This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
It's the most wonderful time of the year - Santalytics 2020 is here! This year, Santa's workshop needs the help of the Alteryx Community to help get back on track, so head over to the Group Hub for all the info to get started!
I'm interested in exploring the relationship of a teams performance against an entire set of data that is available with each record. This is really an open ended question as to how you prefer to examine relationships in perofrmance records and what associations can be made. What are your go-to tools for this? What about in a time series as well?
We have a method of scoring a record for how well it performed along with a subset of data points associated with that record. I can make reasonable assumptions as to what would likely cause increases/decreases in performance, but want to take a data pure stance to this and remove any biases.
Questions we're after:
What are strong correlations to increased/decreased performance
What has caused performance to increase/decrease over time (assuming performance always is static with model match data points). Essentially what has changed outside of our man performance column that may attribute to increase/decreases in the performance measure.
What is your method of visually showing this information to business users?
Let me know if this is too open ended and I can provide some dummy data in a packaged workbook to toy with.
Example data is always helpful. You can easily explore the relationships between variables with the Data Investigation tools. I love Association Analysis to explore the relationship between variables, and the resulting scatterplot view can help you avoid multicolinearity.
Along those lines, to correlate data points, you also have the Scatterplot tool, the Pearson correlation, and the Spearman correlation, also in Data Investigation.
Then perhaps you'll want to perform a Linear Regression, or another appropriate regression, in the Predictive category. If you want Time Series analysis, that's it's own set of tools in the Predictive Suite.
@EstherB47 - Thanks for this! I did run the association analysis and agree its going to be useful! My first run had too many variables that I don't care about and wouldn't matter. After only selecting ones that could be of meaning its making much more sense and will help guide further areas of focus. Even found one variable I wouldn't have assumed a strong association at the top of the list. Going to explore the other recommendations now. I do have R installed.
If I get stuck with anything along the way here, I'll follow up with a example workbook of dummy data.