Averages can help us understand some big-picture categories and trends in data.
Averages flatten and conceal.
🚨Nerd Alert! 🚨 Both are true. We can use summary statistics like this as long as we are aware of their limitations. Unfortunately, we use them in research and practice all the time as if they were the only truth. And while averages may capture useful information about a population, they can leave individuals unseen and underserved.
We can use these summary methods in health work, but we shouldn’t be dogmatic about them. Clinical guidelines might reflect averages. So might insurance reimbursement. So the set of available options you have as a patient is limited by this lens.
Here's another statistical approach that shapes care: Health researchers love to focus on a single variable— an entity like a disease or a symptom. This is a traditional way to think and analyze. Isolating variables is clean and easy. It can give power and clarity to findings.
But it’s not enough. That’s not how people work, and if you’ve ever engaged with the healthcare system, you’ve probably felt this tension: You aren’t just PCOS. You aren’t just an A1C of 6.2. You aren’t just a migraine. But you’ll likely be offered treatment as if you were.
We have powerful ways to find patterns and clusters and trends that go way beyond basic measures of central tendency or inferential statistics. We can use SEM tools like principal component analysis. We can use person-centered analysis approaches like latent profile analysis. We can use machine learning to find complex profiles we didn’t know we were missing. All of this is rapidly increasing in power, ease, and accessibility.
Let's put these tools to work in service of meeting the needs of complex human being. I don't know anyone who's "average", do you?
#statistics #AI #machinelearning #healthcare
Experienced in Data Analytics, Quality Improvement, Operations and Finance
1moI am so honored to work at InforMD and be a part of CBCs data journey. CBC is a dedicated and innovative behavioral health IPA using data insights to leverage CBOs that provide essential services (behavioral health, primary care, care coordination, housing, and more). Please take the time to read about the amazing work being done by CBC and how InforMD helped create a data driven culture using the BI platforms that fit their need. #Snowflake #Tableau #AmazonS3