Interpreting Statistics
Even in simple tables, with 4 cells, produced by the cross-classifications of two variables with only two categories, it is possible to misinterpret the quantitative pattern in the relationship when present. Researchers need to report more than just an overall chi-square statistic or simple percentage differences.
Common statistics that are used to portray the nature and strength of the relationships in addition to percentage differences are relative risks, and odds ratios. One should also note the difference between statistical significance and substantive significance. Are the numeric differences substantively important in the population? Statistical significance is determined by sample size and degrees of freedom in addition to the quantitative relationships.
|
|
|
|