Question 1 of 15

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

Below is the simple crosstab for Beta Blocker use vs. Low HDL. The natural log of the odds ratio [ln(OR)]=

Question 2 of 15

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

Using the table from question 1, the standard error of ln(OR)=  

Question 3 of 15 1.0 Points

Using the answer for the two previous questions, calculate a 95% CI for OR (Note, you will have to convert your answer from log form before you select an answer).

 A. 1.644 - 14.368

 B. 2.795 - 8.449

 C. .497 - 2.665

 D. 1.028 - 2.134

Reset Selection

Question 4 of 15

In this assignment, you will use the data set “birthwt3.sav”, in homework files under resources. This time we will use the clinical criterion for low birth weight (<2500 grams) to determine whether the baby is of low birth weight. Then you will use logistic regression to examine the effects of risk factors on the chance of a mother having a low birth weight baby. The dataset contains the following variables:

ID: Mother’s identification number

MOTH_AGE: Mother’s age (years)

MOTH_WT: Mother’s weight (pounds)

RACE: Mother’s race (1=White, 2=Black, 3=Other)

dummy variables (Black, White and Other) are already coded in the data set.

SMOKE: Mother’s smoking status (1=Yes, 0=No)

PERM: History of premature labor (number of times)

HYPE: History of hypertension (1=Yes, 0=N0)

URIN_IRR: History of urinary irritation (1=Yes, 0=N0)

PHYS_VIS: Number of physician visits

BIRTH_WT Birth weight of new born (grams)

LBW: Dummy variable, coded 1 if the baby has low birth weight

Run a logistic regression to predict the outcome “LBW” using Smoke, Black and Other as the independent variables. Based on the logistic regression results, match the odds ratio for each of the variables.

A. 3.030

B. 2.957

C. 3.053

1. Smoke

2. Black

3. Other

Question 5 of 15

We cannot make which of the following conclusions based on the results above?

 A. The effect of smoking on the risk of having low birth weight baby is higher among black women than among white women.

 B. Adjusting for smoking status, the odds of having low birth weight baby for Black women and women of other race is about 3 times the odds for White women.

 C. Adjusting for race, smoking increases the risk of having birth weight baby significantly.

 D. The effect of smoking on the risk of having low birth weight baby is the same for women of different races.

Reset Selection

Question 6 of 15

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

The odds of having low birth weight baby for a Black women who smoked during pregnancy is  

Question 7 of 15 1.0 Points

Run a second logistic regression to further include “age” and “weight” as independent variables. Based on the logistic regression results, which one of the following statements is incorrect?

 A. After further adjusting for mother’s age and weight, smoking is still an important risk factor for having low birth weight baby.

 B. After further adjusting for mother’s age and weight, the racial gap between black and white women in terms of their risk having low birth weight baby widened a little bit.

 C. After controlling for age, race and smoking status, mothers who weigh more may have a lower risk of having low birth weight baby. But this finding is barely significant at 5% level.

 D. After controlling for mother’s weight, race and smoking status, older mothers have a lower risk of having low birth weight baby. This finding is significant at 5% level.

Reset Selection

Question 8 of 15

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

Based on the model above, what is the odds of having low birth weight baby for a 25 year-old Black woman who weighs 120 pounds and who did not smoke during pregnancy?  

Question 9 of 15 1.0 Points

Run a third logistic regression to further include variables about medical history including: history of premature labor, hypertension, urinary irritation and number of physician visits. Check “Hosmer and Lemeshow goodness-of fit” statistics in “Options." Check "Probabilities" and "group membership" in "Save."

Which one of the following variables regarding medical history has significant impact (at level 5%) on the risk of women having low birth weight babies?

 A. History of premature labor

 B. History of hypertension

 C. History of urinary irritation

 D. Number of physician visits

Reset Selection

Question 10 of 15 1.0 Points

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

Based on the third model, overall what percentage of the sample is correctly predicted? (Please do not include "%" in your answer. If it is 50.0%, write 50.0)  

Question 11 of 15 1.0 Points

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

For the participant on line 73 (ID number 164.00), what is the predicted probability of LBW using model 3. (Hint, to find the answer you will have to go to data view and pull it out of the dataset under the PRE_ column). Report the number exactly as it appears in the data  

Question 12 of 15 1.0 Points

Is the participant on line 94 of the dataset (ID number 188.00) correctly classified by the model. (Once again you will have to look at the data view and the PGR_ column)

 A. Yes

 B. No

Reset Selection

Question 13 of 15

Accepted characters: numbers, decimal point markers (period or comma), sign indicators (-), spaces (e.g., as thousands separator, 5 000), "E" or "e" (used in scientific notation). NOTE: For scientific notation, a period MUST be used as the decimal point marker. 

Complex numbers should be in the form (a + bi) where "a" and "b" need to have explicitly stated values. 

For example: {1+1i} is valid whereas {1+i} is not. {0+9i} is valid whereas {9i} is not. 

What is the Cox Snell Pseudo R square for model 3?  

Question 14 of 15 1.0 Points

What is the null hypothesis for the Hosmer–Lemeshow test?

 A. The logistic regression model fits the data well.

 B. The logistic regression model does not fit the data well.

 C. All the independent variables are significant.

 D. None of the independent variables are significant.

Reset Selection

Question 15 of 15

Based on the Hosmer–Lemeshow test, the thrid logisitic regression fits data well.

 True

 False

Reset Selection

Read less