Hi,
I have data in form of counts (number) and I wanted to run a statistical test to compare the counts between groups to see if there is statistically significant difference between groups. I was advised to use the Freeman Tukey test but I could not find any tool to run the test, only formulas.
Can anyone recommend any tool or online calculator if they exist? I am not expert in stats and I need something easy to use
Thanks
Francesca
Hi Francesca Panin, I hope you are doing well.
There are some alternatives like datatab.net, which allows you to analyze your data directly from your browser, and has a bunch of statistical tools ready to use (even you can make plots), however the main drawback is that it is not free. Other free alternatives (but limited in features) are statskingdom.com, wessa.net and stats.blue.
Depending on the specifics of your data, I could recommend you to use desktop software instead, some alternatives are JASP or JAMOVI, which are free, open source and easy to use. I personally use R, which is powerful and flexible, but it has a steep learning curve.
On the other hand, maybe a model-based approach can provide you more flexibility by letting you include the effect of confounders on your analysis in the form of random effects. For instance, you may specify a GLM (or GLMM if your data has some inherent structure or clustering) with a poisson (or negative binomial) distribution, and include your grouping variable as an independent variable, this way you could even model distributional parameters and between-groups characteristics, rather than just reporting a Chi square or an F statistic, primarily for reporting purposes.
In case of needing any further assistance, don't hesitate to reach out.
Greetings,
- Matías
Many thanks! I looked at the browser based ones, but I could not find the Freeman Tukey anywhere. I saw the tests for two proportions in the stats.blue, and I wondered if those tests would be suitable for what I need to analyse? I have a count to compare between two groups
Thank you
Francesca Panin
A few options:
1. R: . The `fisher.test()` function in the base R package can be used to perform Fisher's exact test, which is equivalent to the Freeman-Tukey test for 2x2 contingency tables.
2. Python: The `scipy.stats.fisher_exact()` function can be used for the Freeman-Tukey test on 2x2 contingency tables. For larger contingency tables, you can use the `scipy.stats.chisquare()` function.
3. SPSS: IBM SPSS Statistics is a widely used software package for statistical analysis. It offers a user-friendly interface for conducting various analyses, including tests on count data. The "Crosstabs" procedure in SPSS allows you to compare counts between groups and perform chi-square tests or Fisher's exact tests.
4. Online calculators: Websites like VassarStats (http://vassarstats.net/) and GraphPad Prism (https://www.graphpad.com/quickcalcs/) provide calculators for various statistical tests, including chi-square tests and Fisher's exact tests.
Good luck
- Badges
- Science topic