0

Let's take an example from geology.

I describe an area (A) using a topographic map not really well detailed (50m of resolution). In this area, I study variably-sized shapes of geologic structures (e.g., volcanoes). I have also from A, an area B that is way more detailed (1m-resolution) and it has been described by direct observations (on the field). Now, I want to do analyzes of structures within A, given the knowledge from B, and check if similar structures of B are observable in the entire area A (not covered by B) so that the observation is extrapolated over the low-resolution data. I make a description of this in a little section of a paper.

The description is important to make probabilistic estimates. The aim of the paper I have in mind, deals with estimates of (1) where are the wanted structures, and (2) what shape do they have in the entire A.

Problem:

I cannot publish all of what has been done in B (yet), but just enough details (some are published already) to make the main structures wanted being characterized in A.

I want to publish the estimates done within A, but by using low-resolution data, which means, discriminating less characterizable structures; they can still be distinguished though.

Questions:

By doing (a) extrapolation of descriptions, and (b) general estimates on low-resolution data, should I fear about doing to much "speculations"?

I am more concerned about (a) than (b), because for the latter, the results are debatable and the method used is totally new and worthy for research in general.

1
  • 4
    I'm voting to close this question as off-topic because it's about research in a particular field, not about academia in general. Jul 12 '17 at 19:15
0

The legitimacy of such extrapolations is, as far as I know, among the topics studied by statisticians. If I'm right about that, it would be good to ask a statistician about your extrapolations, both to check how valid they are and also, if they are valid, to get some citable support for their validity.

Not the answer you're looking for? Browse other questions tagged or ask your own question.