Answers to Chapter 4
Test Yourself
Question #1:
Using the dummy table given your results should look like this:
Prevalence Underweight |
|||
Educational Attainment |
|||
Bad (none/inc. primary) |
Good (comp. primary +) |
Total |
|
Sanitation |
|||
Poor (bush/no facility) |
32.7 (19/58) |
29.4 (5/17) |
32.0 (24/75) |
Good (flush/pit toilet) |
29.3 (109/372) |
12.8 (56/435) |
20.4 (164/804) |
Total |
29.7 (128/430) |
13.5 (61/452) |
|
p value = 0.000
From this table we can see that there is a decrease in prevalence of underweight in each of the education categories with improved sanitation, however the difference is much greater in the better educated group. This indicates that there may be interaction between the variables, we will explore this in question #2.
Question #2
This graph shows that there is an interaction between educational attainment and sanitation. The most straightforward interpretation would say that an improvement in sanitation will have a larger effect in households with better education than in households with low education.
Question #3:
| Regression Table: | ||||
Does educational attainment have an affect on weight/age z score controlling for SES? |
||||
| Dependent Variable: WAZ | ||||
| Independent Variable | Model Number: coefficient (t,p) |
|||
1 |
2 |
3 |
4 |
|
| Low education | -0.452 (-4.912, 0.000) |
- |
-0.420 (-4.456 0.000) |
-0.411 (-3.714, 0.000) |
| Roof | - |
-0.253 (-2.594, 0.010) |
-0.158 (-1.608,0.108) |
-0.136 (-0.780, 0.436) |
| Interaction (low edn * roof) | - |
- |
- |
-0.03280 (-0.155, 0.877) |
| N | 695 |
695 |
695 |
695 |
| Adjusted R2 | 0.032 |
0.010 |
0.034 |
0.033 |
From this table we can see that education remains significantly associated with weight for age Z score after controlling for socioeconomic status (roof as a proxy indicator). The interaction term is not significant, so we will ignore model four and look at model three alone for the analysis.
The general regression equation for education and SES is:
weight for age z score = A + B1 (low education) + B2 (bad roof/SES)
So if we put in the values from the regression table above then the equation will be:
weight for age z score = - 0.987 - 0.420 (low education) - 0.158 (bad roof/SES)