Greetings all,
I am investigating the topic of emotional contagion. While there are numerous papers that refer to it as having the potential to spread similarly to an infectious disease, I have not yet been able to find any computation models that try to predict the contagion rate.
Essentially what I'm looking for is "are there any computational models that can predict the potential viral impact of a negative mood".
Thanks in advance for any assistance.
Dear Erika, at a first glance I would say yes, you can imagine to apply some infection-related model (like the SIR model, see https://en.wikipedia.org/wiki/Compartmental_models_in_epidemiology) to this process. The real problem is that, while you are likely to obtain results in qualitative agreement with real situations, a quantitatively correct prediction would require to know a number of parameters which are typically difficult (or nearly impossible) to estimate.
This is a very interesting question. I think Roberto Ferretti's comment about parameters points toward an approach that is based on polling and data analysis (unlike SIR). He argues that parameters may be impossible to estimate. However it may be possible to creatively identify parameters that can actually be researched. There may be interacting parameters that are correlated with mood contagion. Here are a few-
1. The number of individuals in a population.
2. Population demographic parameters (gender, age, diversity, etc.)
3. Type of government.
4. Written and spoken languages.
5. Unemployment level.
6. Property crime rate.
7. Rate of crimes against persons.
8. Income distribution.
9. Political party affiliations.
10. Family size and age distribution.
11. Number of families with television, radio and Internet.
12. Number and type of social media.
13. Homeless percentage.
14. Gun ownership percentage.
15. Distribution of education years.
etc., etc.
These parameters, and more, could be entered into a well-designed relational database that can be queried. When a mood altering event or issue occurs, the population can be mood-sampled (by polling), and the samples can be entered into the database as dependent variables. The data can be analyzed statistically to create regression models, or used as training data for an AI model.
Dear Erika, there is something better than Rt that has many limitations, in the case of mood the major limitations is the supposed ergodicity at the basis of Rt that is seldom verified in case of epidemics and never verified by mood spreading that happens on networks or, in any case, with very different spreading potential of the different 'infected' units.
This is why you should look at the papers related to the spreading of a signal across a social network, see for example:
Article Spreading of healthy mood in adolescent social networks
Article Emotions as infectious diseases in a large social network: T...
https://www.bmj.com/content/337/bmj.a2338+
https://www.pnas.org/content/107/12/5334?__hstc=3584879.07430159d50a3c91e72c280a7921bf0d.1523145601939.1523145601940.1523145601941.1&__hssc=3584879.1.1523145601942&__hsfp=1773666937
Ciao
Alessandro
- Badges
- Science topic
- Similar topics
- Papers