I have data showing which month a disease started for 119 subjects. The months are represented by 1-12 (Jan - December). We hypothesize the disease more often onsets during winter months. What test should I use?
This is the data we have.
Month Number of people with disease onset in this month
1 11
2 9
3 10
4 7
5 7
6 5
7 9
8 9
9 16
10 11
11 14
12 11
So far, I've attempted to use a Poisson test in SPSS and my interpretation of the results is that there's no significant difference from a random distribution (Omnibus test 0.155 significance).
However, I was hoping I could confirm with someone that this is the correct test and that I'm interpreting this accurately? Just eye-balling the data suggests there does seem to be less counts in the summer months, so I was a bit surprised there was no significance.
If there's a different test I should do in SPSS, please let me know!
Thank you.
You can try a K-means clustering by setting K=2 , initial cluster centers around x= 8 and x=12
use modulus function as the distance measure.
Plot your data and look for the peaks
If they're in the months you expect them come back with your question. In the meantime see the current issue of the journal, . SIGNIFICANCE, published by the Royal Statistical Society and the American Statistical Society jointly.. it contains on the cover Sir David Cox's picture.. Best wishes, David Booth
Since dates are circular – month 1 follows month 12 – you need statistical methods for circular data. If you hypothesise that there is a single peak in the data, then the Rayleigh test seems to be the preferred option. If you think that there are several peaks, then the Kuiper test is for you. As you can conclude from those two names, circular statistics are heavily indebted to astronomy.
I think this confirms your idea pretty much. I used sqrt(n) for the precision of the n values, as is usual. Done with FittingKVdm (www.lerenisplezant.be/fitting.htm).
. circsumm angle, ray
Circular statistics for angle
-----------------------------
Number of data 119
Mean direction, deg 307.86
Vector length 19.81
Vector strength 0.167
Circular range, deg 330.0
Rayleigh test: unimodal not uniform?
P-value 0.037
I had to convert the months to angles to use the Rayleigh test, thus :
gen angle=(month*30)-15
So I assume each observation occurs in the middle of the month.
The Rayleigh test is against a null of uniformity and an alternative of unimodality. .
Ronán Michael Conroy That is more or less the same as fitting a sine wave through the data (see above).
The original reference for the Rayleigh test is rather lovely:
- Rayleigh, L. (1919). On the problem of random vibrations, and of random flights in one, two, or three dimensions. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 37(220):321-347.
Don't you just love that!
- Badges
- Science topic
- Similar topics
- Correction