Replica reposted this
Sharing this important post from my Replica colleague, Albab Noor. Understanding this misconception is crucial for any public sector leader making informed policy decisions using data we increasingly rely on to shape our built environment, including transportation policies like congestion pricing. Excited to be working with agencies around the U.S. who are embracing the power of these models to turn raw observations into actionable, meaningful insights.
Misconceptions about Transportation Data: There is a misconception among people about observed data vs. modeled data (some even call it synthetic data) in the transportation big data space. By the time they’re ready for use by an agency, big data products from all companies are MODELED data. There’s no such thing as observed data at that point. One may not call it modeled data, they may call it “expanded data” or “weighted data” or “scaled data”. All of these are different ways to describe the same thing. Remember, if you observe a sample (which you do, because you can’t observe every device) but report a number that tries to describe the population, then you just used a MODEL (even if you don’t realize it) to scale that observed value up to make inference about the population. That’s no longer observed data, it’s modeled/scaled/weighted/expanded data. #TRBAM #Data #Models ———————— Reposting with visibility set to public since people who wanted to share the post couldn’t.
