#statistics
12 articles
- The Telephone Survey Is Dying: Reading the Task Force Report on What Replaces It
Cell-only households went from 3% to 57% in fifteen years. The profession's own task force says the replacement is unsettled, and is honest about it.
- Normal Is Not the Same as Precise: The Histogram I Misread for Three Years
Simulating 1,000 samples finally made the central limit theorem click — and then showed me I'd spent three years misreading my own histogram.
- Shared Vulnerability: The Model, the Audit & the Studies They Both Point At
Carson cites Andreasen and Jamison as warrant. Schlesinger goes back to those originals. One number sits in both papers, meaning different things.
- The Opposing Processes Model Explains Why Competition Helps and Hurts at Once
I wrote that people thrive under competition, then scored full marks on a quiz saying the opposite. Same evening, same student, same model.
- Mental Health in Tech: The Stakes I Named, and the Sample That Couldn't Carry Them
Before touching the data I named stigma, geography, and legal stakes. The country breakdown then showed which of the three this sample could not carry.
- DORA Quick Check - A Critique
A benchmark that scores your team's delivery performance in under a minute. The speed is the product — and why two developers score it differently.
- Lost in Conversion
A number is only as honest as its unit, currency, scale, and sample. I'd converted units for a decade — and still shipped the mistake in my own paper.
- No Root Cause: What Coincidence Analysis Taught Me About Organizational Interventions
Coincidence Analysis taught me that organizational change may not have one root cause — just different sufficient recipes for the same outcome.
- Religiosity vs. GDP: A Regression That Passed Every Check — and a Causal Story It Can't Tell
A clean regression linked religiosity and GDP. That statistical cleanliness is exactly why the causal story needs more caution.
- MGNREGA: The Regression That Held — and the State Where the Budget Line Goes to Zero
A regression linked MGNREGA labor budgets to women’s paid workdays. Then West Bengal’s frozen funds tested what the chart could honestly say.
- Catching False Passes in a QA Selection Plan
I designed a QA selection funnel with validity coefficients and cutoff scores, then found the statistical error and bias I had built in.
- VulnerABLE Pan-European Survey - A Critique
A European survey set out to hear from isolated and vulnerable people. Its own method may have quietly changed what those people felt safe saying.