1

I have no experience with data analysis but a data analysis project was offered to me in the past by one of my math professors. It would have been a PCA project, where PCA means "Principle Component Analysis" -- equivalent to the "singular value decomposition" in linear algebra.

It was a biology data set.

My question is: would a biologist professor use this data set in some different way? If yes, then in what ways other than something linear algebraic?

What would / could a statistics professor do with this biology data set that would be different from what a mathematician and biologist does with it?

Thanks,

  • 3
    I really don't understand the question. Presumably all different kinds of people could do different things with the dataset, depending on what methods they know and what methods seem useful. It doesn't depend on what department their business card says they belong to. – ff524 May 13 '16 at 22:05
  • 2
    Biologists might to PCA and/or something else. Mathematicians also might do PCA and/or something else. You're basically asking, "What are all the ways in which a dataset might be analyzed?" – ff524 May 13 '16 at 22:16
  • 1
    I just have a raw idea that might be nonsense or unrelated but I will share it "just in case". Maybe a biologist will see the data from his point of view (application related to his field), while a mathematician see only numbers? In this case, both will approach the analysis of the data differently. I'm not familiar with this field, but I will assume that the mathematician will have more "out of the box" ideas on how to interpret the data (since he is not concerned with the application but rather trends or relationships between the numbers). – The Guy May 13 '16 at 22:21
  • 2
    The biologist might contribute "Duh, we've known that for decades" or "That relationship is obviously due to a confounding variable X that isn't measured", for example. Also see the anecdote in this video - watch through 8:33. (But presumably the mathematician in this example didn't invent PCA, and the biologist also knows how to do PCA, so it is equally valid to ask: what did the mathematician bring to the table?) – ff524 May 13 '16 at 22:53
  • 1
    I think you should rephrase this question (toward the "what value does a non-mathematical or computer science based analyst add to the overall analysis of data") which is basically answered by @ff524's comment and video. As an analyst, I can tell you that I have personally had to go to the field expert and see what is known, what was likely biased or an error, and ask about a bunch of other issues after my initial analysis. – JGreenwell May 15 '16 at 3:34
1

Yes. The number of ways to analyze data is very large. If the data is complex, it would be unlikely any two people would pick the same approach. Keep in mind that analysis methods are selected from the knowledge of the user to suit the goals of the user. Analysis methods are only partially selected based on the actual content of the data set.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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