I kept only the three reader letters that agreed with my article. What is that called, and how do I fix it?

Confirmation bias — I noticed only the evidence that fits what I already believed. The fix is to go looking for the letters that disagree.

Follow-up probe: "what would you expect to find if you were wrong?" That is the move that breaks the bias.

due 2026-09-21 09:54Z · stability 0.00 · difficulty 0.00 · 0 reps · 0 lapses

0 reviews

Never reviewed. It is waiting in the new-card stack.

Evidence is the tutor's one-line justification, written with the grade (review_logs.evidence). “Needed a hint three sessions in a row” is the signal that the card's notes or answer need work — the loop the authoring chat is for (EXAMPLE-LESSONS, “Authoring feedback”).