Causal inference for people analytics — A/B, diff-in-diff, and the questions correlation can't answer
Most 'people analytics' is correlation in a dashboard. The questions leaders actually want answered — does our new manager program reduce attrition?
- Correlation ≠ causation, but you can earn causation with the right design.
- A/B tests are gold-standard but rarely possible in HR (ethics, scale).
- Difference-in-differences is the workhorse: compare change in treated vs control over time.
- Regression discontinuity exploits sharp cutoffs (promotion thresholds, tenure bands).
- Instrumental variables handle confounding when randomisation isn't possible.
Boards no longer accept 'engaged teams have lower attrition' as analysis. They want to know: does our intervention cause the outcome? That's a causal question, and causal questions need causal methods.
Why this matters
Every people initiative competes for budget against revenue-generating projects. 'We did this and the number moved' is not enough — the number might have moved anyway. Causal methods are how you defend the spend.
Four methods, plain English
| Method | When to use | Example |
|---|---|---|
| A/B test (RCT) | You can randomise who gets the treatment | Two onboarding flows assigned at random |
| Difference-in-differences | Treatment rolls out to some groups, not others | New manager program in 2 BUs; compare attrition trends |
| Regression discontinuity | There's a sharp cutoff in eligibility | Promotion bonuses above a calibration score |
| Instrumental variables | Non-random assignment + a clean 'instrument' | Office reopening timing as instrument for in-person days |
Building the capability
- Hire or partner with one analyst with statistics training (not just SQL).
- Pick 2-3 high-stakes questions per year that deserve real causal analysis.
- Pre-register: write down the hypothesis and method before you look at the data.
- Publish internally with limitations explicit — credibility compounds.
- Connect to the evidence-based management loop.
Running 12 correlations, picking the one that confirms your narrative, and presenting it as insight. That's not analytics — it's confirmation bias with a chart.
Read next
All playbooksMost HR practice runs on tradition, vendor pitches, and case studies of n=1. Denise Rousseau's evidence-based management (EBMgt) framework asks for something…
How to build a People Analytics function from scratch — the maturity model, first hires, tech stack, data foundations, the four use cases that buy credibility…
A short, defensible dashboard for leaders. Skip the vanity metrics and track what predicts hire quality and retention.