Forum Discussion
Helper IDifficulty in comparing same measure but different values
Helper IHello again,
There you go - I kept only 2 courses for the sake of brevity, and added some data around the same dates for 2021 and 2022.
- lbendlin
Super User
Please help me understand what I am looking at
for course 1 and 2021 we ended up with 41 enrollments and 6.9% . What is 6.9 the percentage of? What value represents 100 % ? The amount of slots overall for the training?
Are you saying you want to normalize the 6.9% of 2021 against the 4.37% of 2022? A weighted comparison?
- RenanPNog
Hello,
Here are the answers, I hope I was able to clarify them!
"for course 1 and 2021 we ended up with 41 enrollments and 6.9% . What is 6.9 the percentage of? What value represents 100 % ? The amount of slots overall for the training?"
So the percentage is the evolution of the course, and it's related to the date; and each 3 days is around 1% from the total. For ex, here we see that the Course 1 started at 3%, which means the campaign started really at 08/01/21 (or earlier):While for this year, the campaign started even earlier for the same course (around the end of July):
So, considering that each campaign lasts around 6 months, we can say the 100% ends at 02/01/2023 ( in this example, usually campaigns happen within a specific year.
As for the slots, there are no limits for enrollments, so the percentage is really the evolution of that course in a given campaign over time.
Are you saying you want to normalize the 6.9% of 2021 against the 4.37% of 2022? A weighted comparison?
Context: Our baseline for metrics calculation is the previous campaign, so we always look at the past in order try to measure the current campaign; since the courses have different duration times (it's possible to see that with the dummy), the way we found in order to compare is through the percentage.
Our goal is to compare how we are today (percentage of current campaign (2022)) to with how we were in the last campaign ( percentage of previous campaign (2021)), but since they're different we need to find the most approximated value - so with that we can see the running totals we had until that point in time (2021) and then we'll have an idea if this year's performance is good or bad, and take appropriate action accordingly.
I'm not sure if this would be a weighted comparison, it's more like a approximated value comparison to get another value, that change with time.Let me know if you have any further questions - I reckon this is a bit complex!!
Thanks in advance for your support!- lbendlin
It would be much easier if the course start and end dates were available for each campaign.
