Relationships between students and their teachers are unique and varied, and there is clear evidence that strong classroom relationships can be associated with positive schooling outcomes. In this research we examine the quality of student–teacher relationships as perceived by young adolescent male students and their teachers, and consider whether student perceptions and teacher perceptions match. Understanding some of the bases for both agreement and disagreement between student and teacher perceptions of their relationships will add to current understanding of the complexity underlying the assessment of the quality of student–teacher relationships.
Studies of student–teacher relationships have regularly over-relied on teacher reporting about students to establish relationship quality (Mercer & DeRosier, Reference Mercer and DeRosier2010). However, some studies have compared data capturing student perceptions of their relationships to teacher ratings of student attributes. Murray and Greenberg (Reference Murray and Greenberg2000) reported that primary school students who perceived feeling a close bond with their teacher had fewer teacher-reported problem behaviours and were reported as more socially competent. Wentzel (Reference Wentzel2002) demonstrated student–teacher relationships that Year 6 students perceived more positively were linked to increased motivation, academic achievement and prosocial behaviours.
Adolescent student–teacher relationships have, in general, been researched less widely. However, some studies have indicated that when students perceive positive and caring student–teacher relationships, the outcomes are likely to be increased academic interest and motivation (Wentzel, Reference Wentzel1997, Reference Wentzel1998). In a large Australian qualitative study, Trent and Slade (Reference Trent and Slade2001) reported that adolescent boys’ descriptions of their attitudes towards school were influenced by the perceived quality of their relationships with their teachers, with poor relationships often detrimental to their school engagement and achievement.
Student–teacher relationships perceived by students to be characterised by high conflict and antisocial behaviours consistently predict poor academic achievement and school engagement (Lee, Reference Lee2007; Pomeroy, Reference Pomeroy1999). The same pattern is reported in studies where researcher ratings have characterised student–teacher relationships as high conflict (Burchinal, Peisner-Feinberg, Pianta, & Howes, Reference Burchinal, Peisner-Feinberg, Pianta and Howes2002; Wentzel, Reference Wentzel1993). Nevertheless, caring and supportive relationships can mitigate the effects of disadvantaged backgrounds. For example, Murray and Malmgren (Reference Murray and Malmgren2005) found that an intervention with teachers to foster caring student–teacher relationships with disadvantaged students in a high-poverty urban US secondary school improved academic performance.
Major reviews of the literature (Davis, Reference Davis2003; Pianta, Hamre, & Stuhlman, Reference Pianta, Hamre, Stuhlman, Reynolds and Miller2003) indicate that overall, student–teacher relationships that are perceived positively (by one or both parties) predict good academic and behavioural outcomes for middle school students. However, few studies have directly compared student and teacher perceptions of the same relationships. When they are reported, findings have been inconsistent. In a study with kindergarten children, low correlations were found between student and teacher ratings of support, closeness and conflict (Murray, Murray, & Waas, Reference Murray, Murray and Waas2008), whereas in a study with at-risk, urban youth, there were indications of some concordance between teacher and student perceptions for the positive dimensions of relationships (Murray & Zvoch, Reference Murray and Zvoch2011). The diversity of findings suggest that what is important for student–teacher relationships may depend on whether the perspective of the students or the teachers is considered. Thus there is a need for a better understanding of where student and teacher perceptions agree and differ.
In educational circles in the developed world, the middle years of schooling (students in Years 5 to 9 who are aged between 10 and 16 years) have been recognised as a period of high risk for student disengagement from school, with associated decreasing standards of academic achievement (Victorian Government, 2003). Widely reported evidence of declining male students’ academic results during these years indicates that boys may be at a particular disadvantage (Martin, Reference Martin2002; O'Doherty, Reference O'Doherty1994). In Australia, the Victorian State Government's ‘Connecting through the Middle Years’ (2003) project demonstrated that nurturing what both students and teachers considered a close and supportive one-to-one relationship between adolescent students and a staff member strengthened social connectedness to school and encouraged positive attitudes towards learning. Hence, there is a need for greater understanding of how students and their teachers each perceive the quality of their relationships in order to facilitate such interventions.
Gender and Classroom Interactions
The literature indicates that gender plays a significant part in patterns of classroom interaction. In a classroom observational study designed to investigate the number of interactions between teachers and secondary school students, Duffy, Warren, and Walsh (Reference Duffy, Warren and Walsh2001) observed that female mathematics teachers and both male and female literature teachers interacted more frequently with boys than girls. Other researchers have suggested that a higher frequency of interactions between teachers and boys may occur because boys often exhibit more antisocial behaviour and thus attract more teacher attention (Pianta et al., Reference Pianta, Hamre, Stuhlman, Reynolds and Miller2003). Morgan (Reference Morgan1998) found that teachers’ interactions with adolescent boys involved more overt control efforts, such as gaining adherence to directions. Conversely, girls were more likely than boys to receive only positive competence feedback. Similarly, other researchers have observed that male students receive more criticism and negative feedback about conduct in both secondary school (Duffy, Warren, & Walsh, Reference Duffy, Warren and Walsh2001; Ilatov, Shamai, Hertz-Lazarovitz, & Mayer-Young, Reference Ilatov, Shamai, Hertz-Lazarovitz and Mayer-Young1998) and primary school (Foote, Reference Foote2002). These differences in gender patterns of interaction are important because positive teacher feedback has been shown to have positive effects on student learning (Swinson & Knight, Reference Swinson and Knight2007) and negative feedback is one of the most consistent predictors of poor academic performance and problem social behaviour (Wentzel, Reference Wentzel2002). This trend warrants closer attention, and the presence of gender differences in adolescent student–teacher interactions suggests that the classroom relationships male and female students have with their teachers requires separate examination. Hence, in the current investigation, it is the quality of student–teacher relationships as perceived by young adolescent male students and their teachers that is the focus of attention.
The Current Study
While previous research indicates that stronger relationships between adolescents and teachers are likely to result in more positive outcomes, more research that sheds light on what makes these relationships ‘strong’ is required. To address this limitation in the research literature, the current study gathered data on how adolescent boys and their teachers in one school perceive their relationships. The analyses employed aimed to establish which elements are associated with stronger relationships from the perspectives of both the boys and their teachers.
The central issue in considering these two perspectives on student–teacher relationships was to establish whether perceptions matched. As indicated in this review, studies that directly compare student and teacher perceptions of the same relationships have seldom been reported and when they are reported, findings have been inconsistent. Within a sample of adolescent male students from the same school we expected students to report a number of different types of student–teacher relationships. Simultaneously, we expected to find different types of student–teacher relationships within the group when the informant was the teacher. The data were explored to identify which elements of student-teacher relationships were associated with strong student–teacher relationships from the perspective of the male students and from the perspective of the teachers.
Method
PARTICIPANTS
Nine teachers and 70 boys from a secondary school in suburban Melbourne, Australia participated in the study. In most Australian secondary schools, students have a ‘home group’ to which a specific ‘home group teacher’ is assigned. At the participating school, teachers met their home group at the beginning of each day and also taught the group for two core subjects (two of Maths, English and Social Studies). Only home group teachers were invited to join the study and three declined. Six female and three male teachers participated. This gender composition reflected the higher proportion of female teachers in Australian schools (Australian Bureau of Statistics, 2008).
Only boys whose home group teacher had consented to participate were invited to take part and thus 117 boys in six Year 8 classes, and three Year 9 classes were invited to participate. The sample of 70 reflected a response rate of approximately 60%. Boys were aged between 13 and 15 years old with a mean age of 13 years and 10 months.
MEASURES
Subscales of the ‘my english class’ survey
English is a core subject with most contact hours between students and teachers in the middle years and thus a good opportunity for establishing student–teacher relationships. Hence, to measure boys’ perceptions of their relationships with their teachers, this study used four subscales of the ‘My English Class’ Survey; a questionnaire developed and validated with a large sample of Australian students in the middle years of secondary school (Frydenberg et al., Reference Frydenberg, Freeman, Care, Tobin, Begg and Chan2008). The questionnaire captured student reflections on their level of engagement, wellbeing and the student–teacher relationship. While the questionnaire was developed with English teachers, it was designed with a view to adaptation for more generalised use. The subscales used in this study concentrate on relationships, not English instruction, and can be used with teachers of other core subjects. The four subscales are: Attitudes towards English Teacher, Student Perceptions of English Teacher's Attitudes Towards Them, Student Perceptions of English Teacher's Helpfulness, Student Perceptions of English Teacher's Positive Feedback.
At the target school, the four My English Class subscales were used to measure boys’ perceptions of their relationship with their home group teacher. This teacher taught the boys for two core subject (two of English, Maths and Social Studies). Thirty items were administered; however, following the results of a principal components analysis in the Frydenberg study, only the 17 items that loaded on the student–teacher relationship factor were used in the current analyses. Cronbach's alpha reliability coefficients for the four subscales were: Attitudes towards English Teacher (.86), Student Perceptions of English Teacher's Attitudes Towards Them (.92), Student Perceptions of English Teacher's Helpfulness (.90), Student Perceptions of English Teacher's Positive Feedback (.86) (Frydenberg et al., Reference Frydenberg, Freeman, Care, Tobin, Begg and Chan2008).
Teacher student relationship inventory
Most quantitative studies that have measured perceptions of relationships have used subdimensions of larger social support scales or items extracted from other scales (Ang, Reference Ang2005). In a review of such research, Ang concluded that the Student–Teacher Relationship Scale (STRS: Pianta, Reference Pianta2001) was the only valid self-report dedicated to measuring teachers’ perception of their relationships with their students (Ang, Reference Ang2005). However, the STRS has only been validated in the United States with teachers describing relationships with children from kindergarten to Year 3. Student–teacher relationships with adolescents are likely to be different from those of younger students as adolescents rely on teachers for emotional support in different ways than do younger students (Wentzel, Reference Wentzel1996). For example, the Dependency dimension of the STRS may not be relevant to adolescents as they become developmentally more aware of their own autonomy (Shaffer, Reference Shaffer2002). Ang, therefore, developed the Teacher–Student Relationship Inventory (TSRI) for use with teachers of students from Year 4 through to middle and junior secondary school.
The current study used the TSRI (Ang, Reference Ang2005) to measure teachers’ perceptions of their relationships with participating students. Consistent with the literature on student–teacher relationships (Davis, Reference Davis2003; Pianta, Reference Pianta2001), factor analysis has confirmed three factors: Satisfaction, Instrumental Help and Conflict. The Satisfaction dimension indicates the degree to which the teacher experiences a positive and satisfactory relationship with the student. An example item is ‘I enjoy having this student in my class’. The Instrumental Help dimension reflects the teacher's perception of how willing the student is to turn to them for advice, sympathy or help; for example: ‘If the student has a problem at home, he/she is likely to ask for my help’. The final dimension, Conflict, reflects the teacher's perception of how unpleasant and conflictual their relationship with the student may be. An example item states: ‘This student frustrates me more than most other students in my class’. Teachers respond on a 5-point Likert scale from 1 = Almost never true, to 5 = Almost always true. Ang reported reliability coefficients (alphas) of .72 for Instrumental Help, .88 for Satisfaction, and .90 for Conflict. Satisfactory construct validity has been reported when scores on the TSRI were compared with responses to the Aggression Questionnaire (Buss & Warren, Reference Buss and Warren2000) with Conflict scale scores positively correlated with aggression (r = .21) and Satisfaction scale scores negatively correlated with aggression (r = −.20).
PROCEDURE
Ethical clearance and permissions were obtained before the study was presented to all staff at a weekly staff meeting. In addition, the researcher spoke with all eligible teachers individually. The researcher then regularly visited the classes of eligible students over 15 weeks. The regular presence of the researcher in the school provided opportunities for questions to be answered, and for the researcher to become a familiar figure for potential participants. It also provided opportunity to reissue lost permission forms. As the researcher became more familiar to the students, more consent forms were returned.
Participating boys completed the subscales of the ‘My English Class’ Survey in class-time, in groups of between two and six. Instructions were read aloud by the researcher. Teachers were given one TSRI form to complete for each participating student. Each questionnaire had a cover sheet containing the student's name. After matching the student and teacher responses, cover sheets were removed and destroyed to protect confidentiality.
Separate cluster analyses were used to identify groups of students who reported similar relationship profiles with their teacher, and sets of teacher perceptions identifying similar profiles of relationship with their students (Hair, Tatham, Anderson, & Black, Reference Hair, Tatham, Anderson and Black1998). The complete linkage (or furthest-neighbour) procedure was used for the cluster algorithm and the most commonly used squared Euclidean distance was employed as the distance measure. These methods eliminate the ‘chaining’ problem that can occur in single linkage cluster analysis, and avoids the bias of Ward's method toward clusters with approximately the same number of cases (Hair et al.).
Results
Initial data screening resulted in the exclusion of one case where the participant was observed to circle responses without reading any of the questions. Missing values for two cases on the My English Class survey (student data) and five cases on the TSRI (teacher data) resulted in the exclusion of these cases from the cluster analyses. Screening for multicolinearity resulted in the exclusion of responses for item 11 of the TSRI in the analysis which shared a high correlation with item 7 (r = .86) and item 8 (r = .91). After its removal, acceptable tolerance was found for the remaining 13 items. Table 1 displays response frequencies for boys on the 17 items of the My English Class subscales.
TABLE 1 Descriptive Statistics for the Student-Reported Data (n = 70)
Overall, responses from boys indicated that they generally perceived their student–teacher relationships positively. This was particularly true of items on the Attitudes of Teachers to Boys scale and items on the Positive Feedback scale suggesting the boys generally perceived that their teachers acted positively towards them. The means were towards the positive end of the Likert scale falling between the Sometimes and Always categories. However, a small proportion of responses indicated some negative perceptions of student–teacher relationships.
Overall, teachers perceived their relationships with their male students positively. They were satisfied with their relationships; believed students came to them for help, and perceived student–teacher relationships to be generally free of conflict. Table 2 indicates that there was little variability among the teachers in their responses, particularly on the Conflict scale items, where the vast majority of responses were skewed towards disagreement with the negatively phrased items.
TABLE 2 Descriptive Statistics for the Teacher-Reported Data (n = 70)
However, as with the boys’ perceptions, a small number of teachers gave responses that contrasted with this general picture. In two cases teachers indicated a student Almost always was frustrating and five responses indicated teachers would Almost never miss a student if they were absent.
WHAT PROFILES OF RELATIONSHIPS WERE PERCEIVED?
Separate cluster analyses were run for the boys’ responses on the ‘My English Class’ subscales and for the teachers’ responses on the TSRI. For both cluster analyses, examination of the dendrogram and homogeneity of the clusters indicated that two-cluster solutions were optimal. The two-cluster solution summarising profiles of each boy's perceptions of their relationship with their home group teacher is presented in Figure 1.
FIGURE 1 Mean item scores for the two clusters identified in boys’ perceptions of their relationship with their home group teacher
Figure 1 shows that one cluster of boys (n = 49) indicated in their responses that they had a relatively strong positive relationship with their home group teacher across all four areas of teacher perceptions; Attitudes to Teacher, Attitudes of Teacher to Boys, Teacher Helpfulness, and Positive Feedback. Therefore this group of students were characterised as having ‘strong’ student–teacher relationships. For all of the items, scores were higher than for the second cluster (n = 18). While care should be taken in using multiple t tests with the same population, separate t tests were used here only to check the significance of differences between the item responses for the two profiles. All differences were significant, indicating adequate clustering. The mean item ratings for boys in the smaller second cluster were mainly in the Seldom to Sometimes range and these relationships were characterised as ‘moderate’ or moderately positive perceptions of their student–teacher relationships. The least positive responses within these profiles were for the Teacher Helpfulness items. Overall, these boys perceived their relationships with their home group teacher positively. The analysis did not identify a group that could be considered to have ‘poor’ student–teacher relationships.
The results for the cluster analysis of teacher perceptions of their relationships with the students are presented in Figure 2.
FIGURE 2 Mean item scores for the two clusters identified in teachers’ responses of their relationships with participating students
Responses to the Satisfaction items were generally very high with responses skewed towards viewing all relationships positively with mean responses falling between Sometimes and Almost always true. The Instrumental Help item responses were lower and around the midpoint of the scale, while the responses to the Conflict items, apart from item 4, were very low. On the Conflict scale items 7 and 8 displayed evidence of positively skewed distributions with positive kurtosis. Teachers overwhelmingly responded negatively to items indicating conflict: Almost never on item 7 ‘I cannot wait for this year to be over so that I will not need to teach this student next year’, and Almost never on item 8 ‘If this student is absent, I feel relieved’. This pattern suggests that teachers were unlikely to report conflict in relationships.
Figure 2 shows that teachers rated their relationships with one cluster of boys (n = 37) significantly higher than the other (n = 27) for all but three items. Again, one cluster was characterised as indicative of strong student–teacher relationships and the other of moderate student–teacher relationships. As with the data from students there was no group identified as having poor student–teacher relationships. Unlike the student profiles, statistically significant differences between clusters were not observed for all items. This was most pronounced on the conflict scale where two of the three items showed no difference in teacher perceptions of the boys from the two clusters.
DID THE PERCEPTIONS OF BOYS AND TEACHERS MATCH?
A key question posed in this research concerns the synchrony between student and teacher perceptions of their relationships. To test whether membership of the strong and moderate clusters identified from student and from teacher perceptions matched, we looked to identify the level of contingency between clusters identified from the two separate perspectives. A chi-square test for independence was conducted to determine whether there was a significant contingency between membership of the two strong clusters and the two moderate clusters. A 2 × 2 chi-square test for independence (with Yates Continuity Correction) indicated that the pattern of contingency was no different to what was expected under the independence hypothesis, χ2 (1, 63) = .015, p = .90. The contingency table showing the pattern both of observed frequencies and the expected frequencies under the independence hypothesis is presented below (Table 3).
TABLE 3 Contingency Between Membership of Strong and Moderate Clusters From the Perspective of the Students and the Teachers
Table 3 indicates that slightly more than half the boys (n = 35 or 56%) were identified to be in the cluster that represented matching student and teacher perceptions of their relationship. Equally this means that for just under half (44%) of the boys, the perceptions of boys and teachers did not match. It is informative to explore the differences between the perceptions of these unmatched cases to identify trends that may explain the discrepancy in perceptions. Therefore, comparisons were made between the mean item responses for the matched and unmatched groups. Table 4 presents the mean item scores for the students who were identified as having strong student–teacher relationships from the perspective of both student and teacher compared with the mean item scores for students identified by themselves as having moderate student–teacher relationships but where teachers’ responses indicated strong student–teacher relationships. Small numbers precluded application of tests of significance and so the comparison is made in terms of effect size (difference between the means expressed in SD units).
TABLE 4 Item Comparisons of Relationships Identified as Strong by Both Boys and Teachers (n = 27) With Relationships Identified as Strong by Teachers but Moderate by Boys (n = 9) (Student Data)
Note: 1Difference between the means expressed in SD units.
AT = Attitudes to Teacher, ATB = Attitudes of Teachers to Boys, TH = Teacher Helpfulness, PF = Positive Feedback.
As can be seen in Table 4, large differences were observed for items that related to boys’ feelings. The largest discrepancies were for items 16 ‘My teacher tries to understand how I feel’ and 17 ‘My teacher cares about my feelings’, where boys’ responses for the unmatched group were more likely to be in the lowest rating categories. Similarly, the unmatched group was less likely to agree with the statement: ‘My teacher says something positive to me when I have tried hard’ (item 28) or ‘My teacher makes him/herself available if I want to talk about my concerns’ (item 21). These boys were also less likely to agree with item 11 ‘My teacher really cares about me as an individual’ and item 10 ‘My teacher treats me fairly’.
Table 5 displays the effect size in standard deviation units for teacher data comparisons between the mean item scores for relationships that were perceived as strong by both teachers and boys, with the mean item scores for relationships that were identified as strong by boys but moderate by teachers. Consistent with lower variability in the teachers’ responses, the effect sizes here are smaller. However, discrepancies were still evident and the major discrepancies occurred with Satisfaction items. In the unmatched group, teachers were less likely to agree with item 13 ‘I am happy with my relationship with this student’, item 3 ‘I would describe my relationship with this student as positive’, and item 1 ‘I enjoy having this student in my class’. These relationships were also more likely to be a cause of frustration to teachers: ‘This student frustrates me more often than most other students in my class’ (reverse scored).
TABLE 5 Item Comparisons of Relationships Identified as Strong by Both Boys and Teachers (n = 27) With Relationships Identified as Strong by Boys but Moderate by Teachers (n = 19) (Teacher Data)
Note: 2 Difference between the means expressed in SD units.
S = Satisfaction, C = Conflict, IH = Instrumental Help
In sum, overall the student–teacher relationships reported by participants in this study were generally positive. However, within this general pattern there were important instances of mismatch between student and teacher perceptions of their relationship. Examination of these instances of mismatch highlighted differences in the bases of their judgments; the boys were more concerned about what they perceived as the teacher's lack of care and understanding for them, while the teachers expressed concern over the relative lack of satisfaction they got from these relationships.
Discussion
A significant aspect of the findings was that both boys and teachers mostly rated their relationships positively and this is encouraging for the participating school, which has a strong focus on student welfare and pastoral care. As pointed out earlier, previous research indicates how beneficial positive relationships can be for students’ development (Murray & Malmgren, Reference Murray and Malmgren2005; Victorian Government, 2003). However, despite the overall picture of positive student–teacher relationships, it was possible to identify two different profiles of student–relationships, strong and moderately strong, from both student and teacher perspectives.
According to the boys’ responses on the ‘My English Class’ subscales, one profile of relationships was rated significantly higher than the other on every item. This indicated that these boys perceived their relationship with their home group teacher as more positive and stronger than did the boys with the more moderate profile.
According to the teachers’ responses on the TSRI, one profile of student–teacher relationship was also rated significantly higher than the other on nine out of ten items on the Satisfaction and Instrumental Help scales. On the Conflict scale, the same group was rated as lower on item 4 ‘This student frustrates me more often than most other students in my class’. There was no difference between these profiles on the other two Conflict items where all the ratings were very low.
SOME MISMATCH IN THE PERCEPTIONS OF BOYS AND TEACHERS
In just over half of the cases student and teacher perceptions of their relationships did match. These results are in line with a recent study that found some concordance between teacher and adolescent student perceptions for the positive dimensions of relationships (Murray & Zvoch, Reference Murray and Zvoch2011). For the remaining 28 students, boys’ and teachers’ perceptions did not match.
Student data indicated that the boys in relationships that teachers perceived as strong but boys identified as moderate were characterised by boys considering teachers less likely to understand or care about their feelings. These boys also thought teachers were less likely to care about them, make themselves available or help them. On the other hand, teacher data indicated relationships that boys perceived as strong but that teachers considered moderate were characterised as being less satisfying and more frustrating.
The incidence of mismatch between some boys and teachers’ perceptions of their relationships is a very important finding given that much of the previous research on student–teacher relationships has relied solely on teacher reports to establish relationship quality (Mercer & DeRosier, Reference Mercer and DeRosier2010). Thus our findings provide a salutary caution for research and development of interventions designed to improve the quality of student–teacher relationships. Both student and teacher perspectives on student–teacher relationships should be considered.
ELEMENTS THAT BOYS AND TEACHERS ASSOCIATED WITH STRONGER RELATIONSHIPS
In the current study, relationships considered strong by both boys and teachers were characterised by greater help-seeking, were rated higher on items from the Satisfaction scale by teachers and were described by boys as involving more positive feedback. Relationships considered moderate by both groups involved less frequent help seeking and positive feedback and were scored lower by boys on perceived teacher attitude towards them.
However, examination of the cases that were perceived differently suggested that some relationship elements were more important for each group. Satisfaction with the relationship with the student was an important factor for teachers. On the other hand, the teachers’ perceived caring attitude towards boys and understanding of their feelings was considered important to boys. In relationships that teachers considered stronger but boys did not, boys described teachers as less likely to care, notice problems or provide help than in relationships that both groups saw as strong.
LIMITATIONS
Our results indicated some limitations with the instruments used in the present study. In particular, items on the Conflict scale of the TSRI indicated concerns with the discriminant validity of the survey. The majority of teachers surveyed used the Almost never response category for items indicating conflict; ‘I cannot wait for this year to be over so that I will not need to teach this student next year’ (item 7) and ‘If this student is absent, I feel relieved’ (item 8). Responses on item 4 ‘This student frustrates me more than others in the class’ were also positively skewed. These items required teachers to be very negative in their description of their relationships with their students, which challenges the underlying assumption that teachers treat all their students equally. The pattern of results suggests that items may need to be reworded to tap into teachers’ perceptions of potential conflict with their students. Despite the reported construct validity of the scale (Buss & Warren, Reference Buss and Warren2000) our results suggest a need to reexamine the validity of the Conflict subscale.
The exploratory analysis of the bases of mismatch between perspectives of students and teachers on their student-teacher relationships should be verified by replicating the study with different and larger samples. A response rate of 60% means that teachers and students who did not have strong student-teacher relationships were probably less likely to have participated and this will have added a selective bias to our assessment of student-teacher relationships.
CONCLUSION
Despite the limitations we have identified, an overall positive pattern in perceptions of student–teacher relationships was evident and within these data two distinct profiles of student–teacher relationships from both boys’ and teachers’ perspectives were identified. While there were indications of agreement between the perceptions of boys and teachers for just over half the cases, there were some important bases of difference in how student–teacher relationships were perceived. In light of these results, the tendency observed in previous research (Mercer & DeRosier, Reference Mercer and DeRosier2010; Murray, Murray, & Waas, Reference Murray, Murray and Waas2008) to rely solely on teacher reporting about student–teacher relationships is indeed problematic.
Where there was mismatch, teachers appeared to consider satisfaction a key element of strong relationships, while boys appeared to consider positive feedback and a caring, helpful attitude towards themselves as important elements of a strong relationship. While it is established that positive student–teacher relationships increase the likelihood of positive outcomes, this and further research can enable more effective relationship building between teachers and their male adolescent students. Interactions that focus on elements which are important to boys and to their teachers will be most effective.
