Log-linear analysis

Loglinear analysis is a technique used in statistics to examine the relationship between more than 2 categorical variables. This technique examines if the categorical variables interact. (A Pearson's chi-square test could be used instead, but this technique only allows for 2 of the variables to be compared at a time.)^[1]

It uses a likelihood ratio statistic^[2] , χ^2: χ^2=2∑(O_ij×ln O_ij/Ei_j ) where ln is the natural logarithm, O_ij is the observed frequency in cell_ij (i = row and j= column), and E_ij is the expected frequency in cell_ij.

References

^ Howell, D. C. (2009). Statistical methods for psychology (7th ed.). Belmot, CA: Cengage Learning. pp. 630–655.
^ Field, A. (2005). Discovering statistics using SPSS (2nd ed.). Thousand Oaks, CA: Sage Publications.

[1] Howell, D. C. (2009). Statistical methods for psychology (7th ed.). Belmot, CA: Cengage Learning. pp. 630–655.

[2] Field, A. (2005). Discovering statistics using SPSS (2nd ed.). Thousand Oaks, CA: Sage Publications.

[1]

[2]