Graph: Mean WAZ Score by Measles Immunization and Age

Return to Two-way Confounding

The purpose of running a graph of these variables by age is to provide a comparison of the groups of immunized and not immunized for various age groups, so that proper age groups can be chosen for stratification in the regression analysis. It helps to see the graph of waz score (mean values) by age groups with separate lines for immunized and not immunized. Here is how:

1.  Open keast4j.sav

2.  Click on Graphs, Line…, Multiple and then in the area marked "Data in Chart are…" click in the round dot marked Summaries for groups of cases.

3.  Click on the box Define…

4.  In the area marked "Lines represent", click on the dot marked Other summary function and enter the variable à waz from the list to the left.

5.  In the box marked Category axis, enter the variable --> agegp (5 month age groups) from the list on the left

6.  In the box marked Define lines by , enter the variable  --> hmeasyn (measles immunization)

7.  Click on OK

The output graph will initially need a few changes to clean it up. Do the following three or four edits to clean up the graph: 

1.  Double click on the graph to open the Chart Editor (marked in the upper right hand corner of the screen)

2.  Double click directly on the lines inside the chart to open the Displayed data box. In the Series, Display box click on the missing mean waz and move it to the Omit box, then click OK.

3.  Click on Chart, Reference Line, Scale and Add 0, to give a reference line at the zero mark.

4.  Close the Chart Editor in the File menu to go back to the output screen. Does the output look like this? (The red marks are added to make a point, but should not appear on your output.)

wpe7.jpg (20303 bytes)

INTERPRETATION:

The output graph shows that measles immunization does not appear to have a profound association with nutrition status until around 12 months of age (see the difference in the pattern of the line for measles immunization versus no measles immunization).  This is likely due to the immunologic protection that a child has from the mother during the first year of life and then the vulnerability that is experienced as the child tries to begin tackling the "germs" in the world as he/ she begins eating other foods and contacting other pathogens in the environment.  A vulnerable child around 12 months to 36 months would seem to benefit most from the protective effect of the measles immunization.  It might helpful to look at these children as a separate strata (12-36 months) to see what the effect of measles is within that group.  Look also at a mean outcome of the measles variable to see the percentage of children in different 6 month age groups that actually receive immunization (below, multiply the mean value by 100 to get a %). Very few children are immunized before 12 months of age (most of those in the 6-11 month category received at 11 months, just before they turned a year).

wpe8.jpg (11608 bytes)

Return to Top of Page