Word-of-Mouth (Turtles) Experiment
Return to Current Research
Return to Winzar Research
An idea disseminates through a community through strong ties (friends) and weak ties (casual acquaintances.
What affects the speed of dissemination?
created with NetLogo
view/download model file: Word-of-Mouth-Replication.nlogo
Word-of-Mouth-Replication is a replication of a study reported in:
Jacob Goldenberg, Barak Libai and Eitan Muller (2001), "Talk of the Network: A Complex Systems Look at the Underlying Process of Word-of-Mouth</a>," Marketing Letters, (12), pp. 209-221.
--- quote ---
The strength of weak ties
There is scant evidence on the breakdown of the personal communication between closer and stronger communications that are within an individual's own personal group (strong ties), and weaker and less personal communications that an individual makes with a wide, often random, set of other acquaintances and colleagues (weak ties). We model these two phenomena and show that the influence of weak ties is at least as strong as the influence of strong ties. Despite the relative inferiority of the weak tie parameter in the model's assumptions (strong ties reflect greater probability for an individual-level transformation), their effect approximates or exceeds that of strong ties, in all of the process stages.
--- unquote ---
Five parameters control the diffusion of knowledge through a community:
In the early stages of new product diffusion then advertising clearly has the greatest influence (as no-one knows about it they can't talk about it), but word-of-mouth quickly takes over.
Market Size adjusts the number of people in the market. A very large number makes the model VERY slow with a very small improvement in consistency of output. I suggest leaving it at about 1000.
Size of Personal Network is the number of people in constant, highly-regarded, contact. It ranges from 5 to 29 consistent with Goldenberg et al. A larger number implies faster communication within the clique.
Size of Weak Contacts is the number of people that each person speaks with in each period from outside the clique. It ranges from 5 to 29 consistent with Goldenberg et al. A larger number implies faster communication between cliques.
Weak Ties Effect is a parameter affecting the influence of the number of weak contacts on awareness likelihood.
Strong Ties Effect is a parameer affecting the influence of the number of strong contacts on awareness likelihood.
Advertising Effect is a parameter representing the level of information received by all citizens in each time period.
Note how quickly the diffusion process works when more people are added to the Strong and Weak ties.
If you have the patience, you'll see that the total number of people in the "world" makes little difference to the number of clicks (or time periods) that are taken in any one experiment - but it does have a huge impact on processing time.
I have experimented with using a much wider range than is used here. This follows the Goldenberg paper fairly closely. However, it is clear that the relationships are logarithmic rather than linear, so I've played with multiples.
This section could point out any especially interesting or unusual features of NetLogo that the model makes use of, particularly in the Procedures tab. It might also point out places where workarounds were needed because of missing features.
This section could give the names of models in the NetLogo Models Library or elsewhere which are of related interest.
Jacob Goldenberg, Barak Libai and Eitan Muller (2001), "Talk of the Network: A Complex Systems Look at the Underlying Process of Word-of-Mouth</a>," Marketing Letters, (12), pp. 209-221.
(c) Hume Winzar