The major purpose of the current investigation was to examine empirically whether the use of negative terminology can bias perceptions and judgments of adult-nonadult sexual interactions. It was hypothesized that negative terminology would lead to more negative impressions of these interactions than neutral terminology would. 

Results of the experiment conducted to test this hypothesis showed that students' impressions of both specific and general cases of sexual contacts between male adolescents and adult males were negatively biased by negative terms, providing support for the hypothesis. 

Biasing effects, however, were weak and inconsistent in the specific cases. In contrast to this, biasing effects of negative terms on students' impressions of general cases were stronger and quite reliable. It is significant that students' impressions of general cases of adult-adolescent sexual relation- ships were especially affected by negative terminology. 

This finding indicates that, by describing specific cases of adult-nonadult sexual relationships with negative terms, researchers can give their readers the impression that general cases of these relationships are abusive even when the evidence in the specific cases points to neutral or even positive outcomes, as in the shortened article that students read in the current study. 

Thus, researchers who firmly hold that such relationships are abusive regardless of the empirical evidence (e.g., Finkelhor, 1979b; Maltz, 1989; Russell, 1986), and who consequently use negative terms indiscriminately, may be creating and maintaining in their readership a biased perception of adult-nonadult sexual relationships. 

It is also important to note that the biasing effects of negative terminology in this study seemed to occur without the awareness of the students. Students considered the terminology employed in the shortened articles to be appropriate, whether the terminology was neutral or negative. Thus, students perceived as equally appropriate terms such as "sexual relationship" or "sexual assault," "men" or "child molesters," and "boys" or "victims." 

When the language employed by the media and by most professionals

[Page 267]

is considered, the finding that the students in this study saw nothing unusual in reading a professional journal article replete with negative terms is not surprising. On the other hand, students did not find neutral terms any less appropriate. This result was most likely the case because terms such as "sexual relationship" are compatible with negative, neutral, or positive effects, and as such are not inappropriate regardless of the readers' expectations. Importantly, despite the fact that students perceived no difference in the appropriateness of the terms used, their evaluations and perceptions were nevertheless affected by the type of terminology to which they were exposed.

The results of this investigation provide some initial empirical justification for the concerns recently expressed by authors such as Kilpatrick (1987), Okami (1990), and Nelson (1989) regarding the use of negative terminology in describing all adult-nonadult sexual encounters. The finding that the mere choice of terminology used to describe such contacts can lead to biased judgments and perceptions suggests that a central goal of science - to communicate evidence in an unbiased manner - is not being served when negative terms are used indiscriminately.

The possibility that terms such as "participant" and "relationship" - as opposed to "victim" and "exploitation" - are themselves biased, carrying positive rather than neutral connotations, must also be considered. Although this is a valid concern, the results of the current study suggest that these terms do not necessarily carry positive connotations. Students in this study considered these terms to be appropriate despite their negative views of these sexual contacts - their negative views are indicated by their consistently negative mean evaluations provided in Table 1. If students in the neutral terminology conditions had considered the terms to be positively biased, it is likely that they would not have judged the terms to be appropriate, given their negative views.

A secondary hypothesis of the current investigation was that nonnegative outcome information would be processed in a biased manner. Two measures were critical for assessing biased processing: the long-term effects and the therapy recommended in the specific cases. 

Analysis of the first measure provided mixed support. No bias was exhibited by students in the neutral terminology condition who were exposed to the nonnegative outcome information. Their conclusions concerning long-term effects were consistent with the outcome information. 

However, students in the negative terminology condition who were exposed to this information did exhibit biased processing. Their negative assessments of the long-term effects were inconsistent with the outcome information. 

Thus, for this measure, pre-existing negative attitudes, beliefs, and expectations alone did not lead to biased processing; the addition of negative terms was necessary for this to occur. The negative terms may have had a priming effect (cf. Herr, 1986; Higgins, Rholes, & Jones, 1977), prompting students to interpret the long-term outcome information in a biased fashion. 

Analysis of the second measure provided clear evidence for biased processing. Students' perceptions about the amount of therapy that the adolescents described in the case studies should have received had their sexual relationships been discovered were not consistent with the outcome information. Students who read the nonnegative outcome information recommended extensive therapy, despite being exposed to information that implied therapy was not essential. They recommended as much therapy as did students who were not exposed to the outcome information. 

This bias may be the result of a cultural belief that associates disapproved sexual behavior with sickness (Bullough & Bullough. 1977), resulting in the perception of a need for treatment, regardless of the observed effects of the behavior. The finding of biased processing is consistent with previous research demonstrating the biased processing of information that contradicts strongly held attitudes and beliefs (e.g., Lord et al., 1979).

The current investigation was conducted using a sample of undergraduate college students who read case studies of male adolescents sexually involved with male adults. Several questions of generalizability of the findings in this study thus need to be addressed. 

One question concerns the representativeness of the sample of students who participated. Research has shown that volunteers for studies involving sex tend be more liberal and unconventional than non-volunteers (Rosnow & Rosenthal, 1976). Thus, it is possible that a group of non-volunteer students might have responded differently.
Another issue concerns whether professionals or people from nonacademic populations would have responded similarly. These are empirical questions. Importantly, the sample examined in the current study came from the same population (i.e.. college students) that is likely to be exposed to these articles at some point (e.g., as secondary sources in college textbooks). Therefore, the impact of an author's choice of terms is an important matter in a practical sense.

Another question of generalizability concerns what the effects of the experimental manipulations would have been had the boys been preadolescents, the younger partners been girls; or the older partners been women. Researchers examining attributions of responsibility to younger persons for their sexual encounters with adults have found that adolescents are seen as more responsible than preadolescents (Waterman & Foss-Goodman, 1984) and that boys are seen as more

[Page 268]

responsible than girls (Broussard & Wagner, 1988). Thus, because Tindall's (1978) cases involved male adolescents, students in the current study may have attributed more responsibility to the younger persons than they would have for other types of cases. These considerations suggest that students' evaluations might have been more negative for younger boys or for girls, irrespective of the experimental manipulations. The impact of these age and gender differences is unclear. Further research is needed to examine the generalizability of the current findings to other cases.

A final question of generalizability concerns the sample of adolescents described in Tindall's (1978) report. These adolescents were apparently consenting and successful in terms of adjusting to their sexual experiences. These cases fell on the positive end of Nelson's (1989) continuum of experiences. Additional research might address biasing effects when cases are used that fall along other parts of the continuum.

The current investigation added empirical data to the recent debate among a number of professional sex researchers concerning whether the use of negative terms to describe all adult-nonadult sexual interactions has biasing effects. Our experiment conducted to address this issue demonstrated that the use of negative terms can negatively bias readers' impressions of these interactions.

This negative biasing effect was particularly evident for inferences about general cases of sexual interactions even though students were exposed to information involving only a few specific cases. The finding of a biasing effect indicates that an author's choice of terms can affect readers' impressions irrespective of the actual empirical findings of the author's study.

We also examined the issue of biased processing of nonnegative information regarding the outcome of these interactions. We found, based on an examination of two critical measures, that students did exhibit biased processing, although this effect was moderated in the case of one of the measures by the type of terminology to which students were exposed. This finding suggests that negative terminology can have a priming effect on readers such that they are predisposed to process evidence in a biased manner.