In this research project the association between structural and compositional school features and faculty trust in students, parents, colleagues and principals is investigated. Within organizational studies trust is a long-standing topic, but it has only recently been examined in a systematic way within educational research (Bryk & Schneider, 2002; Goddard et al., 2001; Hoy & Tschannen-Moran, 1999; Smith et al., 2001). The importance of research into the determinants of teacher trust is great though, since trust is related to an effective functioning of the school. For example, trust may influence students’ performances (Goddard et al., 2001) and teachers’ functioning by affecting their sense of efficacy (Hoy & Tschannen-Moran, 1999) and job satisfaction (Van Houtte, 2006). Given that the possible impact of school features on teachers’ trust has been examined hardly in earlier research, this project focuses on the association between teachers’ trust and structural and compositional school features. It is not only the case that teacher trust is a quite recent topic within educational sociology, it is in addition largely of American origin. Hence, analyzing teacher trust in its relation to school characteristics within a European context, namely Flanders, is quite innovative. Within organizational and educational studies trust is described as “confidence that expectations will be met” (Bryk & Schneider, 2002; Goddard et al., 2001; Rousseau et al., 1998; Smith et al., 2001; Tschannen-Moran & Hoy, 2000; Tarter et al., 1989). Some authors have shown that trust may not only be considered as an individual teacher‘s feeling but also as a collective feeling, namely faculty trust (cf. Goddard et al., 2001; Hoy & Kupersmith, 1985; Hoy & Tschannen-Moran, 1999; Smith et al., 2001; Van Houtte, 2006). Therefore the central research question in this project is: “Does there exist something as collective trust, that is, faculty trust?” If faculty trust seems to exist, a subsequent step is to look at its determinants at the school level. If results do show that there is something as faculty trust within secondary schools in Flanders, this is interpreted as an evidence for teachers sharing a feeling of trust at the school level, indicating that faculty trust is part of a school ’s staff culture (Van Houtte, 2005). Four dimensions of teachers’ trust (Hoy & Tschannen-Moran, 1999) are taken into account: trust in the principal, trust in colleagues, trust in parents and trust in students. Next, each of these dimensions is analyzed in its relation to structural and compositional school features, which have been reported in previous research as possibly associated with faculty trust. School size and school sector, dichotomized in private and public schools, are taken into account as structural school features, whereas the compositional school features consist of the schools’ socioeconomic, gender and ethnic composition. Previous research has indicated a positive relation between schools’ socioeconomic composition and teacher trust in students, whereas a negative relation has been shown for schools’ ethnic composition (Goddard et al., 2001). Schools’ gender composition has been related to teacher trust in students as well (Van Houtte, 2007). Combining structural and compositional school features in their relationship to the four dimensions of faculty trust is quite unique in an European context, especially within a large inquiry covering 84 secondary schools.

Data were collected in the school year 2004-2005 within the context of the Flemish Educational Assessment (FlEA). Anonymous data are at hand of 2104 teachers instructing the third and/or fifth grade in a representative sample of 84 secondary schools in Flanders. To conceptualize the faculty trust, we examined the teacher trust by means of the scales developed by Hoy and Tschannen-Moran (1999). At this point, we have measures of trust at the individual teacher level, but since we intend to measure faculty trust, which is a group feature, the aggregation of the obtained measure is a necessary next step. A customary aggregation strategy is to calculate the mean of the scores of the individual members of the group or organization (e.g., Hofstede, Neuijen, Ohayv, & Sanders, 1990). In doing this, one has to be sure this aggregation is permitted. To examine if it is allowed to speak of faculty trust, i.e. something ‘shared’ at the school level, an index of “mean rater reliability” based on the intraclass correlation (ICC) from a oneway analysis of variance (Glick, 1985) is opted for. The interrelatedness between the dimensions of faculty trust - faculty trust in students, parents, colleagues and principals - is analyzed by means of bivariate correlations. Finally, multiple regressions are used to assess the associations between the four dimensions of faculty trust and the included structural and compositional school features.

The applied trust scales were all reliable (Cronbach’s alpha’s > 0.76). Results do show that ‘faculty trust’ is an existing concept within Flemish secondary schools. Teachers’ trust is shared at the school level in all its dimensions (ICC’s > 0.72). The explained variances at school level (ρ’s > 0.13) indicate that taking into account school features as determinants of faculty trust is reasonable. Faculty trust in parents was only significantly correlated to faculty trust in students, indicating that faculty trust in parents and students is a unitary concept, namely faculty trust in clients. A significant regression model was found to explain the variance of faculty trust in students, parents and colleagues by means of structural and compositional school features. This is not the case for faculty trust in principals. This finding suggests that faculty trust in the principal is probably more determined by individual characteristics than by school characteristics. The model ‘s explained variance is remarkably greater for faculty trust in students and parents than for faculty trust in colleagues. With respect to faculty trust in students, school size has a negative relation and the proportion of female students a positive one. SES context is positively related to faculty trust in students, parents and colleagues. Faculty trust in colleagues is lower in public schools and positively associated with the proportion of pupils from an ethnic minority. The effects on faculty trust seems to be stronger for compositional school features than for structural ones.