Question

What statistical test should I run?

I ran multiple ECGs on three individuals over a series of 6 months. One individual had 8 tests, one individual had 7 tests, and one individual had 5 tests. Within each individual, variance between tests was very low. How can I compare variance between them?

I want to run an ANOVA but I am not sure if the conditions are met?

Condition 1: Independence of cases - cases are not normal, they come from the same individual.

Condition 2: Normality - the values are not in a normal bell curve, they are all very similar to each other because they are from the same individual

Condition 3: Homogeneity - I don't even know what this means

Lastly, for whatever test you recommend, can you please tell me how to interpret results?

Thank you!!!

Answer 1

First and foremost, what is the hypothesis you want to test? Do you want to check if the mean ECG result for the 3 patient is the same?

If yes, then ANOVA is the test you need to run.

Next about the assumptions you need to check

The assumption of independence: You cannot directly check test but it assumed that the experiment is randomised and any bias to the results are eliminated.

The assumption of homogeneity of variance :
This is to check that the samples have equal variance.

A quick check for this makes a boxplot to check for each of the samples and see how they placed and if you see there are a huge difference in their ranges.

Next is to use Bartlett’s test.
Bartlett’s test tests the null hypothesis that the group variances are equal against the alternative hypothesis that the group variances are not equal.

In case you reject the null hypothesis take the log of the data points and try the test again.

After that, you can proceed with the ANOVA.

Hope this helps.