Have a look at the three graphs presented in exam question 6 on this paper (just the graphs):

enter image description here

I want to use that as an example for the catastrophe that I am stuck in as a high school student who hasn't delved much into data analysis:

How am I supposed to deduce a correlation?

The framework that I am using right now is to look for an overall change between two points and then say that the graphs are correlated within that interval. But then again, at first site, I wouldn't be able to tell that temperature is not correlated with NPP. I mean, there are fluctuations, but usually when reading graphs we tend to ignore the fluctuations. I have no idea why fluctuations are usually ignored, unless it's a physics experiment where they just want to throw it on experimental error.

The case is even worst with a question like the third one, on the same paper. There are great fluctuations and, while not having been taught the specifics of graph analysis, I find it difficult to know what should be ignored and what shouldn't be, when I read graphs.

enter image description here

  • 1
    $\begingroup$ That is a quite long paper! Can you cut out the graph(s) your question is about and include it in your post? We prefer that questions be self-contained, it should not be necessary to follow links to understand a question. $\endgroup$ Apr 20, 2017 at 18:55
  • $\begingroup$ I don't know how to. $\endgroup$ Apr 21, 2017 at 10:17
  • $\begingroup$ Under your post there is an edit tag. Tap that, put the cursor where the graphics file should be, above there is s small menu, use the include pictire icon. Work about as in facebook $\endgroup$ Apr 21, 2017 at 11:02
  • $\begingroup$ I added the images -- I think there is a minimum rep needed that prevents new members from adding images. $\endgroup$
    – xan
    Apr 21, 2017 at 13:25
  • $\begingroup$ I guess you've overthinking this. For those unfamiliar with the set-up here, I take it that these are British A level exams, taken in secondary school at age 17 or 18 or so. What I am seeing in the first few graphs is a seasonal pattern of sunlight and temperature (as forcing factors) and NPP (as response). I expect you're just asked to comment on whether the peaks coincide (they don't quite and there is biology there too). Measurement error is not germane to the question, I suspect. There are small measurement issues of different kinds, but the response to seasonal rhythms seems central. $\endgroup$
    – Nick Cox
    Apr 21, 2017 at 13:55


Your Answer

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