0

I came across this task that I have no idea how to solve, because I'm not very good at statistics, so I was wondering if someone could help me understand it.

7 scientists with very different experimental skills were trying to find the same parameter. The answers they got were D = {-27.020,3.570,8.191,9.898,9.603,9.945,10.056}.

I need to find a posterior distribution of probabilities for this parameter.

After that, the 2nd, the 4th and the 7th scientists tried to find another parameter and the results that they got were {18.752, 12.450, 11.832}. Find a posterior distribution in 2 ways: 1) taking into account the first series of experimental data 2) and ignoring it. Summarize the results.

Glen_b
  • 257,508
  • 32
  • 553
  • 939
Pavel
  • 1
  • 5
  • 1
    Please read the `[self-study]` tags [wiki](http://stats.stackexchange.com/tags/self-study/info). Then tell us what you understand thus far, what you've tried & where you're stuck. We'll provide hints to help you get unstuck. – gung - Reinstate Monica Jul 05 '15 at 14:51
  • @gung if only I knew where to start, I took this task from my probabilistic programming summer practice class, where I just had 3 lectures and all the lecturer did was show us some code in python, with no proper explanation of what's going on, essentially I have to write a python program with plots, etc. What should I do? At first I thought I have to do curve fitting to understand what distribution this is, and then somehow go from there, but this goes way over my head, because I am bad at statistics. – Pavel Jul 05 '15 at 15:18
  • I have read this http://stats.stackexchange.com/questions/58564/help-me-understand-bayesian-prior-and-posterior-distributions but some of the things the person, who answered says don't make sense to me, because I'm not aware of how to guess what kind of likelihood this is, for example. I guess it's binomial because there are left handed and right handed people, but what is it my in case? – Pavel Jul 05 '15 at 15:27
  • But there are plenty of answers for that question, it would be closed with a link, saying "this has been answered here, please read examples and try to understand it yourself". You can close it if you want, I don't care. Also, following the guide lines of stack exchange, an title should be a question, and not a demand. – Pavel Jul 05 '15 at 23:02

1 Answers1

3

Though it's important to know what kind of parameter it is, this should be enough to get you started:

All these scientists are measuring the same parameter (as you're taking a practice class, I'll assume the measurements are from the same population and follow a gaussian distribution). Measurements... same mean, different standard deviation... Those couple key words should be enough to get your brain started and think of some possible ways to tackle this (I hope).

That being said, they should really give you and your colleagues the necessary statistical basis before asking you to translate statistics into software/code.

Glenn
  • 673
  • 8
  • 18
  • Yeah, I guess I can gather further from google. It's weird because I've googled this before and it didn't find anything. – Pavel Jul 05 '15 at 18:47
  • yeah I didn't use braces, that's why, I just entered "-27.020,3.570,8.191,9.898,9.603,9.945,10.056". – Pavel Jul 05 '15 at 18:58