Conference rusure::math

231.0. ""Infection" model for Knowledge" by ASGMKA::WHITE () Thu Mar 07 1985 13:47

I am looking for help/suggestions/criticism of a description of the knowledge
transfer process. I think an understanding of how knowledge transfers between
people will help us make more effective use of technology (at least here in PSG)
There may be many stones left unturned, but as far as I looked I could find no
epidemiological model for disease/infection transmission that may be modified
for this purpose. A brief description follows and I will be happy to send a
copy of the derivation and equations to anyone interested.

Assumptions: 1)That individuals have a co-efficient which describes their
               immunity from "infection" (I) which represents their resistance
               to learning. 
             2)That each body of knowledge (disease) has a co-efficient of
               contagion (C) which represents how easy or difficult it is to
               acquire a useful working familiarity with the knowledge.
             3)Both co-efficients are normally distributed among the population
               with a range between 0 and 1 (mean 0.5, SD 0.1667).
             4)Knowledge is transferred as a result of contact with an infected
               individual. Thus as the infection spreads, the probability of
               contact with an infected individual increases.
             5)The initial contacts from whom the knowledge spreads are n
               individuals from a population N.

I ended up with:

     P(transmission) = I * {1- sum from j=0 to pc [(1-Cj)]}

           where pc  = p * sum from k=0 to t [p^k-1 * (n * u)/ N]

                  p  = Mean number of contact events per time unit t
                  u  = Mean of joint probability distn of I and C


*****NOTE ERROR IN 1ST EQUATION ****  "sum" should read "product"

The two co-efficients contain a multitude of components including the background
of an individual, their need to learn, skill of the "teacher", documentation
for the tool or technique being taught, etc. 
If anyone has any comments or ideas, please let me know. Derek White.

       How do you distinguish between "knowledge transfer" and background
    noise, which is not accounted for in your equation?