# Follow up tests after pearson chi square

#### paulgooney

##### New Member
Hello,
I am looking for some advice on statistical analysis using SPSS. What I am trying to do is quite simple to something in the know of statistics but my knowledge of statistical analysis is quite limited.
Basically my data set is foaling records from a stud farm over a 10 year period with variables such as the weight of the foal, gender of foal etc
I have carried out multiple chi squared tests to see if there are any significant relationships and there are multiple thankfully.
But I am not sure what tests I need to carry put exactly after the pearson chi square.
Just for example 2 things I am looking at is the gender of the foal and the time it took the foal to stand up after birth. I have carried out a person chi square which has identified there is a significant relationship between the gender of the foal and the time it took to stand post partum
I have my time to stand parameters broke up into <2o minutes, 20-4-0 minutes and >40 minutes.
What test do I now need to run to identify exactly what the significant relationship is? as it does it take colts (males) or fillys (females) longer to stand after birth?
Any help is hugely appreciated, I understand this is quite basic level statistical analysis but I have very limited knowledge on this front

#### Miner

##### TS Contributor
If you still have the raw data (not binned), use a 2-sample t-test.

#### katxt

##### Active Member
Generally speaking, chi squared tests with continuous data put into bins (such as weight and time as you have tried) are a weak way to look for differences. You would be better off with a two sample style analysis - a t test if the data are suitable, or Mann-Whitney test if they aren't.

#### paulgooney

##### New Member
Hi guys, thanks for your input, have tried all mentioned thing above and still cannot make heads nor tails of it
This is what I get when I do a Mann-Whitney test

#### Attachments

• 131 KB Views: 3

#### katxt

##### Active Member
Try using the original times, not the coded times with the MW test, or the t test, and look for a p values less than 0.05. Toy will surely get one from the look of things.