general advice please

[Katie Martin]

Thank you for sharing this platform! I have been using null hypothesis significance testing throughout my PhD and now want to include estimation statistics in my thesis following feedback, but am quite unsure of the most correct way to do this. The behavioural study that I want to add estimation statistics for is longitudinal, with the same behavioural assays conducted for each mouse pup on consecutive days. There are 3 experimental groups in total, one control and two treatment groups comprising different numbers of pups (n = 18-24) and maternal litters (n = 3-4).

I can generate Hedges g values for the individual pup and litter average data using the "shared control" function, however, am unclear on whether analysing the data corresponding to each test day separately is the correct/best option available in this case? There is no repeated measures option that I am missing is there? If not, is it considered acceptable to report the individual values?

Thanks!

[Adam Claridge-Chang]

Hi Kate,

You're right, the app and code are missing an essential sixth function: the repeated measures shared control plot. Since this project lost its funding, we haven't been able to implement it. (Note: If you, a colleague, or anyone reading this can code this new function, we should be able to assist with a pull request in github.)

I'm sorry I can't be more helpful at this point. I will continue to look for ways to make this happen.

Best luck,

Adam

[Adam Claridge-Chang]

Hello,

In response to your question, I am recruiting a statistics intern to work on the repeated measures function over the summer.

We have a few fly datasets they can use to test the code, but it would be great to have a real-world mouse dataset as well. Would it be possible to share with me one of your data tables? It can/should be anonymized for genotype or any other identifying labels. Please email directly if sharing is possible.

BTW, I realize that people are currently just using the paired data function for >2 repeated measures, and simply using the baseline values for both plots (in the case of three measurements). It might be inelegant for more than three time points, but at least it gives a visualization with the correct information. Sorry I didn't think of this obvious stopgap before.

Many thanks in advance,

Adam