Session: 1234567891011121314151617         Page 102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139 of 755

Or: “I didn't like what she said.” So that the answers broke down -- now I'm getting into the findings -- but answers broke down into two broad categories: questions having to do with form, and questions having to do with content. And if they didn't like the music or the sound effects, or they didn't like the announcer, that was on the form side. If what they said, or the story line, or something else, that obviously was on the content side. And we could begin to look at those things and make suggestions to the writer or to the producer.

At the same time that that tape was moving we were making a summary of all the movements of the needles. And we could give you a curve at the end of the broadcast, that showed where there were more pluses and minuses, and we had a profile of that particular program. I did more with it inside of CBS than Paul ever did with it. I don't know whether Paul did much with it in the Bureau, I've forgotten. If he did -- I don't even think he had a machine. I think he borrowed the one I'd developed. And by this time I hired someone in military research from Stockholm. God knows, I've forgotten how I got to even know the man. [] Torquist was his name.

Torquist?

I think Torquist. Don't ask me to spell it. But we had a section in the Research Department that did program testing. And the technique was to get a reasonable cross section of -- or a diverse group of people into a room, play the program (this is all radio), and have them sit there and make their reactions. Artificial to a certain extent -- sure. Better