Xbar & R/S Charts assume normality. We can use the Central Limit Theorem to approximate normality for non-normal data to use these Xbar Charts.

The P Chart is for a binomial distribution.

The C Chart is a count of errors based on a Poisson distribution.

Hi Robert,

I have a question regarding the P-chart. I created a P-chart for one binary variable. Attached is the screenshot of that; as you see, I am interested in monitoring improvement quarterly, but as I have a cohort for two years, there are only 8 data points/ 8 quarters; I was just wondering if this plot is acceptable?

Hi Robert,

I've created a p-chart to monitor a binary variable on a quarterly basis. But I noticed a p-chart requires at least 20 data points. So, I've shifted to monthly reporting.

with this adjustment I have reduction in the number of obs per subgroup when moving from quarterly to monthly charts. Could you please see the attached file and advise if the sample sizes per subgroup are adequate for generating a reliable p-chart?

what is the minimum required number of observations per subgroup to effectively utilize a p-chart?

Thank you!