From kindergarten through graduate school, the most effective learning is usually, in various respects, social: in motivation and continuing stimulation from other pupils, teacher, parents, community; in guidings and elucidations from both classmates and teacher; in goal-settings and appraisals.

At this comparatively early and conceivably sometimes overenthusiastic stage of attempts at automation, might this "null hypothesis" be desirable: that students and teachers can (on the task in question) do at least as well without the special device or programming as with it! Only if results have warranted the rejection of this hypothesis, can development of special devices or programmings, or comparisons of short versus longer steps or objective versus write-in answers, continue without risk of being found fatuous.

A second "null hypothesis" is here ventured as desirable to keep in mind: that the device and program in total takes so much time and trouble (or is so expensive) that in cold fact it is impractical for everyday use by the people and in the circumstances for which it is intended. And more important than the expense is the trouble! Unless some of the current materials and procedures (hundreds of programming charts or frames to be kept in order and used just right) are made a lot less clumsy and easier to use, this second null hypothesis will get them.