The International Potato Center (CIP) was established in 1971 and seeks to reduce poverty and achieve food security on a sustained basis in developing countries through scientific research and related activities on potato, sweet potato, other root and tuber crops, and on the improved management of natural resources. The centre has been running a five year initiative, the Sweet potato Action for Security and Health in Africa, designed to improve food security and livelihood of poor families in sub-Saharan Africa by tapping the unexploited potential of sweet potato. However, the program face several challenges in design, implementation, monitoring and evaluation of its projects which bordered on among others; lack of centralized project design and planning office which would be responsible for provision of guidelines and creation of a data base of projects progress, absence of data management support assistants to oversee the design of data collection instruments across the projects, capacity building of enumerators and data entry clerks, data collection, data organization, entry and analysis. The objective of this study was to enhance effective use of research methods among researchers through training of researchers in data management practices, reinforce the data checking and organization before analysis and evaluation and reviewing of project proposals and literature on sweet potato seed systems. Training followed the training cycle spanning training needs assessment whereby training activities, needs and goals were determined through administration of a structured questionnaire to nine data entry clerks and four researchers. The aim was to establish knowledge gaps among other areas of interest in use of CSPRO in data management. Training evaluation was done through administration of a structured questionnaire to establish the achievement of training objectives. Finally, an adoption survey was done as part of follow up to gauge the acceptance of the data management tool among researchers. The role of data entry checks was demonstrated through analyzing the Tanzanian baseline survey data where two case scenarios were looked into. In one case data was entered into CSPRO without data entry checks and the second case the same data was entered with entry checks inbuilt before data entry. Two categories of checks were used; checks before data entry which include range checks, automatic skips, numeric codes checks, missing and not applicable checks were used. The second category was checks after data entry; double data entry and exploratory data analysis which were used to compare the output from the two case scenarios. A criterion to evaluate project proposals was developed and used to evaluate two proposals. It considered general components of the proposals as well as specific components, which were used to grade the proposals. Literature review was done on sweet potato seed systems. Three steps were adhered to in compilation of the sweet potato seed system paper; first, both published and unpublished material on sweet potato seed system were identified, second step entailed gathering the relevant information from the materials and finally writing up of the paper. Results from the follow up study of the training showed that data from twelve projects were managed by researchers using CSPRO. The quality of the data output and limited data queries during analysis were some of the reasons behind majority recommendation of CSPRO software. However, the researchers identified some of the challenges to adoption of the software as resistance from scientists and lack of a functional data management unit among others. Data entry checking was found to play a significant role in data analysis through increasing plausibility of the output and subsequent conclusions as evidenced by significant difference in output. At 5% both t test and chi square tests performed on two cases (with and without entry checks) gave a p value of <0.0001 which led to the conclusion that indeed there is a difference between output from checked and unchecked data. The project proposals evaluated were ranked above 50% threshold on both the general and specific criterion with proposals reviewed scoring 100% on the specific criteria but <100% on the general criteria. In order to enhance effectiveness of the research teams involved in these research projects it was therefore, recommended that CIP should design and promote training on data management, scientists to embrace data validation and organization and formulate a standardized criteria of evaluating and reviewing proposals and literature.
Date of publication:
RUFORUM Theses and Dissertations
Dr. Aggrey Bernard Nyende, JKUAT, and Dr. Julius Sindi Kirimi, International Potato Centre