Findings, implications and lessons learned

Some statistical implications of this case study are as follows.

  1. The case study has demonstrated how to handle multiple response data and describes how they can often be mistreated, especially when the wrong choice of the denominator is made in calculating percentages. It is important to code the questions in a way that recognises the multiple levels in the data. For example, the data in this case study were recorded at the homestead level, as homesteads were the ultimate sampling units, but for certain questions such as the members of the homestead responsible for cattle activities there was the activity level underneath the homestead level, followed by the homestead member category level (adult males, adult females, boys, girls or hired labour) underneath the activity level.

  2. The case study has also demonstrated the value of combining multiple ranked responses into the form of an index which adds to the interpretation of the relative importance of different responses.