Skip to lesson
Exit
Mixed Methods & Quant Literacy1 / 2

1 min lesson

When signals conflict

Work through this: "Your survey says 80% of developers are satisfied with the agent, but telemetry shows a high revert rate on agent edits and interviews surface distrust. How do you reconcile this in your readout?"

Step 1 of 2

When signals conflictthe part that reveals seniority

Conflicting data is the normal case, not a failure. Resolve it explicitly by deciding which source is authoritative for which kind of claim, then say so out loud.

For a claim about…
What people actually do
Trust most
Analytics / experiment
Because
Behavior beats self-report every time.
For a claim about…
Why they do it
Trust most
Qual
Because
Only interviews recover the mechanism.
For a claim about…
How widespread it is
Trust most
Survey / analytics
Because
Scale is exactly their strength.
For a claim about…
Whether a change caused an effect
Trust most
Experiment
Because
Randomization is the only clean causal evidence.

Name the authority for each claim type rather than averaging conflicting sources into mush.

Calibrate your confidence

Tag every claim with how sure you are and why. “Strong: experiment plus matching telemetry.” “Suggestive: consistent across nine interviews, not yet sized.” “Speculative: one vivid quote.” Over-claiming from thin data is the fastest way to lose a technical team's trust and you rarely win it back.

Say it like this

“Interviews told me developers distrust silent agent edits. The survey showed it's a top-three concern for power users. Telemetry confirmed they revert more after large diffs. So we tested a preview and acceptance rose while reverts held flat. Four sources, one story and I'm confident because they converge.”