How to tell if my trend is real?

[kop442000]

Hi guys,

I'm working on a research project that looks at the amount of cosmic rays coming out of a particular region of the sky in certain time intervals.

There are 13 readings taken:

9,9,6,6,7,1,6,6,1,6,4,1,0

When this is plotted, it looks like a downward trend - meaning the source could be becoming less active - but it's difficult to say with such small statistics.

I though perhaps I should try to quantify the likelihood of getting such numbers if the source was essentially a constant source - ie the trend is just due to fluctuations.

Could someone please suggest how I might think about doing this?

Thanks in advance for any suggestions!

 

[Stephen Tashi]

Browsing the various named statistical tests on the web, the Spearman's Rank Correlation Coefficient is only relevant one I found.

Usually when people state real world problems involving probability they leave out a lot of information. In my opinion, if you want a good answer to a real world problem you should try to make a probability model of it that incorporates most of your knowledge. (For example, you probably know whether those numbers are discrete "counts" from a detector or continouous quantities that have been rounded off, whether there is a reasonable upper limit for the numbers, whether there are other associated known variables that might affect the counts such as the position of the detector and what kind of errors the detector might make.) Think about it as if you were going to write a computer simulation of the process, even if you don't actually do that.