A generalized index of dissimilarity

Area of Science:

• Statistics
• Social Sciences

Background:

• The index of dissimilarity quantifies uneven distribution between two groups.
• Existing methods are limited in scope for complex data.

Purpose of the Study:

• To generalize the index of dissimilarity for multiple groups.
• To develop a significance test for two-way contingency tables.

Main Methods:

• Generalization of the index of dissimilarity for >2 groups.
• Development of a significance test for contingency table cells.
• Implementation in a computer program named DISSIM.

Main Results:

• The generalized index accurately reflects the movement needed for even distribution across multiple groups.
• The significance test allows for exploration of specific cells within contingency tables.
• The DISSIM program facilitates these complex statistical calculations.

Conclusions:

• The generalized index of dissimilarity offers a more versatile tool for analyzing distribution.
• The new significance test enhances the exploration of contingency table data.
• The DISSIM program provides a practical solution for applying these advanced statistical methods.