Idea generation & Style

Arthur Tullett, 1997

Dr Tullett died before he could complete some projects that fascinated him. Knowing he was failing he sent these papers to us in 1998. We are now publishing them in this section of the website. This paper showed his interest in the A-I theory that the more innovative proliferate ideas but accept a higher casualty rate among them. The level-style distinction suggests that that the rate of success is greatly influenced by level, but the kind of success by style. He was unable to complete this study but below are some encouraging notes.

First he noted from our files data from three other sources.

Clapp:

In his studies in Shell, he noted that when the company appealed to the staff for ideas (in a project on cost saving) there was a large response. The more innovative produced more ideas than the more adaptive; the correlation was .36. He also noted that adaptors were more likely to produce adaptive ideas, as judged by him, with a correlation of .4.

Dubras:

In an exercise in Canadian Atomic Energy, a request was made from among R&D personnel for volunteers to form loose idea generation groups. R&D personnel return means of circa 102, and some data suggested the general group conformed to this mean. However, the mean of the volunteers was 114. No other data has been preserved, including total N.

Rosenfeld:

Noted that those who volunteered ideas to his idea generation unit in the R&D department in Kodak yielded a mean, of 113, which is remarkably similar to Dubra’s volunteers. He went further and divided this group into those contributing (over a few years) 1 or 2 ideas and those (fewer) offering 3 or more. Their means were 108 compares to 124; note, also, the closing of the gap between technicians and professionals (and managers, i.e., senior professionals).

Unfortunately no further data is available in our records. In none of these studies are success rates given; although practitioners report that adaptor rates of success, with their fewer ideas, is higher, as expected.

Tullet, working in the R&D of a company that wished for complete anonymity, collected the data below from 157 personnel, involved in a similar project.


No. of ideas Submitted

N
Mean
s.d.
None
(Believed to be mainly technicians)
34
103
17
1 or 2
Managers
8
113
12

Professionals
111
114
13

Technicians
18
104
18
3 or more
Managers
2
123
12

Professionals
15
123
15

Technicians
3
121
14