Computer Science
8 years ago

Misclassification errors reduce accuracy of statistical tests in contingency tables.

Classification errors in contingency tables analyzed with hierarchical log-linear models. Technical report No. 20
Evelyn Korn

Paper Summary

Related papers

Theme background image
Mathematics
3 years ago

Unlocking Relationships Among Variables: Log-Linear Models Reveal Hidden Connections

Missing Data
Multicollinearity
Multivariate Analysis
Theme background image
Computer Science
8 years ago

Misclassification errors reduce accuracy of statistical tests in contingency tables.

Bayesian Networks
Multicollinearity
Multi-Criteria Decision
1 view
Theme background image
Mathematics
8 years ago

New method reveals hidden patterns in categorical data for better decision-making.

Multicollinearity
Bayesian Networks
Sensory Analysis
1 view

Classification errors in contingency tables can affect the analysis using hierarchical log-linear models. These models help describe relationships in tables. Some models are not affected by misclassification, while others are. Tests between models may have reduced power due to misclassification. More data may be needed to compensate for this loss of power.