Question

A group of Maternal and Child Health public health practitioners are interested in the relationship between bacterial vaginosis (BV) and a number of health outcomes in a developing country. Suppose the research team gathers information on a group of participants, and constructs a multiple linear regression model looking at the relationship between BV and tuberculosis. The following is a computerized output displaying the results of their analysis including the following potential confounders: Race (White/African American), Poverty (dichotomized above and below the poverty line), Age at pregnancy, and Alcohol (Since pregnant, had any alcohol? Y/N). (1 pt)

Parameter	Estimate	Standard Error	t Value	Pr > |t|
Intercept	0.2652506938	0.10534055	2.52	0.0120
TBresult	-.0529842698	0.02596828	-2.04	0.0417
poverty	-.0359729022	0.04124888	-0.87	0.3835
race	0.0742883690	0.03360015	2.21	0.0274
ageatpregnancy	-.0042718151	0.00254591	-1.68	0.0939
alcohol	0.0083611736	0.02669668	0.31	0.7542

1. What are the independent and dependent variables?
   .

B) Which potential confounders were significantly affecting the relationship between the exposure and the outcome variables?

C) Write out the model in symbols. Round to 3 decimal places.

Answer 1

A)What are the independent and dependent variables?
Dependent Variable:
Bacterial Vagninosis

Independent Varaibles:
Tuberculosis
Race
Poverty
Age at pregnancy
Alchol

B) Which potential confounders were significantly affecting the relationship between the exposure and the outcome variables?

To identify the potential confounders we look at the pvalue. If it less than 0.5, they are significant.

Hence TBresult and race are significant variables.

C) Write out the model in symbols. Round to 3 decimal places.

y = 0.265-0.053   TBresult-0.036   poverty+0.074   race-0.004 ageatpregnancy+0.008   alcohol