Data management

We do not go into detail here about the way the data were managed, except to report that difficulties were encountered in handling the vast amounts of quantitative data, in particular those arising during the longitudinal monitoring. To some extent this delayed analysis and the researchers had to rely on talks with farmers to gather information on the progress of the trial.

Where survey data are collected, especially of the longitudinal nature that featured in the main part of this study, it is essential that there are adequate human and computer resources to enter and process the data at the speed required by the project.

In retrospect, the researchers were perhaps somewhat ambitious in the quantities of data that they expected to be collected. Less intensive data collection would have resulted in an easier study to manage and could have achieved the same objectives.

Both the changes in the study design brought about by some control farmers deciding not to adhere to the feeding protocol, and the heavy data handling requirement that resulted in long gaps in getting reports from the analysis, made it necessary to obtain additional information from farmers and co-operatives

Additional data had to be collected retrospectively in a questionnaire prepared for the final cross-sectional survey to evaluate sustainability of the impact when it was realised that additional information on concentrate purchase and use was needed to understand some of the results that had been obtained.