Hi all,
I have a database containing the history of personal evaluation of services. I would like to use the Bayesian average to compute the average evaluation for each service, in order to rank them. However, doesn't seem right to use Arithmetic means. Imagine that service s1 received 1 evaluation of 90%, so their evaluation rate will be 90%, but service s2 received 10 evaluations - 0.9, 0.5, 0.9, 0.9, 0.8- so their rate will be 80%.
Any help
I am not sure what you want. You could use a beta distribution to calculate the mean and variance/precision. Alternatively, you could do this in r with the brms package and estimate the posterior distribution. I did it with your data, see attachment. (I included marker for 90% and 95% credible interval, as well as an artificially chosen ROPE from 0.4 to 0.5, just for illustration).
Bayesian statistics is an approach to data analysis based on Bayes’ theorem, where available knowledge about parameters in a statistical model is updated with the information in observed data.
Referring to a situation with only one reviewer, the Bayesian Average might not be as effective because it relies on having multiple data points to make an accurate estimation.
- Badges
- Science topic