Horticultural stats idiot needs help!

#1
I've nearly finished my assignment and am just writing up the stats analysis. Its been a year since I last did any, so forgive me if this appears to be a stupid question. My hypothesis is that there will be a correlation between the length of time seeds are cold stratified, (independent variable, continuous ratio) and the germination rate, (dependant variable, discrete count). Which test should I be using? I have 6 pairs of data.

Thanks in advance for any help.
 
Last edited:

Karabiner

TS Contributor
#2
You can use Pearson-correlation or Spearman's rank correlation.
If you additionally want to test statistical significance, then with
such a small sample ususally one would recommend to test
Spearman's correlation.

With kind regards

Karabiner
 
#3
You can use Pearson-correlation or Spearman's rank correlation.
If you additionally want to test statistical significance, then with
such a small sample ususally one would recommend to test
Spearman's correlation.

With kind regards

Karabiner
Thank you so much for your help. I did look at Spearman's correlation; but I thought the data needed to be ordinal? A bit more info: I have 6 petri-dishes, each with 50 seeds in. one is control, the others are subjected to 2wks, 4wks, 8wks, 12wks and 16wks refrigeration before being germinated. What i'm trying to find out is the best length of time to treat the seeds to produce the highest germination. I will also have data showing the day by day rate of germination over a max 28 day period, but that's another matter.
 
#4
I need to go to work right now for a couple of hours, so I apologise in advance if it takes me a short while to get back to you. I assure you, your time and patience with me is very much appreciated.
 

Karabiner

TS Contributor
#5
I did look at Spearman's correlation; but I thought the data needed to be ordinal?
Yes. Usually, the statistic software will carry out the transformation of interval scaled
data into ranked data automatically.

I have 6 petri-dishes, each with 50 seeds in. one is control, the others are subjected to 2wks, 4wks, 8wks, 12wks and 16wks refrigeration before being germinated. What i'm trying to find out is the best length of time to treat the seeds to produce the highest Germination.
Ok, then this is another situation than I imagined. This could instead be a matter
for a Chi² testb of homogeneity. You have got 6 cells, each with a count between
0 and 50 germinated seeds, and what you expect under the null hypothesis is that
the total count is distributed evenly between cells. If that hypothesis is rejected,
then you can perform pairwise comparisons between cells, in a similar fashion.

With kind regards

Karabiner
 
#6
Ah. yes, that sounds more like it! I have been giving it some more thought, and I'm thinking that if i make a daily count of germination for each group then I could also glean much more from the experiment. in this case, there is a body of research regarding methods of analysis. It seems that measurement and analysis of germination is considered quite a problematic area, so at least i'm in good company.

Many thanks for your time and help with this. Your insight has been invaluable to me. I think I have more of a sense of direction now.