Session: 1234567891011121314151617         Page 102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139 of 755

Right. Yes. There is a big difference. Well, could you talk a little bit about that, about what you did. How would you do that? How would you take the data that you gained from the program analyst and --

Well, first of all, you had the overall raw vote of liking it or disliking it. One of the first things that has to do with success is the subject has to like the program. And if you added the pluses and minuses, you had a composite curve, a graph that went up as the program went on, or it went down. And if it went down, that was the first sign that you were in trouble, or that the program was in trouble. That was a crude rough composite measure. But the main thing that came out of the research was the places where people went from one position to the other in terms of liking or disliking. And that gave the writers and the producers insight about what was on the record, or what was the record in those days -- we didn't have tape.

END OF SIDE TWO, TAPE ONE; BEGINNING OF SIDE ONE, TAPE TWO

Okay. Did you use some of these insights in terms of the development of new programs, or in trying to analyze generally what kinds of programs succeeded over other kinds of programs?

No, and I think I would be -- as dedicated as I was at that time to research -- I think I would be perhaps not the first, but I would certainly be in the vanguard of saying research doesn't create. I remember an interesting discussion I had with Gene [Eugene W.] Kettering, [or Charles F.? -- an editor] who was then a superstar at General Motors. He had developed a lot of their engineering leadership and market research on automobiles, not on