Given Hapunda

88 Normative Questions These define what is best in a given situation. They simply compare what is with what should be. They compare the current situation with the ideal situation i.e., a specified target, goal, or benchmark. They usually answer questions about inputs, activities and outputs. Types of Evaluation Study Designs After identifying the evaluation questions, one can now choose the evaluation design most appropriate for each given question. A design is simply a strategy that one uses to integrate all components of a study in a coherent and logical way for purposes of effective problem-solving (Leedy & Ormrod, 2012; Morra-Imas & Rist, 2009). Evaluation designs can be grouped into three categories: Experimental designs, Quasi-experimental design and Non-experimental designs. Each design has problems related to validity. However, among all the evaluation designs, the randomised experiments are the strongest in terms of ascertaining causality as they have the least threats to external and internal validity (Marlow, 2011). Experimental Design Also known as True experiments or Randomised controlled trials (RCT) or Randomised clinical trials (RCT) in medical settings (Leedy & Ormrod, 2012). These are designs that are used when answering cause-and- effect related questions. They are often referred to as the “gold standard” designs, to which all other designs are compared. These designs are stronger than quasi-experimental designs in terms of establishing causality or making causal inferences (establishing cause-and-effect relationships), and eliminate majority of the threats to external and internal validity. It must be noted that in evaluation, these designs are only used for new interventions that have not been applied before (Marlow, 2011). Once the intervention has been established that it works, designs that apply normative questions are then used in subsequent evaluations. In true experimental designs, evaluators must be able to show that the intervention is the cause of the results obtained, and that without it the desired results would not have been acquired (Bryman, 2008). To prove that the intervention is the cause of the result, an experimental design must compare the results of two groups; one that receives the intervention (intervention or treatment group) Focus Box 3: Examples of Normative Questions 1. Did we empower 90% of the youths as required? 2. Are we heading towards our target(s)? 3. Are we doing what we are supposed to be doing? 4. Did we reach our objective of vaccinating 20,000 cattle?