2 Broken Hearts In a study of 900 randomly selected Americans, data are collected for whether...

Transcribed Image Text:

2 Broken Hearts In a study of 900 randomly selected Americans, data are collected for whether or not the subject has had a heart attack (attack = 1 if had a heart attack, 0 otherwise) along with variables age, systolic blood pressure, diastolic blood pressure, if they smoke or not. A subset of those data are presented in the table. Heart attack age systolic bp diastolic bp smoke 55 1 43 60 50 65 62 47 45 1 0 0 1 143 122 124 162 130 128 114 127 95 75 83 90 92 87 69 86 1 0 a) With these data, the probit and logit models cannot be used! This is because a single variable perfectly predicts heart attacks!! Notice that heart attack always equals 1 when systolic bp > 128 and never equals 1 when systolic bp <128. For this reason, let's consider the linear probability model. What is the fraction correctly predicted for the model attack = -3.025911+0008822age+.0254432systolic-0014869diastolic +3739527smoke b) The regression results of the LPM show that none of the coefficients are significant at the 10% level (in fact, the minimum p-value is 0.255). The value of the Fa statistic is 1.93. Comment on why you think there is a lack of statistical significance for the LPM. 2 Broken Hearts In a study of 900 randomly selected Americans, data are collected for whether or not the subject has had a heart attack (attack = 1 if had a heart attack, 0 otherwise) along with variables age, systolic blood pressure, diastolic blood pressure, if they smoke or not. A subset of those data are presented in the table. Heart attack age systolic bp diastolic bp smoke 55 1 43 60 50 65 62 47 45 1 0 0 1 143 122 124 162 130 128 114 127 95 75 83 90 92 87 69 86 1 0 a) With these data, the probit and logit models cannot be used! This is because a single variable perfectly predicts heart attacks!! Notice that heart attack always equals 1 when systolic bp > 128 and never equals 1 when systolic bp <128. For this reason, let's consider the linear probability model. What is the fraction correctly predicted for the model attack = -3.025911+0008822age+.0254432systolic-0014869diastolic +3739527smoke b) The regression results of the LPM show that none of the coefficients are significant at the 10% level (in fact, the minimum p-value is 0.255). The value of the Fa statistic is 1.93. Comment on why you think there is a lack of statistical significance for the LPM.

Answer rating: 100% (QA)

a To calculate the fraction correctly predicted for the model we need to compare the predicted values from the model with the actual values of the att... View the full answer