Community Forum
The new layout is in beta testing and we're inviting you to help us try it out! Click here to read the announcement post for details.
Data Collection (Local Comp Testing)
-
Cypress Creek Elites
- Premium
- Posts: 696
- Joined: Mon Mar 08, 2021 5:07 pm
- Location: Oregon
- Visit My Farm
Data Collection (Local Comp Testing)
I recently started collecting Level 5 local comp data for newborns for my own personal project, and I'm leaving this here in case anyone else has interest in gathering data for their own projects.
How I did this:
I set up a form that's connected to a sheet.
The form has one question, which is a short answer question.
The sheet has several equations inputted to analyze the results of the form.
The sheet is linked here, and if anyone wants to make a copy of the sheet, wipe columns A and B (one has date/time data, one has scores) and link that to their own form, go ahead.
So what do the numbers in row C mean?
C1 is the Mean, or Average, of all of the results from column B. Average is a great data point when the entirety of the data set is fairly regular and even.
C2 is the Median, which is essentially the number in the very middle of the data set. It's actually a great way to check if the Mean is being impacted by outliers, because the Median doesn't get impacted by them- so if you've got a drastically different Mean and Median, you've got a data point somewhere that's screwing with your averages.
C3-C5 are the Interquartile Range.
C3 is Quarter 1- the data value exactly 25% of the way into the data set.
C4 is Quarter 3- the data value exactly 75% of the way into the data set.
These two points help control for outliers as well, because they're cutting off the very high and very low end of your inputs- or in this case, you can judge approximately where the scores of a given horse fall in terms of the data you already have input.
C5 is the Interquartile Range, or the difference between Quarter 1 and Quarter 3. Again, you're ignoring outliers here, so you're going to get a more accurate read on where the range of scores should fall than with a normal range function, because a singular really low or really high score won't impact the data set.
Feel free to ask questions!
How I did this:
I set up a form that's connected to a sheet.
The form has one question, which is a short answer question.
The sheet has several equations inputted to analyze the results of the form.
The sheet is linked here, and if anyone wants to make a copy of the sheet, wipe columns A and B (one has date/time data, one has scores) and link that to their own form, go ahead.
So what do the numbers in row C mean?
C1 is the Mean, or Average, of all of the results from column B. Average is a great data point when the entirety of the data set is fairly regular and even.
C2 is the Median, which is essentially the number in the very middle of the data set. It's actually a great way to check if the Mean is being impacted by outliers, because the Median doesn't get impacted by them- so if you've got a drastically different Mean and Median, you've got a data point somewhere that's screwing with your averages.
C3-C5 are the Interquartile Range.
C3 is Quarter 1- the data value exactly 25% of the way into the data set.
C4 is Quarter 3- the data value exactly 75% of the way into the data set.
These two points help control for outliers as well, because they're cutting off the very high and very low end of your inputs- or in this case, you can judge approximately where the scores of a given horse fall in terms of the data you already have input.
C5 is the Interquartile Range, or the difference between Quarter 1 and Quarter 3. Again, you're ignoring outliers here, so you're going to get a more accurate read on where the range of scores should fall than with a normal range function, because a singular really low or really high score won't impact the data set.
Feel free to ask questions!
