MeridianSpark

← HomeAll topics

July 29, 2026

Phenomenon window

The Credential Gap: When Education Data Lies

You've heard that a college degree is the ticket to a good job. But what if the data we use to prove that is hiding something? Imagine you're looking at two maps: one shows urban areas with high college graduation rates, the other shows rural areas with lower rates. The obvious conclusion: city kids have more opportunities. But what if you zoom in and discover that within each city, certain neighborhoods have rates even lower than rural areas? That's Simpson's Paradox—a statistical trick that can flip your entire understanding. In this project, you'll dig into real survey data to see if the 'credential gap' between urban and rural areas is real, or if it's a mirage created by hidden biases. You'll use tools like cross-tabulations and chi-square tests to uncover the truth. Ready to see what's really going on?

  • Think about your own high school: did you notice any differences in how many students planned to go to college based on where they lived in town?

  • Have you ever been surprised by a statistic that seemed to tell one story, but your own experience suggested another?

  • Can you recall a time when a teacher or counselor assumed something about your future based on your background or where you're from?

SociologyEducationSocial ScienceCategorical InteractionarXiv · 2026ap_statistics/exploring_two_variableap_statistics/inference$0 public datasetsSupplies: laptop onlyStandard laptop (CPU only, 8GB+ RAM; Windows/macOS/Linux)Easy

Research gap

Replicate and extend Georgetown's proprietary findings using open-access ICPSR surveys, revealing hidden biases via Simpson's Paradox.

I have a story — start exploring

Unlock the full 8-week scaffold, personal story prompts, and admissions hooks.