0

model and modeling seem identical to me. Aren't those really same ? (or is there any flaws so that they are two different tags.)

And in model tag, it is written that there is a distinct element of $P$ (probability) which generates the observed data. Though statistical model involves a random error term ($\epsilon$), will I be able to generate the same observed data for a certain probability ?

And in Wikipedia, it is written that

A statistical model embodies a set of assumptions concerning the generation of the observed data, and similar data from a larger population.

When should i get the observed data and when should i get similar data ? For different probabilities or only one probability can generates different set of data , as statistical model involves random error term ?

Can i see it empirically ?

Community
  • 1
user 31466
  • 1,197
  • 13
  • 31
  • 1
    The tag descriptions make them seem the same, unfortunately. I'd have said they were *quite* different; one is for questions related to a model (a noun), while the other is for questions related to the process of *modelling* - of arriving at a model (a verb). The destination and the journey are different! [However, if your question is mainly about possible tag overlap it belongs on meta.stats.stackexchange.] I can't figure out what the rest of your question is even asking. Can you clarify? – Glen_b Feb 11 '15 at 07:16
  • @Glen_b I wanted to mean by the rest of the question is that can i generate the same data from a statistical model ? I doubt, because this is not deterministic model. – user 31466 Feb 11 '15 at 07:23
  • @Glen_b Can i be referred of an example of `modeling` please? What are the steps of modeling ? How to arrive at a model(destination) from data ? – user 31466 Feb 11 '15 at 07:28
  • 1
    Questions of the form "help me come up with a model for this" (like [this](http://stats.stackexchange.com/questions/134197/model-suggestion-for-temporarily-broken-device)) or "How do I choose which predictors to include in my model?" (like [this](http://stats.stackexchange.com/questions/134648/finding-significant-predictors-of-psychiatric-readmissions)) are `modeling` questions. Your 2nd and 3rd questions there are *examples* of modeling questions, but they're too broad for a question in comments -- indeed, I think maybe even too broad for either of them posted as a question in their own right – Glen_b Feb 11 '15 at 07:48

0 Answers0