Study questions

5. The researcher reports that the Buhlungu dip-tank area in Lobamba Sub-region was selected purposively because of a particular breed of cattle known only to occur there. The other dip-tank areas were chosen at random. Decide how you would use this information to revise the calculation of the population total and standard error for Lobamba. Obtain new population estimates.
6. In view of the few donkeys kept by farmers in Swaziland it is decided to analyse the proportion of homesteads owning cattle per dip-tank area rather than the number of donkeys owned per homestead. This then reduces to a single-stage analysis at the sub-region level. Write down the formulae you will need to calculate the population average and its standard error. Apply this formula to Malandzela Sub-region. Multiply the population average and its standard error by the number of homesteads to get an estimate of the sub-region confidence range. Compare with that already obtained.
   

7. It is decided to select instead 6 dip-tank areas within the Lobama sub-region at random so that the chance of each dip-tank area being selected is proportional to the numbers of homesteads in that dip-tank area. Describe how you would do this.
   

8. Check the raw data for chickens and delete any homesteads with extremely high numbers of chickens that might be considered to be outliers. Repeat the calculations of the two population estimates for chickens. Do they seem more reasonable?