (p / (1-p)) = -6.30 + 0.15 * 37
ln (p / (1-p)) = 1.16
=> p / (1-p) = exp(1.16)
=> p / (1-p) = 3.189
=> p = 3.189 - 3.189 * p
=>p = 3.189 / 4.189
=> p = 0.7612 = 0.76