Track subgroup outcomes over time: Which control most directly | AIGP
7-day money-back guarantee — full refund within 7 days of purchase if you've completed under 20% of the questions. See pricing →
Certifications Tools Flashcards Career Paths Exam Guides Blog Pricing For Teams About

Language

✓ EnglishDeutschEspañolFrançaisPortuguês
Check readiness — free →

Track subgroup outcomes over time: Which control most directly addresses the concern?

AIGP Understanding How to Govern AI Development Medium

Tracking subgroup outcomes over time directly tests whether the feedback loop is amplifying historical disparities.

The question

A school-district scheduling tool assigns intervention resources using historical referral patterns. Testing, oversight, supplier review, and risk documentation are complete. Over time, assignments generate new referrals that reinforce earlier disparities, yet monitoring tracks only overall assignment accuracy. Which control most directly addresses the concern?

Preparing for AIGP? Take the free 5-min readiness quiz →

  1. Ask teachers to approve every assignment.
    Teacher approval may add oversight, but without subgroup outcome evidence it may not reveal or correct systematic reinforcement patterns.
  2. Replace historical features with synthetic referrals.
    Synthetic data may preserve existing bias or introduce leakage, and replacing features does not itself demonstrate improved subgroup outcomes.
  3. Increase overall accuracy testing.
    Overall accuracy can remain strong while subgroup disparities worsen, so aggregate testing does not directly assess the feedback-loop concern.
  4. Track subgroup outcomes over time. ✓
    Subgroup outcome monitoring can reveal whether feedback-driven referrals amplify historical disparities that aggregate accuracy conceals.
The trap
For feedback loops, monitor evolving subgroup outcomes, not only initial model accuracy or aggregate performance.

How to remember it

Tracking subgroup outcomes over time directly tests whether the feedback loop is amplifying historical disparities.

How many of these would you get right?

One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.

Test your AIGP readiness — free

More Understanding How to Govern AI Development questions

Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development · Every answer, right and wrong, comes with its own explanation.