Teacher/student interactions and classroom behavior: the role of student temperament and gender.

McClowry, Sandra Graham ; Rodriguez, Eileen T. ; Tamis-LeMonda, Catherine S. 等

The purpose of this study was to examine the relationships of student temperament and gender to disruptive classroom behavior in urban primary grade schools. Teacher reports and classroom observations were used. Forty-four teachers and their 152 students participated. A two-step cluster analysis was conducted with teacher reports on their students' temperaments. Three temperament clusters were identified: industrious, intermediate, and high maintenance. ANOVAs revealed that, as compared to students with other temperaments, children who were high maintenance exhibited significantly higher levels of overt aggression toward others, emotional-oppositional behavior, attentional difficulties, and covert disruptive behavior. Teachers reported more difficulty managing the behavior of high maintenance students and were observed to provide more negative feedback to them compared to those who were industrious. Hierarchical and logistic regression analyses demonstrated that temperament mediated the relationship between student gender and disruptive classroom behaviors. Temperament also mediated the association between gender and teachers' difficulty managing students' covert disruptive behavior. Irrespective of gender, students whose temperaments were high maintenance and intermediate were more likely than industrious students to receive negative teacher feedback. Irrespective of students' temperament, teachers were observed to provide more positive feedback to boys than to girls.

Keywords: temperament, disruptive student behavior, teacher/student interactions, school children

**********

Dynamic interactions between teachers and their students occur in elementary school classrooms on a daily basis. Whether engaged in instruction or transitioning between activities, teachers and students have myriad opportunities to interact with each other--a topic that has been the focus of numerous studies (Beaman & Wheldall, 2000; Jones & Dindia, 2004; Kelly, 1988; Sutherland, 2000). Overall, the findings show that elementary schoolteachers provide much more negative than positive feedback to their students. When provided, positive feedback is associated with good academic performance. Students seldom receive positive feedback when meeting their teachers' behavioral expectations (e.g., standing quietly in line). In contrast, negative teacher feedback occurs more frequently and is often precipitated by disruptive student behavior.

Gender is frequently associated with the amount and quality of teacher-student interactions. Two meta-analyses have elucidated under what circumstances teacher feedback differed by student gender (Jones & Dindia, 2004; Kelly, 1988). No gender differences were found in the amount of positive feedback teachers provided their students. Ironically, however, girls who exhibited behavior that teachers valued received less overall attention than boys (Kelly, 1988), although the magnitude of these differences were relatively small (d = . 14) (Jones & Dindia, 2004). In contrast, boys received more negative feedback with effect sizes ranging from small to moderate (d = .34) (Jones & Dindia, 2004). The greater amount of negative feedback that boys received was attributed by Jones and Dindia (2004) to the higher level of disruptive behavior that boys exhibited compared to girls. Teachers used negative feedback in attempting to control their behavior (Broidy et al., 2003; Coie & Dodge, 1998; Jones & Dindia, 2004; Rescorla et al., 2007). This assertion was supported by another study that showed teachers perceived their male students as more difficult to manage than the girls (Childs & McKay, 2001).

The modest effect sizes of gender on teacher/student interactions suggest that other moderating and/or mediating factors may be operating (Brophy & Good, 1974; Jones & Dindia, 2004). A competing explanation is that negative teacher feedback results from disruptive student behavior, regardless of gender. This assertion was supported by Sutherland (2000), who observed that disruptive students, compared to those who were better behaved, had more interactions with their teachers. Likewise, Kelly (1988) found that the difference in the amount of negative feedback boys and girls received was smaller among disruptive students.

Another student characteristic that modifies teacher-student interactions is temperament (Keogh, 2003; Lerner, Lerner, & Zabski, 1985; Pullis & Cadwell, 1982). School-age students whose temperaments were low in task persistence, high in activity, and high in negative reactivity were likely to have negative interactions with their teachers and to exhibit disruptive classroom behavior (Kean, 1995; Prior, Sanson, Smart, & Oberklaid, 2000; Rothbart & Bates, 2006). In contrast, students who were high in task persistence were more likely to experience positive teacher-student relationships (Guerin, Gottfried, Oliver, & Thomas, 1994; Keogh, 2003; Prior et al., 2000). Those whose temperaments were low in activity and negative reactivity, in addition to being high in task persistence, were perceived by their teachers as teachable and competent (Nelson, Martin, Hodge, Havill, & Kamphaus, 1999; Prior et al., 2000; Rothbart & Bates, 2006). Teachers gave such students more positive feedback (Van de Werfhorst, 1986) and perceived them as requiring less supervision (Pullis & Cadwell, 1982).

Another temperament dimension that influences teacher-student interactions is withdrawal. Several studies have found that students whose temperaments were high in withdrawal were shy and reticent to engage with their teachers compared to their more social classmates (Henderson & Fox, 1998; Rudasill & Rimm-Kaufman, 2009; Sanson, Hemphill, & Smart, 2004).

Although student gender and temperament frequently have been examined in relation to disruptive classroom behavior, less is known about their combined contributions to classroom management and to teacher-student interactions. Moreover, no previous studies have examined these constructs in inner-city schools with populations of economically disadvantaged children at risk for developing disruptive behavior problems (Institute of Medicine, 1994). Understanding how temperament and gender influence student behavior can provide teachers with a framework for interpreting classroom dynamics. Such knowledge may lead to the implementation of teacher strategies that could enhance classroom management. The purpose of this study was to examine the relationships of student temperament and gender to disruptive classrooms behavior in urban primary grade schools. The study had three objectives: (1) to identify temperament profiles among 1st- and 2nd-grade students; (2) to determine whether students with particular temperament profiles differed significantly in their levels of disruptive classroom behavior (both teacher reported and observed), in teachers' reported problems managing their behavior, and in teachers' use of positive and negative feedback; and (3) to determine whether the relationships between student gender and disruptive classroom behaviors, teachers' problems managing disruptive behavior, and teacher positive and negative feedback were mediated by student temperament.

METHOD

Participants and Setting

Participants in the study included 152 students and their 44 teachers in general education 1st- and 2nd-grade classrooms. The six partnering schools were from one urban school district in a large northeastern city in the United States. Department of Education statistics reported that all of the schools were low performing and served predominantly African American families. Approximately 86% of the children in the schools qualified for free or reduced-price lunch programs.

Fifty-six percent of the students were male (n = 85) and 44% were female (n = 67). Children ranged from age 5 to 9 (M = 6.61, SD = 0.81). Approximately two thirds of students (n = 101) were in the 1st grade and the remaining one third (n = 51) were in the 2nd grade. The race/ethnicity of the children was reported as 88% African American (n = 133), 9% Hispanic (n = 14), and 3% racially mixed (n = 5). Fifty-six percent (n = 86) of the children lived in single-parent homes, 37% (n = 56) lived in a two-parent home, and 7% of the adult respondents (n = 10) declined to report their family configuration.

Teacher participants included 29 first-grade and 15 second-grade teachers (41 female, 3 male). Seventy-nine percent of teachers (n = 35) reported their race/ethnicity as African American, 7% (n = 3) as White, 7% (n = 3) as Hispanic, and 7% (n = 3) as Asian.

Measurement

Student temperament was assessed using the Teacher School-Age Temperament Inventory (T-SATI), a self-report measure that consists of 36 items rated on a 5-point Likert-type scale (ranging from never to always) (Lyons-Thomas & McClowry, 2012). The T-SATI is an adaptation of the parent report School-Age Temperament Inventory (SATI) developed by McClowry (1995). Both versions have four dimensions: negative reactivity, task persistence, withdrawal, and motor activity. The T-SATI takes approximately 10 minutes to complete. In previous studies, the Cronbach's alphas for the T-SATI dimensions ranged from .89 to .96 (Lyons-Thomas & McClowry, 2012). As shown in Table 1, the alphas in this study ranged from .70 to .94.

Teachers reported on student disruptive behavior using the Sutter-Eyberg Student Behavior Inventory (SESBI) (Eyberg & Pincus, 1999), which is the teacher version of the Eyberg Child Behavior Inventory (Eyberg & Pincus, 1999). The SESBI consists of 36 items that are each rated along two scales: (1) a 7-point Likert-type intensity scale (ranging from never to always) in which teachers report the frequency of occurrence of each behavior in the classroom, and (2) a problem scale in which teachers endorse (yes or no) whether they experience difficulty managing each of the stated behaviors. In a prior paper, subscales of the SESBI were identified using principal components factor analysis with varimax rotation (McClowry, Snow, Tamis-LeMonda, & Rodriguez, 2010). Item loadings ranged from .54 to .86 and closely replicated factors identified by Burns and Owen (1990) and Bums, Walsh, and Owen (1995). The derived subscales were as follows: overt aggression toward others, emotional-oppositional behavior, attentional difficulties, and covert disruptive behavior. The highest loading items for each of the four subscales, respectively, include teases or provokes other students, cries, has difficulty staying on task, and steals. For each of the intensity subscales, scores were computed as the mean of the component items; problem subscale scores were computed as the total number of behaviors endorsed by the teacher as problematic or difficult to manage. As shown in Table 1, the alphas for the intensity subscales ranged from .84 to .96. The Kuder-Richardson reliability for the problem subscales ranged from .74 to .94.

Observational measures of student disruptive behavior and teacher feedback were obtained using the Primary Classroom Observation Scale (PCOS) (Tamis-LeMonda, Briggs, & Carlson, 2008). The PCOS is a partial time-sampling coding system in which the occurrence of specific student and teacher behaviors are observed in 30-second intervals and then recorded during a 30-second off period. During the one-hour observation period, a total of 24 behaviors are coded, along with a narrative description of teachers' responses to student disruptions.

Prior to conducting the observations, two coders were trained over an 8-week period and reached a reliability level of over 90% agreement (based on a comparison of independent ratings of the same classrooms). Reliability checks continued throughout the duration of the observations conducted for the 44 classrooms; specifically, each of the two coders independently rated every eighth classroom as a reliability check, and was required to maintain over 90% agreement on observed student and teacher behaviors. Percent agreements were calculated separately for each coded behavior, as indicated by the number of times both observers marked the same teacher or student behavior as occurring during the same 30-second coding interval.

Because the focus of this study was on student disruptive behavior and teacher feedback, only variables from the PCOS that pertained to these constructs were examined. Student disruptive behavior comprised five observed behaviors: calling out, roaming around the classroom, annoying other students, being physically aggressive toward other students, and exhibiting oppositional/noncompliant behavior in response to a teacher's requests. The occurrences of two types of teacher feedback were also observed. Positive feedback was coded when a teacher's statement communicated something positive about the child or his/her academic performance (e.g., "You really are working hard") rather than merely reinforcing an answer given in response to the teacher's question (e.g., "That's correct"). Negative feedback was coded when a teacher's statement contained a negative evaluation of the child's ability, work habits, and/or motives (e.g., "You really are not working hard"). For each of these observed variables, the total instances observed were recorded during the coding period and then prorated to reflect rates per hour of observed behaviors. These variables are further described in the appendix.

Procedure

Data for this study were obtained at baseline from teachers and children who were participating in a preventive intervention INSIGHTS into Children's Temperament (McClowry et al., 2010). This study focuses on the classroom component, with data collected from participating teachers and classroom observations conducted prior to the initiation of the 10-week intervention. Recruitment of 1st- and 2nd-grade teachers involved a 30-minute information session conducted at each of the six participating schools. Once a classroom teacher provided informed consent to participate, a variety of strategies were implemented to recruit parents, including sending letters, posting information flyers at the school, making telephone calls, and conducting brief presentations at parent meetings. After a parent agreed to participate, his or her child was asked to give assent.

Teachers completed the T-SATI and the SESBI at baseline on each participating student. Neither the teachers nor the coders were aware of the students' scores on these instruments when the classroom observations were conducted.

Classrooms were observed at baseline by trained coders using the PCOS (Tamis-LeMonda et al., 2008). Observations were conducted during morning lesson time and did not include out-of-classroom time (e.g., lunch, gym), special events (e.g., classroom parties, parent visits), or occasions during which teaching assistants or other staff were responsible for leading the class. Observers stationed themselves in an unobtrusive location that provided a good view of the entire classroom and refrained from interacting with students or engaging in any classroom activities that were taking place during the observation period. Each observer had a timer, pencil, and code sheet for recording student and teacher behaviors. Data on the students and participating teachers were later tallied from the coding sheets for subsequent analyses. Classroom observations averaged 52 minutes in length for each target student (SD = 33.18).

RESULTS

Table 1 presents descriptive data on teacher reports of student temperament, teacher-reported occurrences of disruptive behavior and reported difficulty managing disruptive behavior, and observed student disruption and teacher feedback (i.e., positive and negative). Of note, 38% of the students were observed to be disruptive. Fifty-six percent of students received negative teacher feedback, whereas only 20% received positive teacher feedback. Because the distributions of these observational variables were not normally distributed, nonparametric statistics were conducted after dummy coding these behaviors as observed or not observed.

Bivariate associations among all the variables are presented in Table 2. As shown, the temperament dimensions of negative reactivity and activity were positively associated with all teacher-reported measures of disruptive behavior (rs = .48 - .82, ps < .001) and difficulty managing disruptive behavior (rs = .31 - .63, ps < .01). Task persistence was negatively related to these same measures (rs = -.26 - -.79, ps < .001). A similar pattern of associations upheld for measures of observed student disruption and negative teacher feedback. Specifically, student negative reactivity and activity were positively associated with measures of observed student disruption (r = .26 and .25, ps < .01) and negative feedback (r = .35 and .30, ps < .001), whereas task persistence was negatively associated with these same measures (r = -.17 and -.23, ps < .05).

In addition, teachers' reports of student disruptive behaviors were all highly related to teachers' reported difficulty managing these same behaviors (rs = .43 - .82, ps < .001). Moreover, teachers' reports of student disruptive behaviors were positively associated with observed measures of student disruption (rs = .21 - .32, ps < .01) and negative teacher feedback (rs =. 18 - .29, ps < .05). Teachers' reports of difficulty managing student disruptive behaviors were similarly associated to these same observed measures. Positive teacher feedback, however, was not associated with any of the study variables.

Profiles of Student Temperament

To examine the construct validity of student temperament profiles among 1st- and 2nd-grade students, a TwoStep Cluster Analysis was conducted with teacher reports on the T-SATI. The TwoStep auto-cluster procedure offers several advantages over traditional clustering techniques (e.g., k-means, hierarchical). First, the number of clusters does not need to be selected a priori; instead, the algorithm automatically determines the optimal number of clusters based on a number of criteria (elaborated below). In addition, simulation studies have shown that the combination of distance measures and criterion statistics (such as Bayesian Information Criterion [BIC] or Akaike Information Criterion [AIC]) yield better estimation that either one alone (SPSS, 2001).

The first step of the TwoStep clustering procedure involves the formation of preclusters, to which cases are assigned using a sequential clustering approach (Theodoridis & Koutroumbas, 1999). The BIC or AIC for each number of clusters within a specified range is calculated and used to find the initial estimate for the number of clusters. In the second step, the preclusters are clustered using an agglomerative hierarchical clustering algorithm, producing a range of solutions that differ in the number of derived clusters. The algorithm selects the optimal number of clusters based on the Schwarz's BIC; the solution with the lowest BIC coefficient is deemed optimal. Additional criteria used to index fit include large BIC ratio of change and distance measure statistics.

The clustering procedure was conducted using the four dimension scores derived from the TSATI: negative reactivity, task persistence, withdrawal, and activity. An initial inspection of the variable-wise importance plots and associated students' t statistics revealed that the dimension of withdrawal did not contribute to the discrimination of the clusters (i.e., mean values were virtually identical across all cluster groupings, irrespective of the number of derived clusters). Accordingly, the analysis was repeated using the three temperament dimensions of negative reactivity, task persistence, and activity. Using the above-specified criteria, the auto-clustering algorithm indicated that a three-cluster solution was optimal (BIC = 247.375). As shown in Table 3, the first cluster comprised 35 children (23%) with temperament profiles characterized as industrious (high scores on task persistence and low scores on negative reactivity and activity), a second cluster consisted of 56 children (37%) with temperament profiles characterized as high maintenance (low scores on task persistence and high scores on negative reactivity and activity), the final cluster included 61 children (40%) who did not meet the criteria for either the industrious or high maintenance profiles and who were called intermediate. Based on Bonferroni post-hoc comparisons, students' mean scores on the temperament dimensions of negative reactivity, task persistence, and activity each significantly differed across the three profiles (p < .001).

Although the derived solution was objectively determined on the basis of BIC and additional fit criteria, the stability of the cluster groups, as well as the degree to which clusters were heterogeneous on the study outcomes of interest, was examined. Specifically, two steps were taken to validate the three-cluster solution. First, the stability of the cluster groups was cross validated using cut-points derived from a standardized sample of 243 elementary school-age children provided by a national sample of teachers (Lyons-Thomas & McClowry, 2012).

A comparison of the TwoStep cluster groupings with those derived from the standardized cut-off scores revealed a 97.1% concordance for children characterized as industrious, an 83.6% concordance rate for children classified as intermediate, and an 83.9% concordance rate for children classified as high maintenance. However, the pattern of findings for all subsequent analyses was identical when using both methods of classification. Findings using the TwoStep method of clustering are presented in this article.

Finally, because McClowry (2002) found that the proportion of boys and girls differed on temperament profiles, gender also was examined. As illustrated in Figure 1, a significant chi-squared analysis revealed different patterns between girls and boys, [chi square](2) = 13.98, p < .001. Boys were disproportionately represented in the high maintenance profile (71% vs. 29%), whereas girls were overrepresented in profiles characterized by an industrious temperament (69% vs. 31%).

Profiles of Student Temperament in Relation to Classroom Dynamics

The next stage of this study examined how the different temperament profiles were related to classroom dynamics. Specifically, the analyses examined whether students with particular temperament profiles differed significantly on measures of disruptive behavior (both teacher reported and observed), teachers' reported difficulty managing disruptive behavior, and teacher positive and negative feedback.

[FIGURE 1 OMITTED]

ANOVA yielded a significant main effect of temperament profile for each of the four student disruptive behaviors. As shown in Table 4, teachers reported that high maintenance pupils, as compared to industrious and intermediate students, exhibited significantly higher levels of overt aggression toward others, emotional-oppositional behavior, attentional difficulties, and covert disruptive behavior. Further, as compared to industrious children, intermediate students had significantly more occurrences of overt aggression toward others, attentional difficulties, and covert disruptive behavior.

Analyses also examined teachers' reported difficulty managing students' disruptive behavior (also in Table 4). As shown, teachers reported significantly more difficulty managing the behavior of high maintenance students as compared to their industrious and intermediate peers on all four types of disruptive classroom behaviors. Teachers also reported more difficulty handling the attentional difficulties of intermediate children as compared to their industrious counterparts.

As shown in Table 4, chi-square analyses revealed significant differences in disruptive classroom behavior and negative teacher feedback by temperament profile. Specifically, students whose temperaments were characterized as intermediate and high maintenance were disproportionately represented among children exhibiting disruptive behaviors (45% and 45%, respectively) when compared to their industrious peers (10%), [chi square](2) = 9.68, p = .008. In addition, teachers were observed to provide significantly more negative feedback to intermediate (45%) and high maintenance (43%) students as compared to their industrious counterparts (12%), [chi square](2) = 13.97, p = .001. No significant temperament differences were found for positive teacher feedback, [chi square](2) = 1.93, p = .381.

The last stage of the analysis examined whether the relationships between student gender and disruptive classroom behaviors, teacher problems managing disruptive behavior, and teacher positive and negative feedback were mediated by student temperament profiles. Based on guidelines from Baron and Kenny (1986), hierarchical and logistic regression methods were used. All categorical predictors were dummy coded. Specifically, the first step of models included whether the child was male; students' temperament profile was entered in the second step of models (intermediate, high maintenance, with industrious as the omitted reference group). Estimates from models that included only student gender were compared to model estimates that included student gender and temperament. This allowed for examination of the potential role of temperament as a mediator of the relationship between gender and student disruptive behaviors, teacher difficulty managing behavior, and observed disruption and teacher feedback (positive and negative) in the classroom. Hierarchical linear regression was used in models that examined teachers' reported occurrence of student disruptive behavior and difficulty managing these behaviors. To permit for the examination of marginal effects, logistic regression was used for observed measures of student disruption and teacher feedback in the classroom; for these analyses, each observed measure was coded as having occurred (coded 1) versus not (coded 0) during the one-hour observation period.

As can be seen in the top half of Table 5, a full mediation model was supported. Student temperament mediated the relationship between gender and overt aggression toward others, emotional-oppositional behavior, attentional difficulties, and covert disruptive behavior. That is, for each of the four types of disruptive classroom behaviors, the effect of gender attenuated to nonsignificance when temperament was entered into the model. As compared to their industrious peers, children with high maintenance temperaments were reported by their teachers as exhibiting significantly higher levels disruptive classroom behavior. Moreover, students with temperaments characterized as intermediate exhibited more overt aggression toward others, attentional difficulties, and covert disruptive behavior than children with industrious temperaments.

Also shown in the bottom half of Table 5, student temperament was shown to mediate the relationship between gender and teacher management of covert disruptive behavior. As can be seen, the effect of gender on students' covert disruptive behavior attenuated to nonsignificance when temperament was entered into the model. In addition, temperament was significantly related to teacher difficulty managing students' overt aggression toward others, emotional-oppositional behavior, attentional difficulties, and covert disruptive behavior. Consistently, teachers perceived the behavior of their high maintenance students as more difficult to manage than that of their industrious students. They also perceived more difficulty managing the overt aggressive behavior and attentional difficulties of intermediate children compared to students with industrious temperaments. Gender was not related to teachers' management of these disruptive classroom behaviors.

Finally, logistic regression analyses examined observed measures of students' disruptive behavior and teachers' use of feedback in the classroom. As shown in Table 6, the effect of gender on observed disruptive behavior attenuated to nonsignificance when students' temperament was simultaneously entered into the model. This finding indicates that students' temperament mediates the association between student gender and disruptive classroom behavior. Specifically, students whose temperaments were characterized as intermediate (odds ratio [OR] = 4.02, p < .05) and high maintenance (OR = 4.59, p < .01) were more likely than their industrious counterparts to demonstrate disruptive behavior.

Also shown in Table 6 are models that examined teachers' observed use of positive and negative feedback in the classroom. Findings of these analyses did not support a mediation model. Instead, analyses revealed that teachers provided significantly more positive feedback to boys than girls (OR = 3.72, p < .01), irrespective of students' temperament. Moreover, controlling for the effects of gender, students whose temperaments were characterized as intermediate (OR = 3.98, p < .01) and high maintenance (OR = 4.57, p < .01) were more likely than industrious students to receive negative feedback from their teachers.

DISCUSSION

The purpose of this study was to examine the relationships of student temperament and gender to disruptive classrooms behavior in urban primary grade schools. The findings clearly demonstrate the strong associations between student temperament and classroom disruptive behavior. Students whose temperaments were high maintenance--that is, low in task persistence and high in negative reactivity and activity--were more disruptive than students whose temperaments were characterized as intermediate or industrious. Teachers in this study also reported more difficulty managing the behavior of students with high maintenance temperaments. These findings are consistent with previous cross-sectional and longitudinal studies that have shown that children with challenging temperaments, like those described as high maintenance in this report, exhibit more disruptive behavior and are more difficult to manage, both at school and at home, compared to children with milder temperaments, such as those who are industrious (Caspi, Henry, McGee, & Silva, 1995; Keogh, 2003; McClowry et al., 1994; Smart et al., 2003).

The high maintenance and industrious profiles identified in this study are consistent with those McClowry (2002) derived from parent reports. Whereas the 883 children in that study were from ethnically and socioeconomically diverse families, this study included only African American and Hispanic students from predominantly low-income families. In both studies, boys were disproportionately represented on the high maintenance profile, whereas girls were disproportionately industrious. Still, 29% of the students in this study with industrious temperaments were boys and 31% of children with high maintenance temperaments were girls.

The results, however, explicate the critical need to untangle temperament from gender when studying child disruptive behavior. When the effects of gender were examined alone, boys were, as expected, more disruptive than girls. However, when temperament was also taken into account, the effect of gender on student disruptive behavior attenuated to non-significance. In other words, temperament was a stronger predictor of student disruptive behavior than child gender.

Temperament also influenced observed teacher-student interactions. Students whose temperaments were high maintenance or intermediate, compared to industrious students, received more negative feedback from their teachers regardless of their gender. This finding is supported by a meta-analysis of gender differences in teacher/student interactions conducted by Kelly (1988), who concluded that the generality that boys receive more negative feedback from their teachers did not hold among girls who are disruptive. In this study, students with high maintenance temperaments received five times more negative feedback than their industrious classmates. Students with temperaments characterized as intermediate received 4 times more negative feedback than industrious children.

A different pattern of interactions was associated with positive teacher feedback. Notably, only 20% of the students received any positive feedback. Temperament was not associated with positive teacher feedback--just gender. Boys were more likely to receive positive teacher feedback than were girls. These findings are corroborated by other observational studies that found that girls receive little attention from their teachers (Kelly, 1988; Rudasill & Rimm-Kaufman, 2009).

The finding that temperament was not related to positive teacher feedback was counterintuitive. One might have expected that students with industrious temperaments, who were high in task persistence and low in negativity and activity, would have experienced higher levels of positive feedback because these are the attributes that teachers value (Keogh, 2003). However, they did not.

The overall proportion of observed negative to positive feedback is striking. Teachers gave nearly 3 times more negative than positive feedback to their students. Previous studies also have shown that students receive much more negative than positive feedback (Beaman & Wheldall, 2000; Jones & Dindia, 2004; Kelly, 1998; Sutherland, 2000). The amount of negative feedback in this study warrants concern, because it was directed at economically disadvantaged minority children. Such students are vulnerable to the quality of their relationships with their teachers and particularly benefit responsive student-teacher interactions (Meehan, Hughes, & Cavell, 2003; O'Connor, 2010).

The lack of findings regarding the withdrawal temperament dimension requires further consideration. In this study, withdrawal did not contribute to the cluster analysis and was not associated with any of the other variables, with the exception of a small negative correlation with overt aggression toward others. Withdrawal may not have been related to the findings in this study for several reasons. Teachers are more observant of disruptive student behavior than internally oriented states, such as withdrawal (Gresham, Elliott, Cook, Vance, & Kettler, 2010; Kolko & Kazdin, 1993). As a result, students who are high in withdrawal receive less attention from their teachers than their classmates who are not shy (Rudasill & Rimm-Kaufman, 2009). Another inference may be that the withdrawal dimension represents a distinctly different temperament profile from the other three dimensions of negative reactivity, task persistence, or motor activity. Instead, it may operate separately. This conclusion is supported by the extensive longitudinal research conducted by Kagan and his colleagues who focused exclusively on inhibited versus uninhibited children (Kagan, Snidman, & Arcus, 1992). Further research is needed to more closely examine how the temperament dimension of withdrawal is related to student classroom behavior and to teacher/student interactions.

STRENGTHS AND LIMITATIONS

The results of these analyses should be considered in relation to the strengths and limitations of the study. One noteworthy strength was the inclusion of teachers' reports of their perceptions of their students' temperaments and observational data. The two sources of data were consistent in demonstrating that students with high maintenance temperaments demonstrated higher levels of disruptive behavior than their intermediate and industrious peers, were perceived by their teachers as more difficult to manage, and received significantly more negative feedback.

Another strength of this study was its setting in urban primary grade classrooms primarily composed of African American students and teachers. A resulting limitation, however, is the relatively homogenous nature of the student population. Qualitative research is recommended to further explore the cultural implications of teacher/student interactions. Comparisons with suburban and rural classrooms also are needed to assess whether these results generalize to other educational contexts and with students and teachers from various socioeconomic and racial/ethnic groups.

An additional limitation is related to the amount of demographic information obtained about the teachers. Level of educational preparation and years of teaching might have been related to the type of feedback provided by teachers. Some of the variation in the types and amounts of teacher feedback also may have been influenced by when the observations were gathered, given observations were conducted over the course of the academic year (e.g., fall vs. spring). In fact, Chow and Kasari (1999) found that the quality of teacher-student interactions changes notably throughout the school year. The small sample size in this study is another limitation that prohibited formally testing the mediation effect. The magnitude of the effects seen in the logistic models indicated that there was not enough power to detect a statistically significant effect using Sobel's z (Fritz & MacKinnon, 2007). A small to moderate effect size would have required sample size of greater than 421 to detect an effect of statistical significance (power = .80). Based on Baron and Kenney's model (1986), however, the results suggest mediation.

IMPLICATIONS FOR CLASSROOM MANAGEMENT

The findings of this study have important implications for classroom management. Teachers are often unaware of how often they provide negative versus positive feedback (Good & Brophy, 2008; Jones & Dindia, 2004; Sutherland, 2000). Although a 3:1 to 4:1 ratio of positive to negative feedback is recommended (Stichter, Stormont, Lewis, & Schultz, 2009), the opposite pattern was found in this study. The practice implications reverberating from this finding cannot be overstated, because negative teacher feedback has deleterious effects on students. Conflictual relationships between primary grade students and their teachers lay the foundation for compromised academic and behavioral outcomes (Birch & Ladd, 1997; Pianta, Steinberg, & Rollins, 1995).

The importance of positive teacher-student relationships is particularly critical for high-risk students (O'Connor, 2010). Montague and Rinaldi (2001) demonstrated that the window of opportunity to reach high-risk children is narrow. Although children in 1st and 2nd grade were not explicitly aware of their teachers' negative feedback, they were by 3rd grade and, in turn, viewed themselves more negatively (Montague & Rinaldi, 2001).

Frequent negative teacher feedback is counterproductive, because it heightens rather than reduces disruptive behavior (Nelson & Roberts, 2000). Teacher preparation and professional development programs, however, can assist teachers to use evidence-based strategies to better manage student classroom behavior. A limited but expanding number of such programs exist. For example, INSIGHTS Into Children's Temperament, which applied the temperament framework derived from this study, effectively supports teacher efficacy and reduces student disruptive classroom behavior (McClowry et al., 2010). Positive Action is a character development program that enhances student academics and behavior by reinforcing positive actions (Beets et al., 2009). The Classroom Organization and Management Program is a teacher professional development program that assists teachers is creating a classroom environment that fosters student engagement (Evertson & Smithey, 2000). Regardless of the theoretical framework used by such programs, the aim is to enhance teacher/student relationships and classroom management--a goal that is empirically supported by the findings from this study.

DOI: 10.1080/02568543.2013.796330

APPENDIX
PRIMARY CLASSROOM OBSERVATION SCALE (PCOS): OPERATIONAL
DEFINITIONS AND EXAMPLES

Observed Variables Operational Definition Example

Student disruptive Disruptive behavior is
 behavior the sum of the
 following codes:
 Verbal--yelling out an "Can I go to the
 answer out of turn bathroom?"
 or asks to do
 something that is
 unrelated to the
 ongoing classroom
 work.
 Roam--not sitting in Leaving his or her
 the assigned seat. seat or sitting in
 his or her chair in
 an inappropriate
 manner.
 Annoy--a child Calling names or
 intentionally annoys throwing a paper.
 a classmate.
 Oppositional/ Silence or verbal
 noncompliance-- refusal: "I don't
 refusing to comply have to do that if I
 with a teacher's don't want to. You
 request. can't make me."
 Child aggression--more Pushing, pulling, or
 extreme forms of hitting.
 child antisocial
 behaviors.

Teacher feedback:

Positive A statement that "Very good, Sasha, you
 communicates really are working
 something positive hard."
 about the child or
 his/her performance,
 rather than merely
 about the answer to
 the question. This
 feedback is made in
 relation to a
 child's intentions,
 approach to task,
 effort, motivation,
 or behavior.
Negative A statement directed "I can tell, Roland,
 to a child that is that you are not
 characterized by listening. If you
 negative evaluation had been paying
 of the child's attention you
 ability, work wouldn't have gotten
 habits, motives, that wrong."
 etc. These are
 comments that go
 beyond merely
 stating an answer is
 incorrect (even if
 the child is named)
 to statements in
 which larger
 inferences are made.

REFERENCES

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. doi: 10.1037/0022-3514.51.6.1173

Beaman, R., & Wheldall, K. (2000). Teachers' use of approval and disapproval in the classroom. Educational Psychology, 20(4), 431-446. doi: 10.1080/713663753

Beets, M. W., Flay, B. R., Vuchinich, S., Snyder, F. J., Acock, A., Li, K.-K., ... Durlak, J. A. (2009). Use of a social and character development program to prevent substance use, violent behaviors, and sexual activity among elementary students in Hawai'i. American Journal of Public Health, 99(8), 1438-1445. doi:10.2105/AJPH.2008.142919

Birch, S., & Ladd, G. (1997). The teacher-child relationship and children's early school adjustment. Journal of School Psychology, 35, 61-80.

Broidy, L. M., Nagin, D. S., Tremblay, R. E., Bates, J. E., Brame, B., Dodge, K. A., ... Pettit, G. S. (2003). Developmental trajectories of childhood disruptive behaviors and adolescent delinquency: A six-site, cross-national study. Developmental Psychology, 39, 222-245. doi: 10.1037/0012-1649.39.2.222

Brophy, J., & Good, T. (1974). Teacher-student relationships: Causes and consequences. New York, NY: Holt, Rinehart, & Winston.

Burns, G. L., & Owen, S. M. (1990). Disruptive behaviors in the classroom: Initial standardization data on a new teacher rating scale. Journal of Abnormal Child Psychology, 18, 515-525. doi: 10.1007/BF00911105

Burns, G. L., Walsh, J. A., & Owen, S. M. (1995). Twelve-month stability of disruptive classroom behavior as measured by the Suner-Eyberg Student Behavior Inventory. Journal of Clinical Child Psychology, 24, 453-462. doi: 10.1207/s 15374424jccp2404_9

Caspi, A., Henry, B., McGee, R. M. T., & Silva, E (1995). Temperamental origins of child and adolescent behavior problems: From age three to fifteen. Child Development, 66, 55-68. doi: 10.2307/1131190

Childs, G., & McKay, M. (2001). Boys starting school disadvantaged: Implications from teachers' rating of behaviour and achievement in the first two years. British Journal of Educational Psychology, 71, 303-314. doi: 10.1348/000709901158532

Chow, V. T., & Kasari, C. (1999). Task-related interactions among teachers and exceptional, at-risk, and typical learners in inclusive classrooms. Remedial and Special Education, 20, 226-232.

Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology (3rd ed., Vol. 3, pp. 779-862). Hoboken, NJ: Wiley.

Evertson, C. M., & Smithey, M. W. (2000). Mentoring effects on proteges' classroom practice: An experimental field study. Journal of Educational Research, 93, 294-304.

Eyberg, S., & Pincus, D. (1999). Eyberg Child Behavior Inventory and Sutter-Eyberg Behavior Inventory: Professional manual. Odessa, FL: Psychological Assessment Resources.

Fritz, M. S., & MacKinnon, D. P. (2007). Required sample size to detect the mediated effect. Psychological Science, 18, 233-239. doi: 10.1111/j.1467-9280.2007.01882.x.

Good, T. L., & Brophy, J. E. (2008). Looking in classrooms (10th ed.) Boston, MA: Pearson/Merrill Education.

Gresham, F. M., Elliott, S. N., Cook, C. R., Vance, M. J., & Kettler, R. (2010). Cross-informant agreement for ratings for social skill and problem behavior ratings: An investigation of the Social Skills Improvement System-Rating Scales. Psychological Assessment, 22, 157-166. doi: 10.1037/a0018124.

Guerin, D., Gottfried, A. W., Oliver, P. H., & Thomas, C. W. (1994). Temperament and school functioning during early adolescence. Journal of Early Adolescence, 14, 200-225. doi:10.1177/027243169401400206

Henderson, H. A., & Fox, N. A. (1998). Inhibited and uninhibited children: Challenges in school settings. School Psychology Review, 27, 492-505.

Institute of Medicine. (1994). Reducing risks for mental disorders: Frontiers for preventative intervention research. Washington, DC: National Academy of Press.

Jones, S. M., & Dindia, K. (2004). A meta-analytic perspective on sex equity in the classroom. Review of Educational Research, 74(4), 443-471. doi: 10.3102/00346543074004443

Kagan, J., Snidman, N., & Arcus, D. M. (1992). Initial reactions to unfamiliarity. Current Directions in Psychological Science, 1, 171-174. doi: 10.1111/1467-9566.ep10770010

Kean, J. (1995). Teacher ethnotheories and child temperament: Impact on classroom interactions. Australian Journal of Early Education, 22, 13-18.

Kelly, A. (1988). Gender differences in teacher-pupil interactions: A meta-analytic review. Research in Education, 39, 1-23.

Keogh, B. K. (2003). Temperament in the classroom: Understanding individual differences. Baltimore, MD: Paul H. Brookes.

Kolko, D. J., & Kazdin, A. E. (1993). Emotional/behavioral problems in clinic and nonclinic children: Correspondence among child, parent and teacher reports. Journal of Child Psychology and Psychiatry, 34, 991-1006. doi:10.1111/j. 1469-7610.1993.tb01103.x

Lerner, J. V., Lerner, R. M., & Zabski, S. (1985). Temperament and elementary school children's actual and rated academic performance: A test of a "goodness-of-fit" model. Journal of Child Psychology and Psychiatry, 26, 125-136. doi: 10.1111/j.1469-7610.1985.th01633.x

Lyons-Thomas, J., & McClowry, S. G. (2012). An examination of the construct validity and reliability of the Teacher School-Age Temperament Inventory. Journal of Classroom Interaction, 47(2), 25-32.

McClowry, S. G. (1995). The development of the School-Age Temperament Inventory. Merrill-Palmer Quarterly, 41, 27-285.

McClowry, S. G. (2002). The temperament profiles of school-age children. Journal of Pediatric Nursing, 17, 3-10.

McClowry, S. G., & Galehouse, P. (2002). Planning a temperament-based parenting program for inner-city families. Journal of Child and Adolescent Psychiatric Nursing, 15, 97-105. doi: 10.1111/j.1744-6171.2002.tb00332.x

McClowry, S. G., Giangrande, S. K., Tommasini, N. R., Clinton, W., Foreman, N. S., Lynch, K., & Ferketich, S. L. (1994). The effects of child temperament, maternal characteristics, and family circumstances on the maladjustment of school-age children. Research in Nursing and Health, 17, 25-35. doi: 10.1002/nur.4770170105

McClowry, S. G., Snow, D. L., Tamis-LeMonda, C. S., & Rodriguez, E. T. (2010). Testing the efficacy of INSIGHTS on student disruptive behavior, classroom management, and student competence in inner city primary grades. School Mental Health, 2, 23-35. doi: 10.1007/s12310-009-9023-8.

Meehan, B. T., Hughes, J. N., & Cavell, T. A. (2003). Teacher-student relationships as compensatory resources for aggressive children. Child Development, 74(4), 1145-1157.

Montague, M., & Rinaldi, C. (2001). Classroom dynamics and children at risk: A followup. Learning Disability Quarterly, 24, 75-83. doi: 10.2307/1511063

Nelson, B., Martin, R. P., Hodge, S., Havill, V., & Kamphaus, R. (1999). Modeling prediction of elementary school adjustment from preschool temperament. Personality and Individual Differences, 26, 687-700. doi: 10.1016/S0191-8869(98)00174-3

Nelson, J. R., & Roberts, M. L (2000). Ongoing reciprocal teacher-student interactions involving disruptive behaviors in general education classrooms. Journal of Emotional and Behavioral Disorders, 8, 27-37. doi: 10.1177/106342660000800104

O'Connor, E. (2010). Teacher-child relationships as dynamic systems. Journal of School Psychology, 48(3), 187-218. doi: 10.1016/j.jsp.2010.01.001

Pianta, R. C., Steinberg, M. S., & Rollins, L. B. (1995). The first two years of school: Teacher-child relationships and deflections in children's classroom adjustment. Development and Psychopathology, 7, 295-312.

Prior, M., Sanson, A., Smart, D., & Oberklaid, F. (2000). Pathways from infancy to adolescence: Australian temperament project 1983-2000. Melbourne, Australia: Australian Institute of Family Studies.

Pullis, M., & Cadwell, J. (1982). The influence of children's temperament characteristics on teachers' decision strategies. American Educational Research Journal, 19, 165-181. doi: 10.2307/1162563

Rescorla, L., Achenbach, T. M., Almqvist, F., Bird, H., Broberg, A., Dobrean, A., ... Verhulst, F. (2007). Epidemiological comparisons of problems and positive qualities reported by adolescents in 24 countries. Journal of Consulting and Clinical Psychology, 75, 351-358. doi: 10.1037/0022-006X.75.2.351

Rothbart, M. K., & Bates, J. E. (2006). Temperament. In W. Damon & R. Lerner (Eds.), Handbook of child psychology (6th ed., pp. 99-166). New York, NY: Wiley.

Rudasill, K. M., & Rimm-Kaufman, S. E. (2009). Teacher-child relationship quality: The roles of child temperament and teacher-child interactions. Early Childhood Research Quarterly, 24, 107-120. doi: 10.1016/j.ecresq.2008.12.003

Sanson, A., Hemphill, S. A., & Smart, D. (2004). Connections between temperament and social development: A review. Social Development, 13(1), 142-170. doi: 10. 1046/j.1467-9507.2004.00261

Smart, D., Vassallo, S., Sanson, A., Richardson, N., Dussuyer, I., & McKendry, B. (2003). Patterns and precursors of adolescent antisocial behavior: type, resiliency, and environmental influences. Melbourne, Australia: Australian Institute of Family Studies.

SPSS, Inc. (2001). The SPSS TwoStep cluster component: A sealable component enabling more efficient customer segmentation. Chicago, IL: Author.

Stichter, J. P., Stormont, M., Lewis, T. J., & Schultz, T. (2009). Rates of specific antecedent instructional practices and differences between Title I and non-Title I schools. Journal of Behavioral Education, 18(4), 331-344. doi: 10.1007/s 10864-009-9094-5

Sutherland, K. S. (2000). Promoting positive interactions between teachers and students with emotional/behavioral disorders. Preventing School Failure, 44, 110-115. doi: 10.1080/10459880009599792

Tamis-LeMonda, C., Briggs, R., & Carlson, A. (2008). Primary Classroom Observation Scale (PCOS). Unpublished manuscript.

Theodoridis, S., & Koutroumbas, K. (1999). Pattern recognition. San Diego, CA: Academic Press.

Van de Werfhorst, H. (1986). Temperament and teacher-child interactions. In G. A. Kohnstamm (Ed.), Temperament discussed: Temperament and development in infancy and childhood (pp. 141-147). Leiden, Netherlands: University of Leiden.

Sandra Graham McClowry

New York University, New York, New York

Eileen T. Rodriguez

Mathmatica, Princeton, New Jersey

Catherine S. Tamis-LeMonda and Mark E. Spellmann

New York University, New York, New York

Allyson Carlson

Port Jefferson School District, New York, New York

David L. Snow

Yale University School of Medicine, New Haven, Connecticut

Submitted April 19, 2011; accepted August 25, 2011.

This article was supported by funding from the National Institute of Nursing Research (R01NR004781) and the Institute for Education Sciences (R305A080512).

Address correspondence to Sandee McClowry, Ph.D., RN, FAAN, New York University, 246 Greene Street, 507W, New York, NY 10003. E-mail: sandee.mcclowry@nyu.edu

TABLE 1
Descriptive Statistics of Temperament Dimensions, Disruptive
Behavior, Difficulty Managing Disruptive Behavior, and Teacher
Feedback

 Cronbach's
 M SD Range [alpha]

Temperament Dimensions
 Negative reactivity 2.46 0.91 1.00-5.00 .94
 Task persistence 3.31 0.87 1.11-5.00 .93
 Activity 2.46 0.97 1.00-5.00 .90
 Withdrawal 2.61 0.64 1.00-4.40 .71
Student Disruptive Behavior
 Overt aggression toward 2.92 1.59 1.00-6.91 .96
 others
 Emotional-oppositional 2.37 1.49 1.00-6.90 .96
 behavior
 Attentional difficulties 3.15 1.50 1.00-6.56 .94
 Covert disruptive behavior 2.00 1.17 1.00-7.00 .84
Teacher Difficulty Managing
 Disruptive Behavior
 Overt aggression toward 2.69 3.71 0.00-11.00 .93
 others
 Emotional-oppositional 1.59 2.90 0.00-10.00 .94
 behavior
 Attentional difficulties 2.43 3.08 0.00-9.00 .91
 Covert disruptive behavior 0.50 0.99 0.00-4.00 .74
Observed Student Disruption
 and Teacher Feedback
 Student disruption 1.77 5.33 0.00-118.00 --
 Positive teacher feedback 0.31 0.77 0.00-4.00 --
 Negative teacher feedback 1.81 2.77 0.00-18.00 --

TABLE 2
Intercorrelations Among Student Temperament Dimensions, Disruptive
Behavior, Difficulty Managing Disruptive Behavior, and Teacher
Feedback

 2 3 4

Temperament Dimensions
 1. Negative reactivity -.51 *** .59 *** .02
 2. Task persistence -.54 *** -.06
 3. Activity -.18 *
 4. Withdrawal
Student Disruptive Behavior
 5. Overt aggression toward others
 6. Emotional-oppositional behavior
 7. Attentional difficulties
 8. Covert disruptive behavior
Teacher Difficulty Managing Behavior
 9. Overt aggression toward others
 10. Emotional-oppositional behavior
 11. Attentional difficulties
 12. Covert disruptive behavior
Observed Student Disruption and
Teacher Feedback
 13. Student disruption
 14. Positive teacher feedback
 15. Negative teacher feedback

 5 6 7

Temperament Dimensions
 1. Negative reactivity .69 *** .82 *** .66 ***
 2. Task persistence -.60 *** -.42 *** -.79 ***
 3. Activity .77 *** .58 *** .64 ***
 4. Withdrawal -.16 * -.04 .04
Student Disruptive Behavior
 5. Overt aggression toward others .80 *** .83 ***
 6. Emotional-oppositional behavior .68 ***
 7. Attentional difficulties
 8. Covert disruptive behavior
Teacher Difficulty Managing Behavior
 9. Overt aggression toward others
 10. Emotional-oppositional behavior
 11. Attentional difficulties
 12. Covert disruptive behavior
Observed Student Disruption and
Teacher Feedback
 13. Student disruption
 14. Positive teacher feedback
 15. Negative teacher feedback

 8 9 10

Temperament Dimensions
 1. Negative reactivity .51 *** .49 *** .63 ***
 2. Task persistence -.48 *** -.43 *** -.26 ***
 3. Activity .48 *** .52 *** .38 ***
 4. Withdrawal .04 -.12 -.07
Student Disruptive Behavior
 5. Overt aggression toward others .68 *** .80 *** .64 ***
 6. Emotional-oppositional behavior .63 *** .61 *** .82 ***
 7. Attentional difficulties .62 *** .64 *** .52 ***
 8. Covert disruptive behavior .49 *** .50 ***
Teacher Difficulty Managing Behavior
 9. Overt aggression toward others .73 ***
 10. Emotional-oppositional behavior
 11. Attentional difficulties
 12. Covert disruptive behavior
Observed Student Disruption and
Teacher Feedback
 13. Student disruption
 14. Positive teacher feedback
 15. Negative teacher feedback

 11 12 13

Temperament Dimensions
 1. Negative reactivity .45 *** .39 *** .26 ***
 2. Task persistence -.57 *** -.33 *** -.17 *
 3. Activity .42 *** .31 ** .25 **
 4. Withdrawal -.01 -.03 -.11
Student Disruptive Behavior
 5. Overt aggression toward others .66 *** .56 *** .29 ***
 6. Emotional-oppositional behavior .54 *** .55 *** .32 ***
 7. Attentional difficulties .77 *** .46 *** .21 **
 8. Covert disruptive behavior .43 *** .73 *** .22 **
Teacher Difficulty Managing Behavior
 9. Overt aggression toward others .79 *** .60 *** .15
 10. Emotional-oppositional behavior .63 *** .69 *** .19 *
 11. Attentional difficulties .55 *** .12
 12. Covert disruptive behavior .20 **
Observed Student Disruption and
Teacher Feedback
 13. Student disruption
 14. Positive teacher feedback
 15. Negative teacher feedback

 14 15

Temperament Dimensions
 1. Negative reactivity .01 .35 ***
 2. Task persistence .06 -.23 **
 3. Activity .06 .30 ***
 4. Withdrawal .03 -.13
Student Disruptive Behavior
 5. Overt aggression toward others .03 .29 ***
 6. Emotional-oppositional behavior .02 .28 ***
 7. Attentional difficulties -.04 .29 ***
 8. Covert disruptive behavior -.02 .18 *
Teacher Difficulty Managing Behavior
 9. Overt aggression toward others .05 .24 **
 10. Emotional-oppositional behavior .02 .30 ***
 11. Attentional difficulties .04 .28 ***
 12. Covert disruptive behavior .00 .25 **
Observed Student Disruption and
Teacher Feedback
 13. Student disruption -.03 .60 ***
 14. Positive teacher feedback -.06
 15. Negative teacher feedback

* p < .05. ** p < .01. *** p < .001.

TABLE 3
Analysis of Variance: Temperament Dimensions by Student Temperament
Profile

 Student Temperament Profile Means (SD)

Temperament Industrious Intermediate High Maintenance
Dimension (n = 35) (n = 61) (n = 56)

Negative reactivity 1.52 (0.46) a 2.17 (0.38) b 3.35 (0.71)
Task persistence 4.38 (0.46) a 3.25 (0.58) b 2.71 (0.69) c
Activity 1.43 (0.37) a 2.29 (0.53) b 3.30 (0.86) c

Temperament
Dimension F

Negative reactivity 137.14 ***
Task persistence 83.99 ***
Activity 92.95 ***

Note. Means in the same row with different subscripts are
significantly different based on Bonferroni post-hoc comparisons.

*** p < .001.

TABLE 4
Analysis of Variance: Student Disruptive Behavior and Teacher
Difficulty Managing Disruptive Behavior by Temperament Profile

 Student Temperament Profile

 Industrious Intermediate
Outcome (n = 35) (n = 61)

Student Disruptive Behavior
 Overt aggression toward others 1.41 (0.45) a 2.44 (0.88) b
 Emotional-oppositional behavior 1.31 (0.51) a 1.75 (0.69) a
 Attentional difficulties 1.51 (0.53) a 2.90 (0.98) b
 Covert disruptive behavior 1.21 (0.32) a 1.80 (0.89) b
Teacher Difficulty Managing
 Disruptive Behavior
 Overt aggression toward others 0.34 (1.08) a 1.72 (2.76) a
 Emotional-oppositional behavior 0.29 (1.07) a 0.52 (1.73) a
 Attentional difficulties 0.17 (0.57) a 1.93 (2.48) b
 Covert disruptive behavior 0.06 (0.24) a 0.39 (0.84) a

 Student
 Temperament
 Profile

 High Maintenance F
Outcome (n = 56)

Student Disruptive Behavior
 Overt aggression toward others 4.39 (1.43) c 96.40 ***
 Emotional-oppositional behavior 3.71 (1.56) b 70.67 ***
 Attentional difficulties 4.44 (1.22) c 95.31 ***
 Covert disruptive behavior 2.72 (1.37) c 26.14 ***
Teacher Difficulty Managing
 Disruptive Behavior
 Overt aggression toward others 5.21 (4.22) b 30.64 ***
 Emotional-oppositional behavior 3.55 (3.59) b 27.67 ***
 Attentional difficulties 4.38 (3.43) c 29.44 ***
 Covert disruptive behavior 0.89 (1.26) b 9.16 ***

Note. Means in the same row with different subscripts are
significantly different based on Bonferroni post-hoc comparisons.

*** p < .001.

TABLE 5
Test of Mediation Using Hierarchical Multiple Regression: Student
Gender and Temperament in Relation to Teacher Reported Student
Disruptive Behavior and Difficulty Managing Behavior

 Overt Aggression Toward
 Others

 Model 1 Model 2

Disruptive Behavior
 Gender = Male -0.95 (.25), -0.29 (.18),
 -0.30 *** -0.09
 Temperament profile
 Intermediate 0.96 (.23),
 0.30 ***
 High maintenance 2.87 (.24),
 0.87 ***
 [R.sup.2] Total .09 .57
 F 14.55 *** 65.78 ***
Teacher Difficulty Managing
Behavior
 Gender = Male -1.13 (.60, 0.00 (.54),
 -0.15 0.00
 Temperament profile
 Intermediate 1.38 (.68),
 0.18 *
 High maintenance 4.87 (.71),
 0.63 ***
 [R.sup.2] Total .02 .29
 F 3.52 20.29 ***

 Emotional-Oppositional
 Behavior

 Model 1 Model 2

Disruptive Behavior
 Gender = Male -0.66 (.24), -0.12 (.18),
 -0.22 ** -0.04
 Temperament profile
 Intermediate 0.41 (.23),
 0.13
 High maintenance 2.35 (.24),
 0.76 ***
 [R.sup.2] Total .05 .49
 F 7.68 ** 47.06 ***
Teacher Difficulty Managing
Behavior
 Gender = Male -0.81 (.47), -0.06 (.43),
 -0.14 -0.01
 Temperament profile
 Intermediate 0.23 (.54),
 0.04
 High maintenance 3.25 (.57),
 0.54 ***
 [R.sup.2] Total .02 .27
 F 2.93 18.33 ***

 Attention Difficulties

 Model 1 Model 2

Disruptive Behavior
 Gender = Male -0.83 (.24), -0.17 (.17),
 -0.27 *** -0.06
 Temperament profile
 Intermediate 1.35 (.22),
 0.44 ***
 High maintenance 2.86 (.23),
 0.92 ***
 [R.sup.2] Total .08 .56
 F 12.27 *** 63.85 ***
Teacher Difficulty Managing
Behavior
 Gender = Male -0.87 (.50), 0.11 (.45),
 -0.14 0.02
 Temperament profile
 Intermediate 1.79 (.57),
 0.29 **
 High maintenance 4.25 (.60),
 0.67 ***
 [R.sup.2] Total .02 .28
 F 3.04 19.52 ***

 Covert Disruptive Behavior

 Model 1 Model 2

Disruptive Behavior
 Gender = Male -0.63 (.19), -0.31 (.17),
 -.27 *** -0.13
 Temperament profile
 Intermediate 0.51 (.22),
 0.22 *
 High maintenance 1.39 (.23),
 0.58 ***
 [R.sup.2] Total .07 .28
 F 11.51 *** 18.73 ***
Teacher Difficulty Managing
Behavior
 Gender = Male -0.33 (.16), -0.16 (.16),
 -0.17 * -0.08
 Temperament profile
 Intermediate 0.30 (.20),
 0.15
 High maintenance 0.77 (.21),
 0.38 ***
 [R.sup.2] Total .03 .12
 F 4.35 * 6.41 ***

Note. Values are unstandardized B weights, standard error of B (in
parenthesis), and [beta] coefficients from the final regression
equation. Industrious is the excluded reference category.

* p <. 05. ** p < .01. *** p < .001.

TABLE 6
Test of Mediation Using Hierarchical Logistic Regression: Student
Gender and Temperament in Relation to Observed Disruptive Behavior
and Teacher Feedback

 Observed Student
 Disruption

 Model 1 Model 2
 Odds Ratio Odds Ratio

Gender = Male 2.05 * 1.58
Temperament profile
Intermediate 4.02 *
High maintenance 4.59 **
[chi square] (df) 4.37 (1) * 13.64 (3) **
Nagelkerke [R.sup.2] .039 .118

 Observed Positive Teacher
 Feedback

 Model 1 Model 2
 Odds Ratio Odds Ratio

Gender = Male 3.57 ** 3.72 **
Temperament profile
Intermediate 0.55
High maintenance 0.89
[chi square] (df) 8.60 (1) ** 10.06 (3) *
Nagelkerke [R.sup.2] .086 .100

 Observed Negative Teacher
 Feedback

 Model 1 Model 2
 Odds Ratio Odds Ratio

Gender = Male 1.62 1.18
Temperament profile
Intermediate 3.98 **
High maintenance 4.57 **
[chi square] (df) 2.16 (1) 14.32 (3) **
Nagelkerke [R.sup.2] .019 .120

Note. Industrious is the excluded reference category.

* p < .05. ** p < .01.