Can You Trust Online Course Ratings? What Evidence to Look For

When a course page says “4.9 stars” or shows a wall of glowing testimonials, it is tempting to treat that number as a verdict. For parents and students comparing paid online courses, ratings can be helpful—but only as one piece of evidence.

A rating tells you what some reviewers chose to report about their experience. It does not automatically tell you whether the course was right for learners with a similar starting point, whether most enrolled students finished it, how learning was assessed, or whether a claimed result was measured fairly. The most reliable approach is to read ratings as clues, then test those clues against clearer evidence.

The U.S. Federal Trade Commission has rules addressing deceptive practices involving consumer reviews and testimonials, including fake reviews, certain undisclosed insider reviews, and some forms of review suppression. That is an important protection, but it does not turn every real review into a complete evaluation of educational quality. ([ftc.gov](https://www.ftc.gov/system/files/ftc_gov/pdf/r311003consumerreviewstestimonialsfinalrulefrn.pdf))

Start with the right question

Instead of asking, “Is this course highly rated?”, ask: “What is the evidence that this course is a good fit for this learner, for this goal, at this point in time?”

A beginner who needs steady teacher feedback may have a very different experience from an independent learner who already knows the basics. A course can be genuinely valuable for one group and a poor match for another. That is why the quality of the evidence matters more than the size of the star graphic.

1. Look beyond the average rating

An average score hides variation. A 4.8-star average could represent a large number of consistent experiences, or a small number of highly enthusiastic early buyers. It could also reflect feedback gathered immediately after purchase, before learners have completed substantial work.

When reading reviews, look for details that are difficult to infer from a star rating alone:

• Who is reviewing? Look for clues about age, prior knowledge, learning goals, schedule, and whether the learner studied independently or with support.
• What did they actually do? More informative reviews mention assignments, explanations, feedback, practice, projects, or assessments.
• When was the review written? A review written after the first lesson answers a different question from one written after completion or after using the skill in a real setting.
• What did not work? Specific limitations—such as pace, unclear instructions, limited feedback, or advanced prerequisites—can help a family judge fit.
• Is there a pattern? Repeated comments from independent reviewers are more useful than one memorable testimonial.

Be especially cautious if every review uses broad claims such as “life-changing” or “the best course ever” but gives no information about the learning experience. Enthusiasm is not the same as evidence.

2. Ask how many learners the rating represents

A rating based on 12 reviews should not be read in the same way as one based on 1,200 reviews. More reviews can give a broader picture, but quantity alone is still not enough. You need to know which learners were invited to review and which learners responded.

Useful questions include:

• How many learners enrolled during the period represented?
• How many submitted a rating or survey response?
• Were reviews collected from all enrolled learners, only completers, or only people who opted in?
• Are the reviews from a recent cohort, or are they spread across several years and course versions?
• Were reviewers offered a discount, prize, referral benefit, or other incentive?

A provider does not need to publish every internal detail to be credible. But a provider making prominent quality claims should be able to explain, plainly, what its rating is based on. If it cannot, treat the rating as marketing information rather than a dependable measure of learning quality.

3. Compare the course with the learner’s baseline

“Baseline” means where learners were starting before the course: their prior knowledge, confidence, skills, available time, language needs, and access to support. It is essential because a course result has little meaning without knowing the starting point.

For example, a course may report that learners produced impressive final projects. That can be encouraging, but it does not show how much they learned unless you know whether they began as complete beginners, had previous training, or were selected because they were already ready to succeed.

Education evidence standards pay close attention to baseline comparability when evaluating studies that compare groups. The What Works Clearinghouse, an initiative of the U.S. Department of Education’s Institute of Education Sciences, includes baseline equivalence and attrition among the issues considered in assessing study designs. ([ies.ed.gov](https://ies.ed.gov/ncee/wwc/Docs/referenceresources/Final_WWC-HandbookVer5.0-0-508.pdf))

Parents and students do not need to conduct a formal research review. A practical version is to ask:

1. What knowledge or skills does the course expect before enrolment?
2. Is there a diagnostic quiz, sample lesson, or clearly stated prerequisite?
3. What does successful progress look like for a learner starting at this level?
4. Does the course offer different routes, pacing, or support for learners who need them?

A trustworthy course description makes the expected starting point clear. It does not imply that every learner will get the same result regardless of preparation or effort.

4. Treat completion rates carefully

Completion can be useful information, but only if the definition is clear. “Completed” might mean watching every video, submitting all assignments, passing an assessment, attending a final session, or simply reaching the end of a platform sequence. These are not interchangeable.

Ask the provider to define the measure before comparing it with another course. A meaningful completion figure should state:

• the cohort or date range;
• the number of learners who started;
• what counted as a completion;
• whether learners who withdrew, transferred, or stopped participating were included;
• whether the course was self-paced, teacher-led, mandatory, or optional; and
• what support learners received.

For formal postsecondary reporting, the National Center for Education Statistics notes that student-success measures use distinct survey components and cohorts. That is a useful reminder that completion measures depend on definitions and reporting rules; a headline percentage without its method is incomplete. ([nces.ed.gov](https://nces.ed.gov/ipeds/use-the-data/measuring-student-success-in-ipeds))

Do not assume that a lower completion rate always signals a poor course. A demanding, optional, self-paced course may lose learners for reasons unrelated to course quality. Equally, a high completion rate does not demonstrate deep learning unless learners had meaningful opportunities to practise and show what they could do.

5. Separate outcome claims from proof

Course marketing may promise better grades, faster progress, admissions success, a qualification, a job, or higher earnings. These claims deserve the highest level of scrutiny because they can influence major family decisions.

Ask four questions:

1. What exact outcome is being claimed? “Improved confidence” is different from “passed an exam,” and “completed a portfolio” is different from “got a job.”
2. Who was measured? Was the claim based on every enrollee, only finishers, only survey respondents, or a selected success-story group?
3. How and when was it measured? Look for the assessment, survey question, time period, and whether the provider verified the result independently.
4. What comparison was used? Did outcomes improve compared with learners’ own starting point, another learning option, or no comparison at all?

Strong evidence does not require a course provider to run a large experiment for every class. It does require honest boundaries. A careful provider can say what was measured, who took part, what the data cannot show, and why results may differ for future learners.

A testimonial can show that a positive experience happened. It cannot, by itself, show how common that experience is or prove that the course caused the result.

6. Watch for selection bias

Selection bias occurs when the people represented in the evidence are not representative of all the people who enrolled—or of the learners considering the course. This can happen without anyone intending to mislead.

For instance, learners who finish a difficult course may be more likely to leave a positive review than learners who stop early. Students who volunteer for a case study may have unusually strong results. A provider may naturally highlight its most successful projects. Each example can be real, yet the resulting picture can still be incomplete.

Look for signs that a provider is reducing this problem rather than hiding it:

• It distinguishes enrolment, participation, completion, and survey-response numbers.
• It includes constructive feedback or explains common fit issues.
• It presents outcomes for a defined cohort rather than only selected examples.
• It describes the learner profile and prerequisites.
• It avoids implying that a few success stories predict every learner’s future.

Research-quality standards also address attrition because missing participants can distort a comparison. That same principle is useful for families: ask who is absent from the evidence, not only who is featured in it. ([ies.ed.gov](https://ies.ed.gov/ncee/wwc/Docs/referenceresources/Final_WWC-HandbookVer5.0-0-508.pdf))

A practical evidence checklist before you buy

Make the final decision on fit, not hype

The best course is not necessarily the one with the most stars. It is the one whose teaching approach, level, workload, feedback, cost, and support match the learner’s needs—and whose claims can withstand reasonable questions.

Before enrolling, ask for a sample lesson, a syllabus, expected weekly workload, assessment details, and a straightforward explanation of any completion or outcome figures. If answers are clear, proportionate, and specific, that is a positive signal. If the provider relies on urgency, vague superlatives, or unsupported promises, slow down.

Teachers should remain central to meaningful educational decisions: they can interpret evidence, understand a learner’s context, and judge work that a rating cannot capture. If you are looking for a more transparent, teacher-led approach to learning support, explore SubSchool’s information for parents. SubSchool is designed to reduce repetitive teaching work while teachers retain authorship and the final educational decision.