I would include also untreated controls. How many replicates in each treatment?

Do you already have an idea about variability within a muscle, animal or between animals?

If not, I would start there; otherwise you might have to conclude that natural variability is too large to draw any meaningfull conclusion from your experiment.

A friend suggested that I do this instead. Yes of course! untreated controls will be tested too.

-marinate steak chunks in crushed kiwi for 24 hours, 1 hour and 15 minutes.

-measure firmness with a texture analyzer or by applying weight (e.g. 500 g) on the steak chunks and measuring the length with a ruler.

-cook marinated samples in a pan (same temperature and time)

-measure firmness with the weight method or with texture analyzer

I only have one day to conduct experiments in the lab and extract as many findings/information that I can..would it be useful to look at the structure of the meat under the microscope as well?

Quite an ambitious plan! Microscopy does require a lot of time and skills! You might consider dropping one time point and increase the number of replicates per time point to increase your chances for obtaining a relevant/significant result. Good luck!

Thank you! I will stick to the weight method to measure the firmness (see how far it compresses the meat). Any advice on which meat cut to choose?

Choose a large preferably homogenous muscle. That would hopefully decrease variability and increase your chances to observe a treatment effect.

Thanks so much, appreciate it. Would it be useful to measure the mass of the meat before and after treatment?

Both dimensions and weight might be usefull, but do not expect too much effect on these parameters before cooking, but cooking will probably have a pronounced effect in all groups on dimensions and probably to a lesser extent on weight.