Confidence Model

CourseHub assigns a confidence level to each course’s ROI evaluation. This page documents the methodology.

Three Confidence Levels

High Confidence

• Platform-published completion rate
• Verified salary data from LinkedIn, Glassdoor, or Levels.fyi
• Course data updated within 90 days
• Multiple data sources confirm the metrics

Medium Confidence

• Estimated completion rate based on similar courses
• Verified salary data from a single source
• Course data updated within 180 days
• Some parameters estimated

Low Confidence

• Multiple estimated parameters
• Stale data (> 180 days old)
• Salary data inferred from category averages
• Course newly listed with no completion data

Confidence Score Calculation

confidence_score = 0
+ (1.0 if completion_rate_source == "platform" else 0.3)
+ (1.0 if salary_source == "verified" else 0.3)
+ (1.0 if data_age < 90 days else 0.5 if < 180 days else 0.2)

confidence_level = "High" if score >= 2.5 else "Medium" if score >= 1.5 else "Low"

How to Use Confidence Levels

• High confidence: ROI estimate is reliable; can be used for major investment decisions
• Medium confidence: ROI estimate is reasonable; verify with primary sources before large investment
• Low confidence: ROI estimate is rough; treat as a directional indicator only

Improving Confidence

If you have access to better data sources for a specific course, please contact us. We update confidence levels as new data becomes available.

See Also

• ROI Formula
• Data Sources