Data management

For comparisons involving river and human data (see Objective 4: establish if a relationship can be determined between incidence of infection in snails and prevalence of infection in humans) total values summed over age and gender are required by site and month to match the river data stored in CS4Data1. Fortunately, these data had been calculated before the recording sheets had been lost.

It is worth describing the difficulties that occurred in managing the data. Data were collected from the three districts Koboko, Yumbe and Moyo which stretch over a distance of nearly 100 km. The senior author was based in the middle of these districts, namely Yumbe. He was provided with a community office in Yumbe for him to handle his paper work and to keep some field equipment. None of the districts had access to electricity and so, after completing the monthly field recording sheets, he travelled to Arua in another district where thermal power was available. Here he entered the data into his laptop.

The original recording sheets remained in a drawer in the Yumbe office. Later during the analysis for this case study he returned to Yumbe to collect the recording sheets only to find that the community organisation had relocated itself and had discarded a lot of its paper work. His recording sheets were among these papers!! The author's supervisors worked 800 km in Mbale district in Eastern Uganda; during one of these trips to meet them his laptop was stolen.

Although not all students will be expected to work under such difficult conditions as this, there are, nevertheless, many lessons to be learned. In particular, the case study emphasises the need for young researchers to be taught necessary data management skills during their training. Fortunately the mistakes that occurred, i.e. not ensuring that the raw data sheets were safely stored and not adequately backing up the computer data files, did not completely ruin this study and some potentially interesting results have been obtained.