Data management

The raw data are contained within CS8Data1. The series of spreadsheets show how data can be summarised through use of the Excel pivot table facility and put into a separate sheet for analysis. By tabulating three ways by tag no, block and treatment the pivot table can check that the data coding and data entry are correct. For instance, if the contents of the pre-experiment pivot table are change to 'count' from 'average', it can be seen that every frequency count occurring within the body of the table is 15. Had there been any other values then the user's attention will be drawn to possible mistakes in coding or possible missing records.

This case study also provides an illustration of the level of precision to which data should be recorded. Notice that milk yields were recorded to the nearest half a gram and that the average milk yields and the values for dry matter intake are stored with one decimal place. In retrospect the precision with which the data were recorded was greater than was needed. One only needs to look at the range in values to realise that whole numbers would have been adequate. It was appreciated, once this study had been completed, that it was wasteful to have entered more significant figures than were necessary. The level of precision to which data are recorded is an important point to be considered by the researcher when designing a data recording sheet.

The mean values used for analysis are stored in the worksheet 'Mean data' in CS8Data1. Treatments are codes 1, 2 and 3. In order to facilitate interpretation of the output this file has already been read into GenStat and two new columns Diet30 and Diet60 formed to describe the diets as fed according to the letters A, B and C used in the design table. Factors have already been defined and these are indicated by a ! following the name. This file is stored in CS8Data2 and described in CS8Doc2.