Recent interest has been noted in the evaluation community in expanding the focus from program implementation and outcomes to program design and planning. One important step for moving in this direction is to examine existing evaluation models and to assess their relative strengths and weaknesses for planning purposes. This article presents a comparative case study of applying logic models and the action model/change model schema for planning the Learning Community Program in Taiwan. Lessons learned from these applications indicate that logic models are relatively easy to learn and effective for identifying major program components and indicators, but not sufficient for articulating the theoretical significance of the program. On the other hand, the action model/change model schema requires more time to learn and practise, but it has relative advantages for providing theoretical insights into contextual factors and causal mechanisms of the program, unlike logic models. This comparison can serve as a guide for evaluation practitioners when selecting evaluation tools to apply in planning and/or evaluating their programs. Résumé : De plus en plus les évaluateurs semblent orienter leur intérêt vers la conception et la planification de programme plutôt que sur l'implantation et les effets. Une étape importante de ce virage est d' examiner les modèles d' évaluation existants et d' évaluer leurs points forts et leurs points faibles pour l' exercice de planification. Cet article présente une étude de cas comparative de l'utilisation des modèles logiques et des modèle d' action/de changement aux fins de la planification d'un programme d' apprentissage communautaire à Taiwan. Nos résultats indiquent que les modèles© 2018 Canadian Journal of Program Evaluation / La Revue canadienne d'évaluation de programme 33.1 (Spring / printemps), 49-68 figure 1. Action Model/Change Model Schema Using Logic Models and the Action Model/Change Model Schema 57 CJPE 33.1, 49-68 variety of items. This is especially true for the activity component. It seems to me anything we do can be put in this component. In this case, the activity component could become a garbage can. I think the schema overcomes this problem by differentiating components between the action model and change model.

No abstract