But at the same time, these analyses of specific stages in the policy process model had a clear downside in that they oriented scholars toward looking at just one stage at a time (deLeon, 1988), thereby neglecting the entire process. Ultimately, many policy researchers (and policymakers) came to view the process as a sharply differentiated set of activities: First, you define the problem; then, a completely different set of actors implements the chosen policy option; a third stage defines the evaluation; and so on. Likewise, they portrayed a disjointed, episodic process rather than a more ongoing, continuous one, as well as a policy phenomenon that seemingly took place in the relatively short term, one more suitable to the policymaker's rapidly changing schedule than the life span of a given policy. Finally, to many, the policy process/stages image implied a certain linearity-for example, first initiation, then estimation…..then (possibly) termination-as opposed to a series of feedback actions or recursive loops (e.g., estimation can lead back to initiation rather than the next step, selection, and implementation and evaluation insistently feed back and forth on each other) that characterize the operations and politics of the policy process.