If you only have a single, two-category treatment variable, then a t-Test would be appropriate.

It also depends on how the data have been collected. If the same participants have been part of both conditions (treatment/no treatment), then you should use the paired-samples t-test. If you have two separate groups (a control group and an experimental group) you should choose the independent samples t-test.

You could also use point biserial correlation analysis with a continuous outcome variable and an independent dichotomous variable, but with a t-test in most software programs you will also get different measures for Cohen's d, which is a useful effect size.

https://datatab.net/tutorial/point-biserial-correlation

Hello dear researcher, I agree with Martin's opinion.

good luck with your research

Re the possibility of multiple outcomes that was mentioned, see the attached PDF for an interesting suggestion from Frank Harrell's well-known book on regression.

If you want to get a better overview how you could analyze RCT designs, I recommend to have a look at the blog posts by Solomon Kurz. He gives a general overview and compares in a simulation change scores vs. multilevel models and in a complete series he not only shows how to analyse OLS regression models, but also GLM models (e.g. for count data), ordinal models, beta regression for bounded data etc., so it is a very good overview. Additionally, he has R codes for all analyzes and he demonstrates frequentist as well as Bayesian approaches, where the latter may have some advantages for non-standard models.

A general introduction:

https://solomonkurz.netlify.app/blog/2022-06-13-just-use-multilevel-models-for-your-pre-post-rct-data/

And the first part of the complete series:

https://solomonkurz.netlify.app/blog/2023-04-12-boost-your-power-with-baseline-covariates/