1. Which keywords are associated with the highest-performing videos?

• Consider analyzing by views, likes, and comments.
• Are there common patterns in topics or content types?

2. Do videos with high engagement (likes/views/comments) tend to have more positive, negative, or neutral comments?

• Use average sentiment across comments per video.
• Do high-performing videos attract more extreme sentiment?

3. What is the relationship between comment volume and comment sentiment?

• Do videos with more comments have more polarized (0 or 2) reactions?
• Are neutral comments more common on low-engagement videos?

4. Which videos have the most liked comments, and what is the sentiment of those top comments?

• Are the top-liked comments generally positive or critical?
• Do they align with the video's popularity?

5. What proportion of videos have disabled comments or hidden like counts?

• Do these videos still perform well in terms of views?
• Is there a trend by keyword or publishing date?

6. Are there keywords that consistently result in higher comment sentiment or more liked comments?

• Can creators use this insight to drive more engagement?

7. How does the publication date relate to performance?

• Are newer videos trending better or worse than older ones?
• Does sentiment vary over time?

1. Which countries and regions have the highest and lowest prevalence of different mental illnesses over time?

• Are rates increasing or decreasing?
• Which disorders are most common?

2. How does the burden of disease (DALYs) compare to the prevalence of mental illness across countries?

• Do some illnesses cause disproportionate health loss?
• Are there regions with high burden but low prevalence reporting?

3. Where is the treatment gap for anxiety disorders the largest?

• Which countries have the highest percentage of untreated individuals?
• How do potentially adequate treatment rates compare across countries?

4. What percentage of the global adult population is covered in primary data collection for different mental illnesses?

• Are some disorders underrepresented in data?
• Are some regions more frequently studied than others?

5. Which depressive symptoms are most commonly reported in the U.S., and which ones are least prevalent?

• How do patterns shift by frequency (e.g., “nearly every day” vs. “several days”)?

6. Are there mismatches between disease burden (DALYs) and treatment coverage for depression and anxiety?

• Can you spot regions where the need is high but treatment remains low?

7. Which disorders are well-represented across multiple datasets—and which ones might be underreported or underresearched?

• Use `UNION` and filtering to check for consistency in coverage across data files.