This study was concerned with evaluation of Vygotsky's assumption that assessment of how students respond to socially structured instruction within some domain of interest would provide important diagnostic information about them. A kind of titration procedure was used because the objective of this study is not how much improvement one can bring about via a specific amount of intervention but rather how much aid is needed to bring about a specific amount of learning. The subjects were 24 1st grade children and 2 teachers. Children were devided in two groups which were called "ZPD group" and ordinary group respectively. While the ordinary teacher was in charge of ordinary group, the ZPD group were put in an interactive learning situation under the guidance of "ZPD teacher" who was specially t rained in Vygotsky's ZPD theory and its evaluative procedure. The experimental task was arithmatic items structured by homeroom teacher based on 1st grade arithmatics text. Following a pretest, the children entered into the training phase of the test- train -test assessment procedure. In the initial stage a problem presented without aid and the teacher waited until the children solved it or failed it on their own. If the child failed, a series of graded a ids were given until a solution was reached. The results of this study are as following: The mediated aids of ZPD teacher were more appropriate considering Vygotsky's ZPD conceptualization, i. e. , prompts were provided more in the region of sensitivity compared with ordinary teacher. Thus the tested performance of ZPD group were superior to ordinary group. Even though the initial pretest scores were identical among children, the needed input of prompts were quite different, suggesting that amounts of aids were better indices and more diagnostic measure of children's learning potentials.