Using Fisher (1993) Statistical analysis of circular data (p30-35), I calculated the sample circular dispersion of some timestamp data. However, I got circular dispersion greater than $1,000$. Is this possible? Does a circular dispersion of, for example, $361$ equivalent to a circular dispersion of $1$? Or did I do some massive calculation error?

Just as a reference, I also followed this post. Thus, quoting Kees Mulder:

  1. Denote the data by $\boldsymbol\theta = \{\theta_1, \dots, \theta_n\}.$ An estimate of the mean $\bar\theta$ can be calculated with

  2. Calculate $\bar{R} = \frac{\sqrt{S^2 + C^2}}{n}$

  3. Calculate the dispersion as suggested by whuber. I'm not sure why, but Wikipedia's definition seems to differ slightly from Fisher. I will use Fisher's:

    • $\hat\delta = \frac{1 - \left[ (1/n) \sum_{i=1}^{n} \cos 2 (\theta_i - \hat\mu) \right]}{2\bar{R}^2}$
  4. Then, choose some constant $c$ (1 is probably fine, but you may fine-tune). Then, the interval is given by

    • $ \left[\hat\mu - c \hat\delta, \hat\mu + c \hat\delta \right]$

Also, unfortunately, I cannot post my data.

Given the questions and comments below, I just want to clarify what I am looking for. I am primarily interested in whether the circular dispersion has an upper bound and how to interpret large circular dispersions. I am actually fairly confident that my calculations are correct, as long as the notes that I posted are correct.

  • $\begingroup$ What is your mean resultant length and sample size, in this case? $\endgroup$ – Kees Mulder Nov 21 '16 at 19:15
  • 1
    $\begingroup$ Could you add a fuller reference/citation than "Fisher (1995)", as you would in a paper or dissertation? $\endgroup$ – Silverfish Nov 21 '16 at 19:52
  • 1
    $\begingroup$ To make this self-contained, please define circular dispersion and give a small dataset with puzzling result sufficient to replicate (or correct) what you find. $\endgroup$ – Nick Cox Nov 21 '16 at 19:54
  • 1
    $\begingroup$ If you want to know for sure whether you made a calculation error, I can't see an alternative to you posting your data! Or are you just interested in whether the circular dispersion has an upper bound? $\endgroup$ – Silverfish Nov 21 '16 at 19:58
  • 2
    $\begingroup$ You are making a calculation error. You don't have to reproduce your actual data: make a very small dataset with similar characteristics and see whether the problem recurs with it. Then you can post that dataset. $\endgroup$ – whuber Nov 22 '16 at 16:25

Your Answer

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

Browse other questions tagged or ask your own question.