Lekvall & Wahlbin (1973) have suggested that the right-hand skew often observed in growth and diffusion curves can be assumed to be the overt outcome of a combined exponential and logistic process. In this article it is argued that the skew in the cumulative distributions can also be accounted for not only by structural heterogeneity in the target population, but also by dynamic heterogeneity when the population changes as the process goes on, or by changing stimulus effects. Two simple models-one modifying the logistic and the other modifying the exponential function-are introduced. They can account not only for right-hand but also for left-hand skew, and have the logistic and exponential functions, respectively, as special cases. Since both of the models can give rise to S-shaped and J-shaped curves, it is argued that the shape of the growth curve in itself provides little information about the underlying process. The models can be combined, after which they yield a Riccati equation. The first model is illustrated by the spread of TV ownership in Norway.
[1]
Per Lekvall,et al.
A STUDY OF SOME ASSUMPTIONS UNDERLYING INNOVATION DIFFUSION
,
1973
.
[2]
G. Hernes,et al.
THE PROCESS OF ENTRY INTO FIRST MARRIAGE
,
1972
.
[3]
Nathan Keyfitz,et al.
Introduction to the mathematics of population
,
1968
.
[4]
Thomas F. Dernburg.
Consumer Response to Innovation: Television
,
1957
.
[5]
F. E. Croxton,et al.
Applied General Statistics.
,
1940
.
[6]
H. Pemberton.
The Effect of a Social Crisis on the Curve of Diffusion
,
1937
.
[7]
H. Pemberton.
The Curve of Culture Diffusion Rate
,
1936
.
[8]
R. Prescott.
Law of Growth in Forecasting Demand
,
1922
.