0
$\begingroup$

I have been working on a time series-like data set from using calcium indicators (waveform data). A script I was given transforms the voltages (x values) into Z-scores. However, the mean and standard deviation used in calculating these scores are from a baseline period before the trial, which makes it look something like this:

$$ x_{trial} - \mu_{baseline} \over \sigma_{baseline} $$

I know that the "x - mu" represents the difference between voltages during the trial and mean voltages at baseline, but how would using a baseline standard deviation affect the interpretation? Similarly, how would using a mu value from the trial and sigma value from the baseline period be interpreted?

Thanks in advance!

$\endgroup$
0
$\begingroup$

This can be interpreted from the hypothesis testing point of view. The null hypothesis is to assume that your x follows the same normal distribution as your baseline (given $\mu$ and $\sigma$). The z-score tells how many standard deviations your observation deviates from the assumed mean. From it you can estimate the probability of observing such a great deviation if your null assumption was true. However, you may also use z-score just to scale the result and see if you observation is like baseline (near 0 or maybe in range [-2,2]) or exceptional.

It is hard to see what you could test if you assumed the null distribution with the observed mean but the baseline standard deviation.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.