Given Hapunda

102 randomly to the experimental and control or comparison groups. To that effect, they are not suitable for establishing the causal relationship between a programme or intervention and its effects. Additionally, they are unlikely to be statistically-representative, and thus generalising the findings is often problematic. Also, because of the multiple sources of data, the depth of the analysis, and the common desire to include multiple sites, case studies can be time-consuming and costly (Wholey et al ., 2007). Descriptive Evaluation Designs Descriptive designs are typically used for formative or process evaluations. Descriptive study designs can help you show whether your programme is operating as planned, provide you with feedback about the services you offer, determine whether your programme is producing the types of outputs and outcomes you want, and help clarify programme processes, goals and objectives (Project STAR, 2006). These designs include only people from the target population who are eligible to participate in some part of the programme (Campbell & Stanley, 1963). The purpose of a descriptive evaluation design is Focus Box 12: Case Study Design A case study design was employed in an evaluation study in Kenya with the aim of getting in-depth views about the effectiveness and sustainability of the interventions that had been put in place in the sector of housing, water supply and sanitation in Nairobi. This was to assess the medium to long term impact of the interventions. The study focused on seven sites: Two unplanned settlement sites, two middle income cites and three urban project cites. Data was collected through document review, personal and group interviews, observation of infrastructure and facilities, as well as though household surveys. The analysis was done at many levels including: the socioeconomic impact at household level; and the financial, economic, technical, environmental and economic impacts at neighbourhood levels. Sustainability and effectiveness were measured at each level (Morra & Friedlander, 1999) Focus Box 13: Descriptive Evaluation Design A descriptive evaluation was conducted by Anich et al ., (2005) with the aim of evaluating a routine Complete Blood Count (CBC) monitoring in the patients that were receiving anticoagulation therapy. This was done to describe any clinical outcomes related with this routine in a large diverse sample. An established non-profit organisation recruited patients who came in for the routine CBC which were repeated at a 3 month interval from the year 2000 to 2003. 4033 patients with both overt and occult bleeding were recruited for the study, of which 578 (14.3%) had experienced decrease in hemoglobin. Furthermore, 121 (3.0%) patients with occult bleeding also reported a decrease in hemoglobin, with 13 (0.3%) of these reporting major bleeding resulting in a significant drop in hemoglobin. However, it was concluded that although the CBC monitoring provides a way of detecting occult bleeding, the decrease in hemoglobin that was detected was low. Thereby rendering the routine clinically unnecessary as the clinical outcomes were too low, reporting 0.8% annual detection rate of occult bleeding.