Findings from several recent studies suggest that the gender, cultural background and age of observers may influence their perception and evaluation of nonverbal communication. There is ample evidence to warrant using raters of only one sex (preferably female) and only one cultural background when the rating task involves decoding nonverbal communication. An alternate strategy would be to include rater gender or ethnicity as an independent variable. This paper discusses the conditions under which characteristics of raters may influence the decoding of nonverbal messages and describes strategies for overcoming these problems. Since many studies utilize observers or raters to assess nonverbal communication, without taking into account possible systematic differences in their responses to the task, significant interactions between rater characteristics and the independent variables under investigation frequently may go undetected. (Author)