Study questions

5. One way round these small herd sizes would be to include all the data in the analysis, including first parity cows, to estimate mean values, and then to select for ranking purposes only those cows with a parity greater than one. Use the whole data set and include parity as a fixed effect together with herd group and herd within herd group. Select the top 40 cows.

6. By retaining the four cows with high 'outlying' values in the analysis we have raised the average values for the herds in which the animals belong and hence reduced the chances of other cows being selected from the same herds. Delete these four cows from the data set (but keep them as the top ranking cows) and select an additional 36 cows from an analysis of the remaining data set. Is there any change in the result?

7. Question 6 could have been done by retaining the key statements in the original Input log produced in the case study, and reusing them with the 4 cows deleted. Redo question 6 applying this approach.

8. Discuss whether you consider this study to be of the form of a survey or observational study. What other information would you preferably need in order to use these results to estimate the average lactation offtake yield of Ankole cows in the district? Do you think that the method of selection of herds for the study justifies estimation of such an average?

9. Design a data recording sheet for recording milk offtakes for this study. Information is also needed on dates of calving, completion of lactation, disposal of cow from herd as these occur. Discuss how you might use Excel to store the data so that they can be used eventually to generate the spreadsheet of CS2Data1. Do you think this is a sensible method of managing the data?