In an education process either in school, in-company or other areas, it is not easy to evaluate human learning effectiveness, This paper presents the results of an experimental study on educating people to read the scale of a measuring device. The device was selected because it was easy to measure people's reactions. Various educational effects such as educating methods, time-passage or psychological aspects, etc., were studied in the experiment. There was a significant range in people's reading error which was one of the largest sources of measuring variation. The characteristics used to evaluate the educational effects were : (1) the final scale reading,and (2) the square-root of"work"done before reading the scale. The former represents the reading observed (consciously) by the people, and the latter is the calculated results indicating the sub-conscious feeling of the people while adjusting the measuring device. ln this study, this represent the measurement stability.