were conducted with Bonferroni correction to control the group-wise error rate. The

mean score difference between property damage and verbal threats (M= 6.23, SD = .79)

and both the physical harm items (M= 5.51, SD = .87) and the murder/rape/torture items

(M= 5.62, SD = .79) was significant < .001. The difference between the physical harm

and murder/rape/torture items was not significant.

 Crime scenarios were also balanced with regard to type of bias crime symbols at

the crime scene. Each protected category includes one item where the offender is a

member of a hate group, one scenario where graffiti or bias symbols are present at the

crime scene and three scenarios where bias language is used verbally or in writing. There

is one exception to this rule. The religion category has two crime scenarios where bias

symbols are present and no indicators where the perpetrators were members of a hate

group.Table 4.6: One-way Analysis of Variance for Mean MHCS Sub-scale Scores by

Severity of Crime



Source df F rlp2 p


 Between subjects

Severity 2 40.97 .13 .001*

S within-group 547 (.67)
 error

Note: Values enclosed in parentheses represent mean square errors rp,2 reported is a partial eta squared. An
asterisk indicates significant results p <.01