Lack of quality data: Which contributing factor most accurately explains the failure?
Underrepresenting accent groups in training data is a data-quality gap that directly explains poor performance for those users.
The question
A speech-recognition feature suddenly performs poorly for users with regional accents that were rare in the original training data. A cross-functional review convenes engineering, data and product teams to understand why the incident arose. Which contributing factor most accurately explains the failure?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Lack of quality data, because the training set underrepresented the accent groups, leaving the model unable to reliably handle those inputs in production. ✓Correct because unrepresentative training data that underrepresents a population is a data-quality gap, directly explaining poor performance for those users.
- Model drift, because the statistical relationship between inputs and outputs shifted gradually after release until the model no longer matched current conditions.Plausible because drift causes post-launch degradation, but the failure stems from an original training-data gap, not a gradual shift in conditions over time.
- Insufficient testing, because the release validation lacked the coverage needed to catch defects that only appeared once the system met real production traffic.Plausible because thin testing can hide defects, but the underlying cause here is the data's demographic gap; better testing would only have revealed it sooner.
- Brittleness, because minor perturbations in the input signal produced disproportionate and unstable failures across the model's normal operating range of inputs.Plausible because brittleness is a real failure mode, but the issue is systematic underperformance for a whole population, not instability from small input perturbations.
The trap
Attributing a training-data representation gap to drift or testing failures, when the root cause is unrepresentative data. How to remember it
Underrepresenting accent groups in training data is a data-quality gap that directly explains poor performance for those users.
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 — freeMore Understanding How to Govern AI Development questions
- A description of the incident: Which set of contents best fulfills the objective of managing and documenting →
- Maintaining a post-market monitoring plan that documents: Which ongoing disclosure obligation is specifically →
- Concrete, measurable success criteria and constraints tied: To define the business context and use case →
- All 426 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.
