Wiggling into Correlation
18 hours ago
- #correlation
- #data analysis
- #statistics
- Jeff Kaufman analyzed contra dance attendance data in relation to surgical mask requirements.
- Initial analysis suggested mask-optional dances were more popular, with 18/20 pairs supporting this hypothesis.
- A different pairing method reduced support to 13/19, making the conclusion less clear.
- Various factors like location, holidays, and marketing affect dance attendance, complicating the analysis.
- Total variation in attendance was estimated, with masking requirements accounting for only 5% of the variance.
- The correlation between mask requirements and attendance was found to be low (0.21-0.29), not statistically significant.
- A larger sample size (over 80 dances) would be needed to confidently determine the effect of masking on attendance.