thanks for the video. But If i try to use it for each row of the table, it gives me errors. I have a dataset of cars and I would like to know the Confidence interval for the median price. How can i do that?
I really appreciated this tutorial. Made it very simple. I have one question though, how can you calculate alpha by itself given a population size and a sample size.
Hi Ruth, excellent video as always! I have one question though. In your example, considering that your sample size is only '89', how accurate this calculation can be considering that your total population is 40k and you didn't inform anywhere your total population of 40k users. I mean, suppose your total population is only about 200 users, so your formula would need to give you a smaller interval because your sample size represents better the entire population. I thought the formulas you used would also receive the information about the total population as well. Did I miss something? Thank you!
Hi Daniel, The dataset I used I took it from Wikipedia, (age of actresses that got an Oscar) , so you are right, the sample size is too small. Didn’t I mention that in the video? Apologies if I didn’t ! /Ruth
The larger the sample the better the prediction. Even though it may sound counterintuitive, the size of the sample needed does not depend on the size of the population it comes from(I.e sampling 1000 from a population of 25000 subscribers would give you an equally good prediction as sampling 1000 from 2 million subscribers). The important think to consider is if the sample itself is large enough (power of sample is called I think) to produce a statistically significant result. And that’s why the population size is not mentioned
@@CurbalEN Hello Daniel, Ruth. Under my opinion, the total population is 89. No more can be considered inside the group. For this reason, I understand that the formula to use should be STDEV.P and not STDEV.S because no more population can be included. Of course, that is only my opinion... Cheers!
I tried to complete your survey, but it was not accepted. I received an error message, "we were unable to submit your survey. contact the person who sent you the survey." I would like to identify and isolate the outliers and examine what they might mean. I could not wrap my head around the DAX formulae for outliers. I calculated outliers and extreme outliers in the quartiles using Excel and created a label to identify each. I could work with that in Excel. Maybe not the best way. Can you address the issues using Quartiles vs. St.D. to work with outliers?
A million thank you's. Had a sqrt calc given to me to get conf level and had no idea how it worked. Brilliant explanation and use case. In excel, they use error bars (whiskers I think) to show conf level, any idea how to visualize the same thing without throwing upper and lower into a line chart? Model is Customer sat over time