Certain students put forth research questions that dictate a research framework capable of evaluating the impact of an agricultural intervention or an outcome. Knowledge of impact evaluation methodologies and methods is key to effectively addressing such questions. This chapter first discusses impact evaluation and argues that it is a research activity involving establishing a causal relationship. The theory of change is an important concept in understanding and conducting impact evaluation. Basic types of research designs for evaluating impact are categorized into three: True experiments, quasi-experiments, and preexperiments. The main types of true experiments include “the randomized controlled evaluations,” normally labeled (RCT), and “the randomized posttest group” and the “randomized pretest–posttest control” group. General guidelines for RCT implementation, randomization, and steps in conducting RCT are discussed. Quasi-experiments are mid-experiments that do not involve randomization of participants. They are, therefore, weak; propensity score matching and difference-in-difference are common methods employed to create a counterfactual and strengthen the study. The preexperimental designs involve simple before-and-after comparisons. Qualitative approaches are indispensable for revealing “how” and “why” changes occurred and understanding participants’ experiences and perceptions. Finally, mixed methods in measuring impact are increasingly gaining recognition as they allow more holistic impact evaluations.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Research Designs for Impact Evaluation Studies

  • Jayne Njeri Mugwe,
  • Steven Runo

摘要

Certain students put forth research questions that dictate a research framework capable of evaluating the impact of an agricultural intervention or an outcome. Knowledge of impact evaluation methodologies and methods is key to effectively addressing such questions. This chapter first discusses impact evaluation and argues that it is a research activity involving establishing a causal relationship. The theory of change is an important concept in understanding and conducting impact evaluation. Basic types of research designs for evaluating impact are categorized into three: True experiments, quasi-experiments, and preexperiments. The main types of true experiments include “the randomized controlled evaluations,” normally labeled (RCT), and “the randomized posttest group” and the “randomized pretest–posttest control” group. General guidelines for RCT implementation, randomization, and steps in conducting RCT are discussed. Quasi-experiments are mid-experiments that do not involve randomization of participants. They are, therefore, weak; propensity score matching and difference-in-difference are common methods employed to create a counterfactual and strengthen the study. The preexperimental designs involve simple before-and-after comparisons. Qualitative approaches are indispensable for revealing “how” and “why” changes occurred and understanding participants’ experiences and perceptions. Finally, mixed methods in measuring impact are increasingly gaining recognition as they allow more holistic impact evaluations.