70
several categories of child verbalizations, as well as Physical Positive and Laugh. One potential problem with this definition of prosocial is that the verbalization categories are not specific enough to reflect the content of the children's speech. For example, as long as the child does not whine, yell, or speak in a sarcastic tone of voice, the statements "I don't want to share the toys" and "I want to share my toys with you" would both be coded as Information Descriptions. In addition, the DPICS II system does not take into account turn-taking behavior or positive affect (e.g., smiling) which are important prosocial behaviors in preschoolers. Research using the Family Intake Coding System (FICS; Stormshak & Greenberg, 1996) found that comparison children displayed more positive affect than clinic-referred children while interacting with their parents (Stormshak, Speltz, DeKlyen, & Greenberg, 1997). Rather than using Information Descriptions broadly as part of the prosocial variable, it might be more useful to use behaviors consistent with "positive elicitation" (Stormshak & Greenberg, 1996). This category would include asking a parent to play or offering a toy. Furthermore, it might be worthwhile to supplement the quantitative Prosocial score with a qualitative rating of positive affect. In the FICS (Stormshak & Greenberg, 1996), parents and children are rated on an ordinal 5-point scale (0 = hostile/angry, 4 = positive affect, happiness).
The present study also replicated some, but not all of the previous findings with DPICS and DPICS II related to parent behavior. Fathers of clinic-referred children exhibited significantly more inappropriate behaviors (i.e., critical and sarcastic statements) than comparison fathers. Over the combined 30 minutes that the dyads were observed, the referred fathers issued approximately 29 critical/sarcastic statements compared to an