A private pilot with an instrument rating, preparing for the commercial checkride, has raised a pointed complaint about Sheppard Air's test-prep software: correct answers are being marked wrong on practice exams. The poster describes a pattern of 3-5 misgraded questions per 100-question practice test, and while that error rate might normally be a minor annoyance, it becomes consequential under the poster's flight school policy requiring three consecutive supervised practice scores of 90% or better before sitting the actual FAA knowledge test. In that context, a handful of erroneously marked-wrong answers is the difference between a passing 92% and a failing 87%, directly delaying a student's progression toward the checkride. The specific example cited—a weight-and-balance CG calculation where the poster selected the correct answer of 44 (having calculated 43.99, correctly rounded) but was still marked incorrect—suggests a genuine grading or rounding-logic bug in the software rather than pilot error, since the poster reports triple-checking the math.
For working pilots and flight instructors, this kind of issue matters more than it might initially seem. Test-prep platforms like Sheppard Air, Sportys, King Schools, and Gleim are foundational tools in ab initio and advanced certification training pipelines, and their credibility rests on accurate, dependable grading logic. Flight schools increasingly build institutional policies—like the "three consecutive 90s" gate described here—around the assumption that these platforms grade reliably. When a testing tool introduces false negatives, it doesn't just cost a student a few minutes of frustration; it can add days or weeks to a training timeline, increase costs for aircraft rental and instructor time, and erode trust in the assessment process at a moment when the industry is already grappling with training bottlenecks and instructor/aircraft availability shortages.
The specific nature of the error—a CG/weight-and-balance moment calculation—also touches on a category of ACS knowledge that FAA examiners and DPEs weight heavily during oral exams and practical tests. Weight and balance errors are a perennial source of real-world incidents and accidents, from overloaded GA aircraft to improperly loaded business and cargo operations. A test bank that mishandles rounding conventions on CG problems, even inadvertently, risks reinforcing confusion in students on a topic where precision has real safety consequences. If the software is truncating or using a stricter rounding threshold than standard aviation convention (i.e., not accepting 43.99 as equivalent to 44 within tolerance), that's a legitimate software defect worth flagging to the vendor, since ACS-based FAA questions typically build in reasonable rounding tolerances precisely to avoid this kind of ambiguity.
Broadly, this incident is a small but telling data point in a larger conversation about the reliability of third-party training and testing software as the aviation industry scales up pilot production to meet airline and business aviation demand. As flight schools lean more heavily on structured, checkpoint-based syllabi (Part 141 programs, ATP-style pipelines, and even Part 61 schools) to standardize outcomes and reduce checkride failures, they are also becoming more dependent on the correctness of automated grading tools. Any systemic bug—even a small percentage error rate—can ripple through a training organization's throughput, and it underscores why instructors and students should treat practice-test scores as directional rather than infallible, manually verify borderline or miscounted answers, and report suspected software defects to vendors promptly. For an industry already sensitive to pilot pipeline throughput and training costs, seemingly minor grading bugs in test-prep software carry outsized downstream effects.