→ Granularity misalignment refers to a mismatch between the level of detail in the predictor variables and the event being predicted.

In this case, flat tires are likely discrete, infrequent events, while tire pressure is measured weekly. If the prediction model is trying to link a specific tire pressure value to a binary outcome (flat tire: yes/no), and the timing doesn't align precisely, the predictor variable (pressure) may not be granular enough to accurately associate with the event.

Why the other options are incorrect:

B: While outliers can exist, they are not the most immediate concern given the time-series nature of the data.

C: While domain expertise is helpful, it doesn't directly address the data structure issue.

D: Lagged observations can be engineered in modeling but aren’t the primary problem here.

Official References:

CompTIA DataX (DY0-001) Official Study Guide – Section 3.1 (Data Granularity):“Granularity misalignment occurs when the temporal or spatial resolution of features does not align with the prediction target.”

Data Science Process Guide, Section 2.3:“Predictive performance can suffer when temporal mismatch exists between observations and outcomes. Granularity issues must be resolved prior to modeling.”

—