Exercises in Targeting


Return to One-Way Page2


The Bangladesh data (bdeshc.sav) will demonstrate the exercise in targeting with consideration to both prevalence and raw numbers. Go through the following exercise to practice this concept, looking carefully at the diarrhea prevalence and numbers in males and females by district. First, use the prevalence and population less than 5 years to compute the numbers of children with diarrhea in the last 2 weeks.

          1.  Open bdeshc.sav

2.  Select Transform, Compute and in the Target Variable enter diarnum (number with diarrhea)

3.  For the population Numeric Expression box by selecting the variable for population-[pop] and prevalence of diarrhea-total- [diaprvfm] from the left hand variable box and inserting them with the arrow button into the Numeric expression box.
Formula: (pop* .2) *(diaprvfm / 100)

4.  Press OK

5.  Check for a new variable added to the end of the data set called diarnum.


Now, create a list of the number and prevalence of diarrhea using the Sort and Case Summaries options:

1.  Click on Data, Sort cases, and insert diaprvfm from the variable box into the Sort by box.

2.  Click on Sort order Descending.

3.  Click on OK. (This will sort from high to low for diarrhea prevalence by district)

4.  Click on Statistics, Summarize, and Case Summaries.

5.  Enter the following variables in this order into the right hand Variable box using the arrow button:


6.  Do not enter a variable into the Grouping Variables box. Click on the Options button and de-select the subheadings for totals. Click Continue.

7.  Click on the Statistics button and remove any variables from the Cell statistics box, then click on Continue and then on OK.


Here is what the result looks like:


wpe5.jpg (50593 bytes)

wpe6.jpg (19462 bytes)


The list is in order of descent by prevalence of diarrhea; therefore the first district Nawabganj has the highest prevalence of diarrhea. Despite this, Nawabganj does not have the highest absolute number of cases of diarrhea. The districts of Chittagong, Comilla, and Kishoreganj have the largest numbers of affected children. If an intervention program can only reach 20 areas, listing the areas in such a way can help in weighting the areas for targeting those that need it the most. It will also depend on the nature of the illness being targeted including the communicability of the illness and the current access to health care.

Return to Top of Page