We can believe that the activities of students working together, asking each other questions, reading and writing from choice lead to a higher level of attention to the language. But this is something intimately connected with the internal consciousness of the students. It cannot be measured directly. (Such methods as measuring the state of arousal through measuring changes in the galvanic resistence of skin caused by varying production of sweat (5) can only be very indirect - it has been described as not much more precise than studying a computer program by measuring the heat it gives off.)

Nor can we define precisely at any given moment what is happening linguistically to a student. Language learning is not definable except in general terms. Hence the difficulty of research into learning in general.
Researchers can only measure what is measurable, but what is not measurable may be more significant. The difficulty is finding a procedure for assessing methods and presenting it to a scientific community mainly concerned with external measures, when education and learning are concerned to a large degree with internal matters of consciousness. Only the external signs of consciousness can be measured or observed. There may be other non-measurable, internal, effects of changes in consciousness.

Not just scientists but also administrators and politicians are usually concerned only with the measurables. In the case of language learning the customary methods of testing used, for example in what was then the East African Certificate of Education, are those of performance of certain skills. Of necessity (mainly cost) the types of performance which can be measured are limited. Public speaking, facility on the telephone or conversational ability cannot be assessed at a cost low enough for a public authority to be able to afford. Enjoyment of literature and, probably, dramatic ability are unassessable at any cost. Yet both of these categories may well be of value to the student in his personal life, and also to society. I think we need to remember that in the face of pressures by such departments as the Treasury there are arguments for not being solely concerned in assessing learning with what is measurable in monetary and other terms. But in a time of fashionable preoccupation with costs, it is difficult politically to justify these unquantifiables.

What might be describable, though not measurable, are the after-school life effects of the method. These also are difficult to define and have to include the former students' own memories and opinions of what happened. This dissertation is concerned with an attempt to find out what the former students think of their English course and whether it has benefitted them.

(5) Robert E. Ornstein ed. Nature of Human Consciousness p. 83