Modeling Epidemics

Posted May 22, 2017 by sgibbs

Reflect on the following questions:

  • What assumptions are made in your model?  
  • Was there any key factor that was left out that might be important in the real world?
  • Would you trust a model if your life depended on it?

Comments

Submitted by eso on Thu, 06/15/2017 - 16:10 · Permalink

This model assumes that people can only get sick through direct contact with an affected individual. This neglects other methods of transmission, such as air exchange and surface contact. It also includes the assumption that recovery is random and doesn't have a minimum or maximum on the recovery time. This makes me also realize that people don't immediately become affected upon contact. The sickness agent has to enter your body and become established before you experience any symptoms. 

I'm not sure I would trust a simple model if my life depended on it because many assumptions are being made and many other factors are being neglected. As for a sophisticated model, I would have more trust if all of our discovered knowledge was incorporated. 

***Something I noticed with my contagion model is that I have to set "Recovery Rate" to a very low value (below 10) and "Transmission Rate" very high (100) to prevent the sickness from disappearing. Even then, the red turtles disappear in less than 10 seconds in most cases. I guess this is a good thing as long as it is a harmful sickness. I'm curious about what actual transmission and recovery rates are. I also realize that population density is a key factor in this simulation. 

Submitted by kcwgtt on Thu, 06/22/2017 - 21:10 · Permalink

  • What assumptions are made in your model?  The turtles can become infected and recover multiple times. 
  • Was there any key factor that was left out that might be important in the real world? If there are turtles that could be immune. The amount of time to become infected or recover.
  • Would you trust a model if your life depended on it? No, because I do not like that the model allows the turtles to be reinfected. Also, the ticks do not correlate to a time period.

I agree with you about the recovery rate. I am wondering how to make the recovery rate more accurate. As it is presently constructed, the recovery rate would not be very helpful in the experiments to be simulated.

Submitted by person11d on Sun, 08/06/2017 - 14:28 · Permalink

What assumptions are made in your model?  

One assumption is that the disease is contracted through contact with the infected individual, where as some diseases can be passed through the air or by coming into contact with an item an infected person has touched. So in reality, the turtles do not need to physically touch to get sick. 

Was there any key factor that was left out that might be important in the real world?

I think a key factor would be knowing exactly how contagious the infection is and how effective the recovery rate or treatment is. If there is a vaccine for it, programming that into some of the turtles would help to determine how likely the disease is to spread through the population. Another key factor is that a person is less likely to contract the infection again as they have hopefully developed an immunity to it, and if they did get sick again their recovery should be quicker. 

Would you trust a model if your life depended on it?

I wouldn't trust a model with my life because in the real world there are small factors of chance that we may not be able to anticipate to program into the model in the first place. The model can only do what we tell it to do, so we cannot program it to include such random changes such as a virus' ability to mutate to be resistant to vaccine or treatment.