ABSTRACT. This study investigated teachers' perceptions of inappropriate student behavior as a function of students' and teachers' gender and ethnic background characteristics. The study used videotapes depicting inappropriate behavior of four different students (i.e., African-American female, African-American male, White female, White male). A total of 87 inservice teachers and 99 preservice teachers viewed the videotapes. Following each videotape, participants were asked to complete a 32-item behavior rating scale focusing on the teachers' perceptions of the student in each videotape. Analysis of the data revealed statistically significant differences related to the gender of the teacher, but not to the ethnic background of the teacher. Statistically significant differences also were found among students depicted in the videotapes. The African-American female student was judged to display more problem behaviors than the White female student. and the White male student was rated as displaying more classroom difficulties than the White female student. Teachers were not found to make judgments based on their ethnic background. The gender of the teacher and the gender of the student appeared to have the most influence on teachers' perceptions of behavior.
Rhodes (1967) characterized behavioral deviance as an interaction between the student's behavior and the response that the behavior provoked from others in the environment. This characterization of student behavioral problems as disturbing to others places emphasis on the investigation of teacher tolerance levels and teacher perceptions of deviant behavior (Kelly, Bullock, & Dykes, 1977; Prieto & Zucker, 1981; Shinn, Tindal, & Spira; 1987; Tobias, Cole, Zibrin, & Bodlakova, 1982; Tobias, Zibrin, & Menell, 1983; Zucker, Prieto, & Rutherford, 1979). According to Edelbrock (1983), teachers vary in their tolerance of student behaviors, in their ability to deal with problem behaviors, and in their expectations related to student behaviors. Similarly, Shinn and colleagues (1987) noted that different degrees of teacher tolerance for behaviors resulted in bias in referral and assessment. These authors also concluded that bias was affected by "prejudicial judgments influenced by naturally occurring pupil characteristics such as ethnic background, sex, and physical attractiveness" (p. 33).
The evaluation of students referred to special education for emotional or behavioral problems has been viewed as difficult and often subjective (Algozzine, Ruhl, & Ramsey, 1991; Schwartz, Wolfe, & Cassar, 1997). This process is even more complicated when addressing students from ethnically different backgrounds (Council for Children with Behavioral Disorders, 1989). Kauffman, Cullinan, and Epstein (1987), as well as Wilson and Bullock (1989), indicated that students' ethnic background and gender were two factors that affected teachers' decision making in student referral and ratings and needed further investigation.
Researchers have investigated the possibility that teachers may be biased in their decisions or assessment of students as a function of their gender or ethnic background. In one study, the teacher's gender was found to affect teacher perceptions of the behavior of students with emotional or behavioral disorders (E/BD) (Kelly et al., 1977). Kelly found that female teachers perceived higher percentages of students as exhibiting behavioral difficulties than did male teachers. On the other hand, Tobias and colleagues (1983) and McGee (1985) did not find significant differences in teacher ratings of the severity of behavior related to the gender of the teacher.
The ethnic background of the teacher also was found to produce conflicting results. Teachers of Hispanic origin, when compared to White or African-American teachers, tended to show a preference for mainstreaming students with disabilities (Tobias et al., 1982) and rated females more severely than males (Tobias et al., 1983). Tobias and colleagues (1982) found that teachers tended to refer students for special education placement more frequently when the students were from ethnic backgrounds other than their own; however, Tobias and colleagues (1983) did not find this interaction. Bahr, Fuchs, Stecker, and Fuchs (1991) found that both African-American and White teachers rated African-American students as more appropriate for referral to special education than White students.
Much of the conflicting data from studies may be explained by the research methodology used. For example, several studies (Prieto & Zucker, 1981; Zucker & Prieto, 1977; Zucker et al., 1979) were conducted that used fictitious case studies to examine the effect of ethnic background of the student on referral patterns. Teachers considered special education placement more appropriate when the students were described as Hispanic rather than as White. Other researchers (Matuszek & Oakland, 1979; Tobias et al., 1982), using similar methodology, found that the ethnic background of the student was not a significant influence on the teacher's determination to refer a student to special education. Two of these studies (Zucker & Prieto, 1977; Zucker et al., 1979) also investigated the influence of student gender on referral; in these studies, student gender was not found to influence the teacher's determination to refer the student to special education. However, Shinn and colleagues (1987) indicated that gender could not be discounted as a factor influencing teachers' referral decisions. Written case studies are limited in that no observation of actual behavioral differences or similarities occurs that might explain contradictory findings.
The purpose of this study was to extend the research regarding gender and ethnicity in the identification of students with problem behaviors. Limitations of previous studies have left the profession currently without a clear answer regarding this relationship. The research questions addressed in this investigation were designed to determine whether differences in teachers' perceptions as measured by ratings of inappropriate behavior of students from different gender and ethnic backgrounds exist (a) as a function of teachers' gender and ethnic backgrounds, (b) as a function of students' gender and ethnic backgrounds, or (c) as an interaction of these conditions.
Methods and Procedures
Participants
A total of 186 preservice (n = 99) and inservice (n = 87) teachers enrolled in graduate and undergraduate education courses were selected as participants in this study. Participants were selected from two state universities in the Southeastern United States, with a minority student representation of at least 24% each. Across both preservice and inservice teachers, 24.55% were African-American females, 11.8% were African-American males, 45.8% were White females, and 17.85% were White males.
Instrumentation
Videotapes
Four videotapes were created for data collection. The tapes featured a White male student, a White female student, an African-American male student, and an African-American female student as the target student for observation. These students were selected due to similar age and physical stature. Only partial views of the students' faces were videotaped in an attempt to control for physical attractiveness.
The videotapes were made in the same classroom and showed the same incidents of undesirable behaviors (e.g., pencil tapping, disturbing others, mild rebukes to the teacher). By staging the videotapes, consistency in factors other than the gender and ethnic background of the student was maintained. For example, all students other than the target student were in the same classroom location engaging in the same behaviors. The target students were dressed alike in jeans and sweaters and were prompted to engage in similar rates of desirable and undesirable behaviors while the teacher engaged in the same instructional and management behaviors across all tapes. This was accomplished by using written scripts. The tapes were reviewed by members of a graduate class in applied behavioral analysis to ensure that the undesirable behaviors were consistent across all tapes.
Rating Scale
The Brief Record of Observed Behavior (BROB), a modified version of the Direct Observation Form (DOF) (Achenbach, 1991), was used in this investigation. For this study, the BROB included only those items from the DOF that pertained to the behaviors observed in the videotapes described earlier. All participants in the study completed four BROB forms related to the behaviors of students viewed in each of the four videotapes. The edited BROB contained 32 items using a 4-point Likert scale with the following descriptors: not observed (0), ambiguous occurrence (1), definite occurrence of mild to moderate intensity (2), and definite occurrence with severe intensity (3).
Cronbach's coefficient alpha (i.e., internal consistency) was used to ascertain the reliability of the modified instrument. Results of the initial analysis indicated a coefficient alpha of .92. This reliability of the BROB compares favorably with the standard version of the DOF. Reliability of the BROB across the order of presentation also was evaluated using Cronbach's coefficient alpha. On the first behavior rating scale completed by the participants, a coefficient alpha of .90 was obtained. The second behavior rating scale had a coefficient alpha of .91, the third had a coefficient alpha of .92, and the fourth had a coefficient alpha of .93. Next, a principal factor analysis with Varimax rotation was conducted on the BROB. This analysis led to the clustering of items into three factors: Inattention, Defiance, and Impulsivity.
Procedures
Teachers participating in the study viewed all four of the videotapes of the students. Within each university class, the order of presentation of the videotapes was counterbalanced. Instructions were read from a script for each class regarding the respondents' task of rating the videos. In short, participants were instructed to watch each video and complete a BROB for the target student (i.e., African-American female, African-American male, White female, White male) after each tape. Thus, each participant completed four behavior rating scales.
Results
Prior to analyses for ethnic and gender differences, an analysis was completed to determine whether the order of presentation or the university site had an effect on respondents' ratings. No statistically significant differences were present in total scores as a function of order of completion or university setting.
A factoral analysis of variance (ANOVA) was conducted to evaluate the effects of teacher gender, teacher ethnic background, and the interaction of these factors on total scores of the BROB. This analysis was performed for the categories (a) African-American male student, (b) African-American female student, (c) White male student, and (d) White female student. The only statistically significant effect was found for the African-American female student as a function of teacher gender, F (1, 185) = 7.32, p < .05. Male teachers rated the African-American female student viewed on tape as more troublesome than did female teachers. Male teachers had higher BROB scores than female teachers for each student rated. However, as noted, the difference between male and female teachers reached statistical significance on only one student, the African-American female.
Student factor (i.e., Inattention, Defiance, Impulsivity) scores derived from the BROB also were analyzed as a function of teacher gender, teacher ethnic background, and the interaction of these factors. Again, the only statistically significant interaction was for the African-American female student, with a statistically significant difference in scores on Impulsivity as a function of teacher gender, F (1, 185) = 6.79, p < .05. Male teachers rated the impulsive behaviors of the AfricanAmerican female student higher than did female teachers.
A series of one-way repeated-measures ANOVA comparisons was conducted to evaluate the difference between student total scores on the BROB. Within-subject contrasts yielded a statistically significant difference between the groups of students, F (3, 183) = 7.26, p < .05. Therefore, a follow-up analysis of the BROB total scores with paired-samples t-tests was conducted to determine where overall differences occurred. Because multiple t-tests were performed, a Bonferroni test (Hocking, 1996) with an adjusted alpha level of .0125 was used (i.e., .05 divided by 4).
A paired-samples Rest for the African-- American female student and the White female student viewed on videotape yielded a statistically significant finding, t (185) of 2.66, p < .0125. Teacher ratings (n = 186) showed that the African-American female student was judged to exhibit more problem behaviors (M = 51.20, SD = 17.11) than the White female student (M = 48.64, SD = 17.27). A paired-- samples Rest (n = 186) for the White female student (M = 48.64, SD = 17.27) and the White male student (M = 51.58, SD = 17.92) viewed on videotape also showed statistically significant results, t (185) = -4.53, p < .0125. Teacher ratings showed that the White male student was viewed as displaying more problem behaviors than the White female student. No differences were revealed for the AfricanAmerican male and White male or for the African-American female and White male comparisons, all ps > .0125.
To determine whether a significant difference in factor (i.e, Inattention, Defiance, Impulsivity) scores existed, a one-way repeatedmeasures ANOVA was conducted for each of the three factor scores. Within-subject contrasts yielded a statistically significant difference between students on Defiance, F (3, 183) = 5.58, p < .05, and Impulsivity, F (3, 183) = 7.89, p < .05. No statistically significant difference was found between the groups of students on Inattention, F(3, 183) = 2.03, p > .05, which measured nonattending behaviors.
The repeated-measures ANOVA of Defiance and Impulsivity was followed by paired t-tests to determine where the differences in student ratings on these factors occurred. For Defiance, the paired samples ttest for the White male student and the White female student yielded a t (185) of -4.10, which was statistically significant at the .0125 level. The White male student had a mean rating of 7.95 (SD = 3.10), whereas the White female student (n = 186) had a mean rating of 7.21 (SD = 3.21). These results show that teachers rated the White male student seen in the videotape as displaying more problems related to defiance than the White female student. No statistically significant findings were found in the comparisons of the AfricanAmerican female to the White female, the African-American male to the White male, or the African-American female to the AfricanAmerican male (all ps > .0125).
On Impulsivity, the paired-samples t-test for the African-American male student and the White male student viewed by teachers yielded a t (185) of -4.52, which was statistically significant at the 0.125 level. The AfricanAmerican male student had a mean rating of 8.04 (SD = 4.87), and the White male student had a mean rating of 9.03 (SD = 4.88). From the behaviors observed in the videotapes, the White male student was judged by teachers to display more impulsive and immature behaviors than the African-American male student. A comparison of the rating of the White female student and the White male student also yielded statistically significant results, t (185) = -3.30, p < .0125. Again, the White male student was observed to display more difficulties in this area. No statistically significant differences were found for the comparisons of the African-American female and the White female or the African-American female and the White male.
Discussion
This study compared the responses of four groups of inservice and preservice teachers on a behavior rating scale. The four groups of teachers (i.e., African-American female, White female, African-American male, White male) rated videotapes of the behavior of four different students (i.e., African-American female, White female, African-American male, White male) engaged in similar classroom interactions. No statistically significant differences were found in ratings of the students related to the ethnic backgrounds of the teachers or the interaction of ethnicity and gender. Statistically significant findings were found related to the gender of the teacher. Male teachers were found to rate the African-American female student as having more behavioral difficulties than did female teachers. This same pattern was also found in the factor scores on Impulsivity. Again, male teachers rated the African-American female as displaying a higher level of impulsive behaviors than did female teachers. In fact, the male teachers' ratings of all students were higher than the female teachers' ratings; however, the only statistically significant difference was for the AfricanAmerican female student.
Interestingly, these findings are contrary to previous research findings. Kelly and colleagues (1977) found that male teachers were more tolerant of inappropriate behavior than female teachers. In his field-based study, teachers were asked to rate their students as having mild, moderate, or severe behavioral disorders. Using fictitious case studies, Tobias and colleagues (1983) and McGee (1985) did not find gender differences in teacher judgments of appropriateness of referral to special education. These studies did not use a behavior rating scale to measure teachers' tolerance for behaviors. It is possible that male teachers have fewer behavioral problems in their classrooms and make fewer referrals to special education for behavioral difficulty than female teachers. However, in this study male teachers' ratings of inappropriate behavior were higher than those of female teachers. Thus, male teachers may actually be less tolerant of inappropriate behaviors than female teachers.
In using case studies in which only the ethnicity and gender of the student were altered, Tobias and colleagues (1982) found that teachers referred students to special education more frequently when the students were from ethnic backgrounds other than their own. However, Tobias and colleagues (1983) failed to find differences in the decision to refer students to special education as a function of the students' and teachers' ethnic backgrounds. The results presented in this study are consistent with the 1983 findings in that the ethnic background of the teachers did not affect their ratings of students.
Bahr and colleagues (1991) found that although African-American and White students had similar scores on the Behavior Problem Checklist, a significantly larger number of African-American students were referred to special education by both African-American and White general education teachers. The authors noted that the differential referral of students to special education was more likely related to academic achievement than to differences in student behavior. Achievement (i.e., correct and incorrect academic responding) was a variable controlled across each of the four videotapes in this study and, therefore, should not have affected participant ratings.
Statistically significant differences in the total BROB score as a function of the students' gender and ethnic backgrounds were evident, with the White male student receiving the highest ratings of disruptive behavior. Additionally, the African-American female had statistically significant higher scores than the White female. These findings suggest that students' characteristics of gender and ethnic background may have an influence on teacher ratings.
Limitations
Clearly, conclusions drawn from these data are subject to several limitations. First, the videotapes showed only a brief sample of student behaviors; and even though the disruptive behaviors were controlled carefully for each student, some subtle differences in topography (e.g., vocal tone) could have been present that contributed to the findings. Extended observations of student behavior would allow a more in-depth investigation. A second limitation is that the data were collected in one geographic area of the country. Thus, students and teachers from the Hispanic culture were eliminated from this investigation due to the difficulty in obtaining a participant sample of Hispanic teachers. Third, to obtain a sample of participants that was reflective of gender and ethnic diversity, both inservice and preservice teachers were included, along with special education and non-special-education teachers at elementary, middle, and secondary levels. Individual studies with larger samples from each group might produce different results. Finally, most of the differences that were found to be statistically different were perhaps numerically different by only a few points; readers are cautioned to evaluate these close differences from a clinical perspective.
Conclusion
The findings in this study have implications for the assessment of students referred to special education for their behavioral difficulties. It appears that the teachers did not make judgments based on their ethnic backgrounds. The gender of the teacher and the gender of the student appeared to have the most influence on teachers' perceptions of behavior.
[Reference]
References
[Reference]
Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991 profile. Burlington: University of Vermont, Department of Psychiatry.
Algozzine, B., Ruh], K., & Ramsey, R. (1991). Behaviorally disordered? Assessment for identification and instruction. Reston, VA: Council for Exceptional Children.
[Reference]
Bahr, M. W., Fuchs, D., Stecker, P. M., & Fuchs, L. S. (1991). Are teachers' perceptions of difficult-toteach students racially biased? School Psychology Review, 4, 599-608.
Council for Children with Behavioral Disorders. (1989). White paper on best assessment practices for students with behavioral disorders: Accommodation to cultural diversity and individual differences. Behavioral Disorders, 14, 263-278.
Edelbrock, C. (1983). Problems and issues in using rating scales to assess child personality and psychopathology. School Psychology Review, 12, 293-299.
[Reference]
Hocking, R. R. (1996). Methods and applications of linear models: Regression and the analysis of variance. New York: Wiley.
Kauffman, J. M., Cullinan, D., & Epstein, M. H. (1987). Characteristics of students placed in special programs for the seriously emotionally disturbed. Behavioral Disorders, 12, 175-184.
Kelly, T. J., Bullock, L. M., & Dykes, M. K. (1977). Behavioral disorders: Teachers' perceptions. Exceptional Children, 43, 316-318.
Matuszek, P., & Oakland, T. (1979). Factors influencing teachers' and psychologists' recommendations regarding special class placement. Journal of School Psychology, 17, 116-125.
[Reference]
McGee, D. W. (1985). The role of gender in teacher perception of disturbance (Doctoral dissertation, University of California, 1985). Dissertation Abstracts International, 46, 06A: 1595.
Prieto, A. G., & Zucker, S. H. (1981). Teacher perception of race as a factor in the placement of behaviorally disordered children. Behavior Disorders, 7, 34-38.
Rhodes, W. (1967). The disturbing child: A problem of ecological management. Exceptional Children, 33, 49-55.
Schwartz, N. H., Wolfe, J. N., & Cassar, R. (1997). Predicting teacher referrals of emotionally disturbed children. Psychology in the Schools, 34, 51-61.
Shinn, M. R., Tindal, G. A., & Spira, D. A. (1987). Special education referrals as an index of teacher tolerance: Are teachers imperfect tests? Exceptional Children, 54, 32-40.
[Reference]
Tobias, S., Cole, C., Zibrin, M., & Bodlakova, V. (1982). Teacher-student ethnicity and recommendations for special education referrals. journal of Educational Psychology, 74, 72-76.
Tobias, S., Zibrin, M., & Menell, C. (1983). Special education referrals: Failure to replicate studentteacher ethnicity interaction. Journal of Educational Psychology, 75, 705-707.
Wilson, M. J., & Bullock, L. M. (1989). Psychometric characteristics of behavior rating scales: Definitions, problems, and solutions. Behavioral Disorders, 14, 186-200.
[Reference]
Zucker, S. H., & Prieto, A. G. (1977). Ethnicity and teacher bias in educational decisions. Instructional Psychology, 4, 2-5.
Zucker, S. H., Prieto, A. G., & Rutherford, R. B. (1979). Racial determinants of teachers' perceptions of placement of the educable mentally retarded. (ERIC Document Reproduction Service No. ED 171 015)
[Reference]
MANUSCRIPT:
[Reference]
Initial Acceptance: 8/23/00
Final Acceptance: 9/18/00
[Author Affiliation]
Paulette B. Taylor
Ware County Schools
Philip L. Gunter
Valdosta State University
John R. Slate
University of Texas at El Paso
[Author Affiliation]
AUTHORS' NOTE:
[Author Affiliation]
Readers desiring a more detailed version of this manuscript with accompanying tables should contact Philip Gunter, Department of Special Education and Communication Disorders, Valdosta State University, Valdosta, GA 31698. E-mail: pgunter@valdosta.edu. This manuscript is taken from a portion of the doctoral dissertation completed at Valdosta State University by Dr. Paulette Taylor.
[Author Affiliation]
AUTHORS:
[Author Affiliation]
PAULETTE B. TAYLOR, School Psychologist, Ware County Schools, Waycross, GA. PHILIP L. GUNTER, Professor and Department Head, Department of Special Education and Communication Disorders, Valdosta State University, Valdosta, Georgia. JOHN R. SLATE, Professor, Department of Educational Leadership and Foundations, University of Texas at El Paso.
