I have a data on the weekly percentage of the total quarter bookings. The data looks as follows (note: weekly percentages add up to 100 for each quarter) : enter image description here (not real data)

I used the Ljung-box test to see if the data resemble random data. The Ljung-box test give me a very high p-value (0.7), indicating that the data resembles complete random data.

However if I use the actual numbers of the weekly bookings (not percentage of total quarterly), I get a fairly low p-value in the Ljung-box test, indicating that the data is not random.

Question: Is this expected when using percentages of the total? If so, why?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.