Teacher Trust in District Administration

Abstract

Purpose: We set out in this study to establish a foundation for a line of inquiry around teacher trust in district administration by (1) describing the role of trust in capacity building, (2) conceptualizing trust in district administration, (3) developing a scale to measure teacher trust in district administration, and (4) testing the relationship between district trust and teacher commitment. Method: Teachers were the unit of analysis. Data were collected from a sample of teachers in one urban school district. Construct validity was assessed by examining content, structural, and convergent validity of the scale. A fully latent structural equation model was used to test the relationship between teacher trust in district administration and teacher commitment. Results: This study makes a strong case for developing a line of research on teacher trust in district administration. It establishes a good measure to use in future research, and it provides initial evidence showing that teacher beliefs are sensitive to the actions of district administrators. Implications: A valid and reliable measure can be used by researchers to study systematically the formation and effects of teacher trust in district administration. Accurate information on district trust also allows central office leaders to formatively assess the capacity of the school system to accomplish reform objectives at scale.

Keywords

trust in district administration district leadership capacity building teacher commitment

Two general findings on school reform provide the backdrop for this study. First, uneven progress in raising achievement and reducing achievement gaps has shifted the reform focus from individual schools to the larger district context in which schools are nested (Daly Liou, & Moolenaar, 2014; A. Hargreaves & Shirley, 2009; Harris, 2011; Harris & Chrispeels, 2006; Sharrat & Fullan, 2009). Second, school systems making measurable and sustainable improvements in teaching and learning have done so by building the capacity of educators to learn from practice (King & Bouchard, 2011). Considerable evidence within the United States and abroad indicates that top performing school systems are distinguished by their capacity to continuously enact changes that produce better processes and outcomes for students across the entire system (Chenowith, 2007; Darling-Hammond, 2005; Fullan, 2010; D. H. Hargreaves, 2011; Mourshed, Chinezi, & Barber, 2010).

Capacity is not a tool or resource that districts purchase and input into schools. It emerges through a relational context the supports information exchange, knowledge creation, and purposeful action (Forsyth, Adams, & Hoy, 2011; Fullan, 2008; Harris, 2011). A move toward capacity building requires careful thought and action for how district leaders organize and coordinate the work of schools and teachers (Sharrat & Fullan, 2009). Leaders who centralize too much control at the top threaten to constrict knowledge creation and adaptation at the school level, whereas too little coordination across schools tends to produce unequal learning opportunities and outcomes (Darling-Hammond, 2005; Honig & Hatch, 2004). The tight–loose balance is a tricky dance for district leaders to master. Successful execution requires a relational context that synchronizes district, school, and teacher actions. Herein enters trust. Trust operates like glue by connecting individuals and groups to a common purpose and as a lubricant that eases collaboration and cooperation among interdependent actors (Forsyth et al., 2011; Tschannen-Moran, 2004, 2014).

The school trust literature is extensive, spanning over 30 years and including diverse school samples (national and international), different trust forms, and a variety of research methods (Forsyth et al., 2011; Tschannen-Moran, 2014). We know from this literature that trust is antecedent to important educational processes and outcomes: professional learning, instructional change, collective action, collaboration, school outreach to parents, knowledge creation, student achievement, school identification, motivation, and school performance (Forsyth et al., 2011; Bryk & Schneider, 2002; Tschannen-Moran, 2014). Even though the trust evidence runs deep, it is limited almost entirely to role relationships within schools (e.g., teacher–teacher, teacher–principal, student–teacher, teacher–parent). Research on trust in district administration is scarce, leading to an impoverished explanation for how decisions and actions made at the executive level affect the attitudes and behaviors of individuals whose collective actions can energize or constrain capacity.

On the surface, relational ties between teachers and district administration would seem to function as a critical conduit for greater school and district capacity. Existing evidence, however, does not provide much explanation for the teacher–district administration relationship. We argue that teacher trust in district administration is a useful line of inquiry for understanding the pathway by which district leaders work through teachers to ensure that students receive a learning experience that prepares them for a purposeful life. We set out in this study to establish a foundation for this line of research by (1) describing the role of trust in capacity building, (2) conceptualizing trust in district administration, (3) developing a scale to measure teacher trust in district administration, and (4) testing the relationship between district trust and teacher commitment.

District Context, Capacity, and Trust

School districts operate as intermediary agents in the larger educational system (Firestone, 2009; Harris & Chrispeels, 2006; Sharrat & Fullan, 2009); they span boundaries between federal/state policy, community interests and values, and local school needs. As intermediary agents, district leaders control how they implement policies and respond to external pressure (Honig & Hatch, 2004). This is an important function. Policy implementation is where the real, messy work of change unfolds (Darling-Hammond, 2005; Darling-Hammond et al., 2006; Honig & Hatch, 2004; McLaughlin, 1990), and district leaders retain considerable discretion in the design and use of strategies to achieve goals and objectives set forth in state and federal policy (Firestone, 2009). Trust in district leaders is a resource that on the surface seems to be a condition that enables teachers and central administrations to work cooperatively toward shared goals and aims.

Firestone (2009) describes three types of district contexts that uniquely affect the role of trust for policy implementation and improvement efforts. First, an accountability context tightens control over the instructional core by standardizing teaching practices and materials, closely monitoring teaching and student outcomes, sanctioning poor performance when goals are not met, and rewarding high achievement. External control, more so than trust, regulates decisions and behaviors in an accountability context (Forsyth & Adams, 2014). Second, a loosely coupled system has many moving parts with little to no coherence at the district level. Loosely coupled districts tend to cycle through programs and interventions, chase money, set unrelated goals, and create silos of decision making and action (Firestone, 2009).

Finally, a learning context, the ideal type of district environment to sustain quality performance (Firestone, 2009), uses capacity building to drive system-wide reform (Darling-Hammond, 2005; Fullan 2008, 2014). Capacity emerges as relational connections enable individuals and schools to learn from their experiences (Adams, 2013; D. H. Hargreaves, 2011). As Sharrat and Fullan (2009) argue, “Capacity is a highly complex, dynamic, knowledge-building process” (p. 8). The work of district leaders is to build the social infrastructure by which schools create knowledge to accurately diagnose if, how, and why improvement strategies are leading to more effective performance (Levin, 2008). For capacity to grow across a district, superintendents and other central office leaders need to be trusted.

There is nothing innovative or particularly new about features of high-capacity school systems. They set clear expectations for quality instruction and student learning, establish a coherent and aligned curriculum, use process and outcome data to study variation in teaching and student performance, and attract, retain, and develop talented educators (Fullan, 2010; Harris, 2011). What these characteristics have in common is the reliance on trust to make structures, processes, and practices functional by igniting behavior that leads to improved teaching and learning. Trust enhances cooperation (Tschannen-Moran, 2014), enriches openness (Hoffman, Sabo, Bliss, & Hoy, 1994), promotes cohesiveness (Zand, 1997), facilitates knowledge creation (Adams, 2013), deepens change (Daly, Moolenaar, Bolivar, & Burke, 2010), and builds school capacity (Cosner, 2009). System-wide reform carried out through a culture of trust is the difference between knowing characteristics that define effective districts and how to actually make schools better places to teach and learn. The latter moves school organizations forward while the former leaves behind many unfilled promises.

Capacity does not form simply by establishing quantitative performance targets, inputting new resources into schools, adopting new teacher evaluation models, and holding people accountable for results (Harris, 2011; King & Bouchard, 2011; Sharrat & Fullan, 2009). Rather, capacity grows as social and psychological barriers to change are replaced by a culture that values risk taking, experimentation, cooperation, and collective problem solving (Schein, 1996; Spillane, Reiser, & Reimer, 2002). Open and connected social networks provide the relational structural for deep knowledge creation (Daly et al., 2010; Daly et al., 2014), while trust establishes the psychological safety to ask tough questions about the effectiveness of strategies, resources, and practices (Garvin, Edmondson, & Gino 2008). If knowledge creation drives capacity as many scholars claim, then trust is the ignition that starts the process moving forward.

Teacher Trust in District Administration: Conceptual Definition

Trust is a complex phenomenon that exists at different analytical levels (Van Maele, Van Houtte, & Forsyth, 2014). It has been studied as a personality trait (Rotter, 1967; Zand, 1972), a relational condition (Bryk & Schneider, 2002), a group norm (Forsyth et al., 2011), and an organizational property (Shamir & Lapidot, 2003). Despite its different forms and characteristics, scholars have reached agreement on some general attributes of trust. These attributes are reflected in the definition used in this study. Trust is a teacher’s willingness to risk vulnerability based on the confidence that district administrators act benevolently, competently, openly, honestly, and reliably (Hoy & Tschannen-Moran, 1999; Mishra, 1996).

Teacher trust in district administration is conceptualized and measured as a type of relational trust that forms as teachers observe and judge the actions and intentions of district leaders. Bryk and Schneider (2002) argue that relational trust is fundamentally an intrapersonal phenomenon that emerges through interactions that occur within defined role relationships. We argue that teacher trust in district administration manifests itself as a teacher perception, not a collective property of a school faculty or a normative condition of the school district. Trust that is measured as an individual belief is substantively different than a collective property (Forsyth et al., 2011). A collective property is the representation of the assumptions, beliefs, and values held in common by members of a role group (Forsyth et al., 2011); it reflects a group norm. We aim to capture trust beliefs of individual teachers, not the collective perception of a teaching faculty in a school.

Relational trust corresponds to the unique social structure of school organizations (Schneider, Judy, Ebmeye, & Broda, 2014). School systems have been described as a complex web of social actors who interact within established role relationships and defined responsibilities (Bryk & Schneider, 2002; Schneider et al., 2014; Van Maele et al., 2014). Critical role relationships affecting school processes and practices extend throughout the larger social system in which schools are embedded. To illustrate, school districts cannot accomplish system-wide goals without teachers, and teachers in turn depend on district leaders to organize teaching and learning in ways that maximize teacher potential and support their effectiveness in the classroom. Trust enables teachers to risk vulnerability by placing themselves in an uncertain position based on confidence that district leaders will respond in ways that are not detrimental to their instructional effectiveness or professional growth (Van Maele et al., 2014).

Social exchanges between teachers and district administrators may not be as frequent or direct as they are between teachers and principals, but hierarchical boundaries and organizational constraints do not eliminate opportunities for teachers to evaluate the collective actions of central office leaders. Teachers follow decisions made at board meetings, read stories communicated in the media, have access to intradistrict newsletters, receive messages through site administrators, and have informal conversations with colleagues. Direct social exchanges combined with other observations build a body of evidence that teachers use to interpret decisions and actions of district leaders. This evidence becomes the wellspring of trust discernments.

Trust in District Administration Scale Development

Before we could proceed with an empirical test, we needed to develop a scale to measure teacher trust in district administration. Similar to existing school trust measures, items forming the Teacher Trust in District Administration Scale operationalize district actions that align with the trust facets. Teacher trust is observable in the perceived benevolence, competence, openness, reliability, and honesty of district leaders. The major conceptual difference between teacher trust in district administration and existing school trust scales (e.g., Omnibus T-Scale, Parent Trust in School Scale, Student Trust in Teacher Scale) is the unit of measurement. Collective trust measures are written so the trustee and trustor reflect the collective group. Teacher trust in district administration is conceived of and measured as an individual teacher belief. The trustor is the individual teacher, not the collective teaching faculty in a school. Thus, items capturing benevolence, competence, openness, honesty, and reliability reflect the collective action of district administration yet are written at a level that measures individual teacher beliefs, not teachers’ shared perceptions of district leaders. We now turn to a description of the facets and proposed scale items.

Benevolence in interpersonal exchanges relates to confidence that one’s interest or something one cares about will be protected by the trustee (Baier, 1986; Mishra, 1996). In an organizational setting, benevolence extends beyond basic care and concern for a single individual to a belief that the interests and well-being of the collective are preserved (Barber, 1983; Ouchi, 1981). Goodwill toward the collective defines benevolence items for trust in district administration. Teachers depend on district administrators to act in ways that express their care and concern for the school community as a whole. Often this involves valuing the expertise and work of teachers. District administration express benevolence by how they work with teachers to build a school environment where teaching and learning thrive. Items capturing benevolence include the following: District administrators value the expertise of teachers. District administrators show concern for the needs of my school.

Competence reflects the trustee’s perceived ability to perform tasks that are required of his/her position (Gabarro, 1987). Trust diminishes if individuals and groups do not demonstrate the competencies needed for successful performance. For educational systems, competence is reflected in knowing how to organize teaching and learning in ways that bring out the best in teachers and students alike (Darling-Hammond, 2005). Teacher discernment of competence teeters on the ability of district leaders to create an organizational environment that develops instructional expertise. Items capturing competence include the following: District administrators demonstrate knowledge of teaching and learning. District administrators have established a coherent strategic plan for the district.

Openness functions like a valve in trust production. When individuals are open their intentions flow freely, but when closed intentions remain hidden, raising doubts about one’s motives and future actions (Adams, 2010; Mishra, 1996). Openness by district administration is observable in two types of actions. The first involves transparency by which decisions are made and performance information is communicated to teachers (Tschannen-Moran & Hoy, 1998). The second is the willingness of district leaders to listen to teachers and to understand their experiences. Openness connects teachers to the larger district vision by providing them with influence over school improvement efforts and regularly communicating important decisions and information (Fullan, 2008). Items capturing openness include the following: District administrators are open to teacher ideas about school improvement. District administrators are transparent in making strategic decisions about district performance.

Honesty is based on integrity and truthfulness (Hoy & Tarter, 2004). Disingenuous actions can elicit a degree of suspicion that raises red flags about the sincerity and intentions of leaders. Suspicion dampens a willingness of teachers to risk vulnerability (Tschannen-Moran, 2004, 2014). In contrast, honesty builds trust by increasing confidence in the actions of the trustee (Mishra, 1996). Honest behavior reflects both the accuracy of information communicated to others and the acceptance of responsibility for decisions and actions (Adams, 2010). District administrators who hide facts, blame others, or cover-up mistakes damage the integrity of the central office and jeopardize its ability to advance an agenda that improves learning for all students (Sharrat & Fullan, 2009). Items capturing honesty include the following: District administrators often say one thing and do another. District administrators take personal responsibility for their actions and decisions.

Reliability relates to predictable and consistent behavior (Butler & Cantrell, 1984). For interpersonal exchanges, reliability gets measured in terms of perceived fairness (Adams & Forsyth, 2009), but at the organizational level, reliability reflects consistent and dependable action (Mishra, 1996). In a school district, reliability can be found in coherent and consistent improvement efforts centered on a predictable instructional focus (Sharrat & Fullan, 2009). Districts that cycle through programs and chase after the latest research-based strategies cannot establish the consistency of practice needed for continuous improvement (Firestone, 2009; Fullan, 2014). Items measuring reliability include the following: District administrators follow through on commitments. District administrators are committed to the stated goals of the district.

Benevolence, competence, openness, honesty, and reliability establish criteria by which teachers judge the collective actions of district administrators. Criteria for trust discernments, however, do explain how judgments of trustworthy behavior interact in the cognitive process to elicit an overall trust belief. Two distinct arguments on the nature of trust beliefs have emerged. First, psychometric evidence on several existing trust scales (see Forsyth et al., 2011; Tschannen-Moran, 2004, 2014) implies that trust is a one-dimensional construct with the five facets sharing variance around a single factor. Second, a more recent belief has been advanced by Romero (2010) and Makiewicz and Mitchell (2014) who conceptualized and measured trust as a second-order construct, specifying trust facets as related but distinct factors. These two different conceptualizations call attention to the factor structure of the trust concept. Do trustworthy behaviors cohere around a single, latent trust variable or do they cohere around distinct trust facets?

Teach trust in district administration is hypothesized to be a single factor construct (Figure 1). Items derive from trust facets, but facets are not independent behaviors; they are interrelated and converge in the discernment process to form a unitary belief (Forsyth et al., 2011; Hoy & Tschannen-Moran, 1999; Tschannen-Moran, 2004). That is, trust facets are inextricably related; the absence of one affects the presence of others (Adams, 2010). To illustrate, district administrators who regularly listen to teachers and solicit teacher feedback are likely to be perceived as open, benevolent, and competent. The weight of the existing empirical evidence supports specifying trust as a single-factor construct. Exploratory factor analyses from multiple samples of the Omnibus Trust Scale, Student Trust in Teachers Scale, and Parent Trust in School Scale consistently find that scale items load strongly on one factor (Forsyth et al., 2011).

Figure 1.

Hypothesized single-factor model of teacher trust in district administration.

Validation Study

The purpose of the validation study was to evaluate the construct validity of the Teacher Trust in District Administration Scale before using the measure in the empirical test. Construct validity refers to the ability of a measure to yield truthful judgments about the object it purports to measure (Messick, 1995; Miller, 2008). While there are different aspects of construct validity (content, discriminate, convergent, consequential, etc.), these aspects share the same fundamental logic—that validity exists to the degree that the measure represents the underlining theoretical construct and informs credible judgments about the phenomenon of interest (Cronbach, 1971; Messick, 1995).

Construct validity was assessed by examining content, structural, and convergent validity. Evidence for content validity comes from two sources. First, similar to existing trust scales, items were written to capture the different facets of trustworthy behavior. Second, the 10 items were submitted to a group of 30 educators to assess item clarity and alignment with the trust facets. This review resulted in support for the theoretical alignment between the items and the underlying trustworthy behaviors. Additionally, respondent feedback revealed behavioral characteristics of district administrators that potentially link to more than one facet. For example, the item, District administrators often say one thing and do another, was identified as representing openness, honesty, and reliability. Similarly, the item, District administrators value the expertise of teachers, was largely identified as benevolence but a few respondents felt it was also a reflection of competence.

Structural validity and convergent validity were examined in a field test with more than 800 teachers from an urban district in a southwestern city. Confirmatory factor analysis (CFA) in AMOS 19.0 was used to test the facture structure of the scale. Building and testing a measurement model a priori has two advantages over a traditional exploratory approach. First, CFA models are guided by theory. The empirical relationships found in the sample data will either support or repudiate the underlying logic of the measure. Second, CFA is useful for evaluating comparative model structure and fit by testing different theoretical specifications of the observed and latent features of the construct (Thompson & Daniel, 1996). In this case, a comparative analysis was used to test the hypothesized model by comparing estimates against a second-order specification of district trust.

Sample

Teachers were the unit of analysis. Data were collected in February of 2013 from a random sample of teachers in an urban school district. A roster of certified teaching faculty in each school was provided to the researchers by the school district. Half of the teachers in each school were randomly sampled to receive a survey with the 10 teacher trust in district administration items and additional survey questions about academic emphasis and perceptions of the teacher evaluation model. This resulted in a sample of 1,305 teachers in 73 schools. Of the 1,305 teachers, 849 completed and returned usable surveys for a response rate of 65%. Teachers in the sample averaged 13 years of teaching experience and 6 years in their current school, approximately 9% were nationally board certified, and 85% were female (Table 1).

Table 1.

Teacher Demographic Information.

	Mean	SD	Min	Max
Years experience	13.11	9.10	1	30
Years in current school	6.13	6.4	1	30
National board	0.09	0.28	0	1
Female	0.85	0.36	0	1

Note. N = 849 teachers.

The school district is located in a southwestern city with a metropolitan population of around 900,000 residents. The district serves approximately 42,000 students across 88 different educational sites. Student demographics in the district include 30% Hispanic, 27% African American, 27% Caucasian, 8% multiracial, 6% American Indian, and 1% Asian. Seventy-nine percent of the students qualified for the federal lunch subsidy. The district employees approximately 2,978 certified staff. Nearly 49% have more than 11 years in the district, 123 are nationally board certified, 30% are racial minorities, and 1,173 hold graduate degrees.

Measures

Convergent validity was examined by correlating teacher trust in district administration with teacher perceptions of the performance-based teacher evaluation system. Teacher evaluation has become a central feature of district reform. Trust in district administration would, on the face of it, appear to be related to teacher perceptions of the teacher evaluation model. Three items measuring teacher favorableness of the evaluation system were taken from Milanowski and Heneman’s (2001) teacher evaluation survey. The items capture teacher perceptions of the system as a whole, the evaluation process itself, and its perceived valence. The items include the following I am satisfied with the new evaluation system. The evaluation process takes more time than it is worth (reversed scored). The new system makes working in the district more attractive to me. The Likert-type response set ranges from Strongly Disagree (coded as 1) to Strongly Agree (coded as 6).

Analysis

For structural validity, a model generating approach using maximum likelihood estimation was used in AMOS 19.0 to evaluate the hypothesized single-factor specification of trust against the alternative second-order model. The first step was to build and test the single-factor model with all 10 items loading on trust. The second step was to build and test a second-order model with each facet conceptualized as a first-order factor of the latent second-order trust concept. This second step addresses Moss’ (1995) argument that “construct validation is most efficiently guided by the test of plausible rival hypotheses” (pp. 6-7). Specifying trust as a second-order model reflects the rival hypotheses as advanced by Romero (2010) and Makiewicz and Mitchell (2014). Fit indices, parameter estimates, and residuals were used to evaluate the two competing models. The absolute fit index was the root mean square of approximation (RMSEA). Relative fit indices included normed fit index (NFI), Tucker–Lewis index (TLI), and the comparative fit index (CFI).

The final step was a model trimming process for both the simple-factor hypothesized model and the alternative second-order construct. Model modification is a conventional and acceptable approach for identifying a parsimonious model with the best fit (Schreiber, Nora, Stage, Barlow, & King, 2006). When trimming models for parsimony, it is critical to preserve the theoretical specification. For the hypothesized model, items were trimmed from 10 to 5. For the second-order model, first-order factors were collapsed from 5 to 3. This decision was based on the models advanced and tested by Romero (2010) and Makiewicz and Mitchell (2014), who measured trust as consisting of three factors: competence, benevolence, and integrity.

For convergent validity, we looked to the relationship between teacher trust in district administration and teacher favorableness of the teacher evaluation framework. Based on CFA results, we specified trust as a simple-factor latent variable composed of five observed variables representing each trust facet. Teacher favorableness of the evaluation framework was also treated as a latent variable so to account for measurement error in the analysis.

Harmon’s single-factor test was used to evaluate the degree to which common method bias may confound the estimated relationships in the convergent validity test. Common measurement bias reflects variance that is attributed to characteristics of the measurement itself and not an underlying relationship between constructs (Meade, Watson, & Kroustalis, 2007). The Harmon test estimates the potential problem of common method variance by subjecting items from all three constructs to an exploratory factor analysis (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). It is assumed that measurement bias exists if a single factor emerges to explain variance in the combined items (Meade et al., 2007). Results of our test showed that two factors were extracted with Eigen values more than one. Items had the strongest loading on their conceptual factors (see Appendix A).

The final analysis was to check our specification of district trust as an individual teacher belief. To do this, we estimated two types of intraclass correlation coefficients (ICCs): an ICC-2 with a one-way ANOVA (Mean square between − Mean square within/Mean square between) and an ICC-1 with an HLM unconditional model. ICC-2 is a measure of within-group homogeneity and assesses the degree to which individual teacher responses cluster around the school mean (Glisson & James, 2002). It is different from the HLM-derived ICC-1 in that the HLM model estimates between-group differences in teacher trust, not the cohesiveness of teacher trust perceptions within schools. Both estimates are necessary to justify aggregation of school-level variables because within-group consistency and between-group variability occur independent of each other (Glisson & James, 2002).

Results

Preliminary data screening confirmed that sample data were normally distributed. Skewness and kurtosis for all trust items were below 1.0 and the Quantile-Quantile plot of the items indicated that sample data aligned with a normal distribution. Furthermore, we did not find any outlier cases that would bias the estimates. Item means, standard deviations, skewness, kurtosis, and correlation coefficients appear in Table 2. Notable from these results is the statistically significant and strong relationship among all 10 items. Next, results are organized by tests of structural validity and convergent validity.

Table 2.

Item Means, Standard Deviations, Skewness, Kurtosis, and Correlations.

	Mean	SD	Skewness	Kurtosis	1	2	3	4	5	6	7	8	9	10
FTDist1	3.09	1.38	.15	−.89	1.0	.53	.58	.66	.59	.63	.59	.57	.60	.63
FTDist2	3.26	1.34	−.08	−.76		1.0	.83	.69	.71	.67	.61	.65	.63	.68
FTDist3	3.49	1.36	−.18	−.75			1.0	.75	.76	.72	.68	.73	.67	.72
FTDist4	3.50	1.32	−.31	−.56				1.0	.76	.85	.76	.70	.76	.77
FTDist5	3.40	1.45	−.13	−.88					1.0	.80	.74	.70	.70	.74
FTDist6	3.71	1.28	−.48	−.29						1.0	.81	.73	.75	.78
FTDist7	3.94	1.29	−.60	−.11							1.0	.74	.77	.75
FTDist8	3.69	1.37	−.43	−.56								1.0	.77	.79
FTDist9	3.64	1.37	−.39	−.62									1.0	.77
FTDist10	3.50	1.45	−.26	−.84										1.0

Note. N = 849 teachers. All correlations are statistically significant at p < .01.

Structural Validity

For structural validity, we first examined the fit indices to compare the hypothesized to the alternative specification. Estimates reported in Table 3 show good fit for the hypothesized model. Chi-square was statistically significant, but RMSEA, NFI, TLI, and CFI all exceeded acceptable thresholds, suggesting a reasonably good fit between the hypothesized model and the observed data. Estimates of the alternative, second-order model did not meet established thresholds for good fit. Chi-square was considerably larger than the single-factor hypothesized model, and RMSEA, NFI, TLI, and CFI were all below the acceptable criteria for a good fitting model.

Table 3.

Model Fit Indices for the Hypothesized and Alternative Models.

Fit Index	Criteria	Hypothesized Model	df	Alternative Five-Factor Model	df	Hypothesized Trimmed Model	df	Alternative Three-Factor Model	df
Chi-square	NS	188.6**	35	384.6**	30	5.0	5	224.7**	32
RMSEA	<.05	.05		.14		.03		.11
NFI	>.95	.97		.93		.99		.93
CFI	>.95	.97		.94		.99		.94
TLI	>.95	.95		.91		.98		.94

Note. N = 849 teachers. RMSEA = root mean square error of approximation; NFI = normed fit index; CFI = comparative fit index; TLI = Tucker–Lewis index.

p < .01.

Parameter estimates and residuals lend further support for conceptualizing and measuring trust as a simple, single-factor construct. For the simple-factor hypothesized model, all parameter estimates were statistically significant and above .70, suggesting strong relationships between the latent trust construct and each item (Table 4). Parameter estimates and residuals for the alternative second-order model point to multicollinearity concerns among the first-order factors. Three of the latent first-order factors had estimates more than 1.0 and negative residuals. Factor loadings over one and statistically negative error variance are not possible outcomes in an actual population (Kolenikov & Bollen, 2012).

Table 4.

Sample Items, Factor Loadings, and Squared Multiple Correlations for Hypothesized Model and Trimmed Five-Item Model.

Latent Factor and Items	Factor Loadings	Squared Multiple Correlation	Residuals
Latent Trust Factor
District administrators value the expertise of teachers	.74	.54	.96
District administrators show concern for the needs of my school	.77	.59	.73
District administrators demonstrate knowledge of teaching and learning	.83	.68	.58
District administrators have established a coherent strategic plan for the district	.90	.81	.34
District administrators are open to teacher ideas about school improvement	.86	.74	.54
District administrators are transparent in making strategic decisions about district performance	.90	.81	.31
District administrators often say one thing and do another	.85	.73	.45
District administrators take personal responsibility for their actions and decisions	.83	.69	.59
District administrators follow through on commitments	.84	.71	.55
District administrators are committed to the stated goals of the district	.88	.77	.48
Five-Item Trimmed Model
District administrators show concern for the needs of my school	.81	.66	.61
District administrators have established a coherent strategic plan for the district	.91	.82	.31
District administrators are transparent in making strategic decisions about district performance	.88	.77	.48
District administrators often say one thing and do another	.78	.61	.74
District administrators are committed to the stated goals of the district	.84	.71	.61

Note. Factor loadings and squared multiple correlations were statistically significant at p < .01; N = 849 teachers. Degrees of freedom for the 5-item model was 5, and for the 10-item model was 35.

Negative residuals indicate a Heywood case stemming from either: a mis-identified model, outliers in the sample data, sampling fluctuations, or mis-specified structural relations (Kolenikov & Bollen, 2012). We can rule out problems with the sample data, model identification, and sampling fluctuation because the hypothesized model did not have the same estimation problems. This leaves a specification problem as the plausible reason for the Heywood case. We can trace the specific problem to a high degree of multicollinearity among the Benevolence, Openness, and Honesty factors. Parameter estimates over 1 for these factors suggest that the factors are measuring the same cognitive discernment. In other words, benevolence, openness, and honesty are not independent dimensions of trust. They are related facets that cohere in the discernment process (see Tables 5 and 6).

Table 5.

Sample Items, Factor Loadings, and Squared Multiple Correlations for Alternative Second-Order Five Factor Model.

Latent Factor and Items	Factor Loading	Squared Multiple Correlation	Residuals
Trust
Benevolence	1.02	1.05	−.04
Competence	1.01	1.01	−.02
Openness	1.02	1.03	−.05
Honesty	.97	.93	.09
Reliability	.95	.91	.95
Benevolence
District administrators value the expertise of teachers	.69	.47	1.0
District administrators show concern for the needs of my school	.78	.60	.70
Competence
District administrators demonstrate knowledge of teaching and learning	.84	.71	.54
District administrators have established a coherent strategic plan for the district	.88	.78	.38
Openness
District administrators are open to teacher ideas about school improvement	.85	.73	.57
District administrators are transparent in making strategic decisions about district performance	.87	.75	.53
Honesty
District administrators often say one thing and do another	.93	.86	.23
District administrators take personal responsibility for their actions and decisions	.88	.77	.38
Reliability
District administrators follow through on commitments	.87	.76	.44
District administrators are committed to the stated goals of the district	.88	.77	.43

Note. Factor loadings and squared multiple correlations were statistically significant at p < .01; N = 849 teachers. Degrees of freedom were 30.

Table 6.

Sample Items, Factor Loadings, and Squared Multiple Correlations for Alternative Second-Order Three Factor Model.

Latent Factor and Items	Factor Loading	Squared Multiple Correlation	Residuals
Trust
Benevolence	.90	.80	.17
Competence	1.01	1.02	−.02
Reliability	.96	.92	.11
Benevolence
District administrators value the expertise of teachers	.66	.44	110
District administrators show concern for the needs of my school	.87	.76	.42
District administrators are open to teacher ideas about school improvement	.93	.86	.26
Competence
District administrators demonstrate knowledge of teaching and learning	.88	.76	.35
District administrators have established a coherent strategic plan for the district	.86	.73	.54
District administrators are transparent in making strategic decisions about district performance	.87	.76	.28
Reliability
District administrators often say one thing and do another	.89	.80	.36
District administrators take personal responsibility for their actions and decisions	.86	.74	.50
District administrators follow through on commitments	.91	.83	.45
District administrators are committed to the stated goals of the district	.88	.77	.49

Note. Factor loadings and squared multiple correlations were statistically significant at p < .01; N = 849 teachers. Degrees of freedom were 32.

Estimates of the trimmed models provide additional evidence to test the validity of the Teacher Trust in District Administration Scale. The hypothesized simple-factor structure was trimmed from 10 items to 5, one item for each facet. The item with the strongest parameter estimate for each facet was retained. Fit indices for the trimmed hypothesized model show a better overall fit compared to the 10-item measure. Chi-square was considerably smaller and not statistically significant at the .05 level. Additionally, RMSEA, NFI, TLI, and CFI improved. All parameter estimates had a strong, statistically significant relationship with trust. Estimates for the respecified alternative second-order model suggested continued specification problems. Fit indices did not reach the thresholds for good fit and parameter estimates and residuals point to continued problem with multicollinearity between the latent first-order factors.

Reliability was estimated through an interitem consistency analysis and split-half reliability test. Both the 10-item simple-structure model and the more parsimonious 5-item model demonstrated strong reliability. Cronbach alpha was .96 for the 10 items and .93 for the 5 items. For the 10 items, Cronbach alphas for Part 1 and Part 2 of the split-half results were .90 and .93. Additionally, the correlation between the split forms was .93 with a Spearman–Brown coefficient of .96 and a Guttman split-half coefficient of .96. For the 5 items, Cronbach alphas were .84 and .91 for the two parts of the split-half with correlations between the forms at .84 and Spearman–Brown and Guttman split-half coefficients of .92, respectively.

The combined results establish empirical support for the structural validity of conceptualizing and measuring Teacher Trust in District Administration as a simple-factor construct with shared variance among trust facets. The five-item trust measure had the best fit with the observed sample data, strong factor loadings, and large square multiple correlations. The five-item measure also displayed strong reliability as assessed through interitem consistency and a split-half reliability test. Results do not lend support for measuring trust as a second-order construct composed of distinct first-order factors. We now turn to evidence of convergent validity.

Convergent Validity

Results for the relationship between teacher trust and teacher perceptions of the evaluation system report good model fit. Chi-square was 36.1 and statistically significant, but RMSEA was below .05, NFI was above .95, CFI was above .95, and TLI was above .96. Factor loadings for the latent constructs were strong, ranging from .70 to .91. As for the correlation result, trust had a statistically significant and strong relationship with teacher favorableness of the evaluation framework (r = .54, p < 01; see Figure 2).

Figure 2.

Test of convergent validity with a fully latent structural equation model using maximum likelihood estimation.

Tests of within-group homogeneity and between-school variability in teacher trust in district administration support the specification of trust at the individual teacher level. An ICC-2 estimate of .50 was not statistically significant, and it was considerably below the .70 threshold set by Cohen, Doveh, and Eick (2001) as indicative of reliable group means. A small ICC-2 indicates weak agreement among teachers within schools as to the trustworthiness of district administrators (see Table 7). The ICC-1 achieved statistical significance at .05, but the amount of variance in district trust between schools was only 4%, a relatively small amount of variance in comparison to the school-level variance associated with collective trust measures (Forsyth et al., 2011).

Table 7.

Intraclass Correlation Coefficients.

Variable	ICC-1	Chi Square	ICC-2	F Ratio
Teacher trust in district administration	.04	82.36*	.50	.99

Note. N = 849 teachers.

p < .05.

To conclude, empirical results lend support for specifying trust in district administration as a simple-factor construct observable in the perceived benevolence, competence, openness, honesty, and reliability of district leaders. Trust can be measured with either the 10-item scale or the shorter 5-item scale. Both scales had good fit with the observed data, strong factor loadings, and high explained variance. Additionally, both scales had strong internal item-consistency and good split-half correlation results. Teacher trust was also related to teacher perceived favorableness of the evaluation framework. Both ICC-1 and ICC-2 estimates support the theoretical specification of district trust as a teacher belief.¹

Empirical Test: District Trust and Teacher Commitment

Teacher trust in district administration represents a line of research that appears capable of deepening our understanding of how school systems continuously get better at teaching and learning. As an initial study, we were interested in knowing if teacher trust in district administration can ignite beliefs and behavior that would enable teachers and schools to flourish. Teacher commitment to the school and its vision stands out as a psychological driver of individual and group capacity (Firestone & Pennell, 1993; Tsui & Cheng, 1999). Just like commitment to a goal motivates an individual to persist in her journey toward success, teacher commitment ignites the desire and determination to see that schools are working to serve students in ways that engage them in deep and meaningful learning (Nordin, Darmawan, & Keeves, 2009).

Commitment is defined as an individual’s identification with the values and goals of an organization, a willingness to work toward achievement of a shared vision, and a desire to remain in the organization (Angle & Perry, 1981; Mowday, Porter, & Steers, 1982; Ross & Gray, 2006). Committed teachers have established a psychological attachment that has deep and wide-ranging effects on their motivation, actions, and performance (Nordin et al., 2009). The effects of commitment also extend to the performance of schools and the larger school system in which teachers work (Reyes, 1992). At the individual level, committed teachers are autonomously motivated, persevere through challenges, and work cooperatively with colleagues (Firestone & Pennell, 1993; Nordin et al., 2009). For schools and school systems, having committed teachers generally leads to better achievement outcomes, a healthy teaching environment, less teacher conflict, and innovative practices (Henkin & Holliman, 2009; Kushman, 1992; Nordin et al., 2009).

Evidence has identified school leaders as an essential factor in getting teachers to commit to a school and its success (Henkin & Holliman, 2009; Nordin et al., 2009; Ross & Gray, 2006). Teachers have stronger commitment when principals are seen as open, collaborative, and empowering; where teachers collectively work to accomplish high academic goals; and where collective teacher efficacy defines the normative school climate (Lee, Zhang, & Yin, 2011; Ross & Gray, 2006; Ware & Kitsantas, 2011). It makes sense that daily interactions between principals and teachers would be a source of teacher commitment. Principals, through their decisions and actions, create an environment that can either reinforce teacher psychological attachment to the school’s vision, or conversely, alienate teachers by leading in ways that contrast with their values, expectations, and beliefs (Henkin & Holliman, 2009; Kushman, 1992; Nordin et al., 2009).

The effect of district leadership on teacher commitment is an interesting relationship to think about. At the executive level, district leaders have fewer interactions with teachers, making it possible for the psychological and behavioral effects of district trust to be attenuated by several school factors. That being the case, we believe that commitment to a school can be deepened when district leaders are viewed by teachers as trustworthy. Low trust would seem to lessen the faith that teachers place in district leaders, thereby jeopardizing their value congruence with the vision of the school, diminishing their willingness to carrying out improvement plans, and reducing their desire to remain in the school. Thus, we predict that teacher trust in district administration has a relationship to teacher commitment over and above the effects of teacher trust in principal.

Data Source

Teachers were the unit of analysis for the study. Data came from 785 teachers in the same urban school district as the validation study. A general teacher survey was administered in February of 2014 to teachers in 73 elementary and secondary schools. Half of the teachers in each school were randomly sampled to receive a survey of teacher trust in district administration, teacher trust in principal, and teacher commitment. The result was a sample of 1,273 teachers in 73 schools. Of the 1,273 teachers, 785 completed and returned usable surveys for a response rate of 62%.

Measures

District trust was measured with the five-item Teacher Trust in District Administration Scale. Teacher commitment was measured with five items adapted from the Organizational Commitment Scale (Porter, Steers, Mowday, & Bouhan, 1974). The scale captures the facets of value congruence, willingness to work toward a shared vision, and desire to stay in the school. Sample items include the following: “I feel very little loyalty to this school” (reversed scored). “I am glad I chose to teach in this school.” “I would probably continue teaching in this school.” Five items on teacher trust in principal were adapted from the Omnibus Trust Scale (Hoy & Tschannen-Moran, 1999) so as to measure individual teacher beliefs not collective teacher perceptions. Sample items include the following: “The principal in this school typically acts in the best interests of teachers.” “The principal in this school shows concern for teachers.” “The principal in this school is competent in doing his/her job.” The principal keeps teachers informed about school issues.” “The principal in this school is honest.” A Harmon single-factor test provided evidence to indicate that common method variance does not seem to be a concern with including all measures on a common survey. Three factors emerged from the extraction and all items loaded strongly on their theoretical property (see Appendix B).

Analysis and Results

We tested a fully latent structural equation model in AMOS 19.0. The fully latent model enabled us to account for measurement error in the estimation of the structural relationship among the two forms of trust and teacher commitment. Results (Figure 3) show that teacher trust in principal and teacher trust in district administration combined to explain 29% of the variance in teacher commitment. The unique effects of district and principal trust were similar. Teacher trust in district administration had a unique effect of .33, whereas the unique effect of principal trust was .32. District and principal trust were strongly related to each other with a parameter estimate of .40. Model fit indices show a strong alignment between the theoretical model and the sample data. Chi-square was statistically significant, but the comparative fit indices all met or exceeded acceptable thresholds for good fit: RMSEA was .05, CFI was .97, TLI was .96, and NFI was .96.

Figure 3.

Results of the empirical test of the fully latent structural equation model.

Discussion

Increasingly, scholars, policy makers, and practitioners have called on school districts to pursue an improvement strategy that builds capacity within schools (Darling-Hammond, 2005; Firestone, 2009; Harris, 2011; Sharrat & Fullan, 2009). Capacity grows out of a relational context that encourages risk taking, problem solving, knowledge creation, and adaptation among individuals and groups (D. H. Hargreaves, 2011). Trust, for its role in knowledge creation and learning, is foundational to high capacity school organizations (Adams, 2013; Bryk & Schneider, 2002; Tschannen-Moran, 2004, 2014). With this in mind, we set out to extend trust research to the teacher–district role relationship by developing a measure of teacher trust in district administration and using the measure in an initial empirical test to evaluate the usefulness of this line of research. Together, the validity evidence and findings from the empirical test lend support for extending trust research to district leadership.

Evidence from the validity study supports the use of the Teacher Trust in District Administration Scale for research and practice. Items for teacher trust in district administration were based on the facets of trust and written to reflect the collective actions of district administrators. For validity evidence, we were interested in three patterns in the data: (1) the relationship between scale items and the latent construct, (2) the factor structure of the scale, and (3) the correlation between district trust and convergent teacher beliefs. Based on findings from other trust measures, we hypothesized that teacher trust in district administration exists as a single-factor construct. This hypothesis stands in contrast to an alternative specification of trust as a multifactored construct with the facets functioning as distinct beliefs.

Results supported our hypothesized model. Each trust facet contributes to a teacher’s perception of district administration. No one facet stands out as more critical in the formation process than others. Benevolence, competence, openness, honesty, and reliability are inextricably related and converge in the discernment process to form a singular belief. While conceptual distinctions can be made, the facets represent highly integrated actions that together elicit teacher judgments of administrators’ trustworthiness. Our findings suggest it is hard to be perceived as benevolent without being perceived as competent, open, honest, or reliable. Similarly, competent behavior likely contains elements of benevolence, openness, honesty, and reliability. The absence of one facet affects the presence of others.

Additional validity support comes from the strong correlation between trust in district administration and perceived favorableness of the teacher evaluation system. We expected trust and teacher perceptions of the evaluation system to go hand in hand. Consistent with our prediction, higher teacher trust in district administration was associated with higher favorableness of the evaluation system, linking teacher trust beliefs to a strategy behind efforts to elevate teaching quality.

A valid measure of district trust by itself does not establish a strong case for developing this line of research. Evidence from the empirical test provides additional support. Results highlight the potential of district trust to explain differences in teacher, school, and school system capacity. At a minimum, capacity building requires a core group of dedicated and committed teachers who have the inspiration and know-how to address problems plaguing teaching and learning (Harris, 2011). Fostering such commitment has generally been viewed as the responsibility of school principals (Leithwood, Louis, Anderson, & Wahlstrom, 2004), and indeed, the evidence indicates that principals have considerable sway over the dedication and persistence of teachers (Henkin & Holliman, 2009; Kushman, 1992; Nordin et al., 2009). That stated, results of our empirical test add district trust to the equation.

We were not surprised that teacher trust beliefs would be related to their commitment. Trust has been described metaphorically as a type of glue that unites individuals and connects individuals to organizations (Bryk & Schneider, 2002; Tschannen-Moran, 2004, 2014). Nonetheless, we did not expect district trust to have as large of an effect on commitment when accounting for principal trust. Principals, for their proximately to teachers, would seem to have greater influence on teacher commitment than district leaders. This was not the case in our sample of teachers, suggesting district leaders may be closer to the psychological sources of motivated and engaged teachers than hierarchical boundaries imply.

The study had limitations that future research can address. First, we sampled teachers from one urban school district, thereby potentially reducing variability in individual teacher beliefs. Future research can broaden the sample to include teachers from different school districts. A sample of teachers across multiple school districts may yield different results. A second limitation was with the simple empirical test. The study was partly based on the argument that capacity building is a viable process for school improvement. Capacity, both as a concept and condition, is more complex and dynamic than what can be captured by measuring trust and teacher commitment. Future research can examine the relationship between teacher trust in district administration and other indicators of capacity and high performance. Finally, we believe evidence on the formation of teacher trust in district administration has utility as well. Understanding practices that build trust has the potential to shape how district leaders lead and manage system-wide improvement efforts.

Conclusion

In conclusion, we offer three implications for research and practice. First, as a simple factor construct, trust facets should not be weighted or ranked in an attempt to understand which actions are more likely to elicit positive discernments. With facets shaping a unitary belief, any attempt to weight the relative importance of specific behaviors would be an arbitrary decision. For example, we cannot conclude that competence is any more instrumental than openness, or that honesty does not carry the same weight as benevolence. Rather, findings confirm the theory that district administrators can build trust by consistently acting in ways that teachers perceive as benevolent, competent, open, reliable, and honest (Forsyth et al., 2011). Each trust facet contributes to teacher discernments of the actions and intentions of district leaders.

Second, trust in district administration has the potential to deepen our understanding of how executive leaders in school districts set a direction for improvement, organize and coordinate work processes, and develop talent and expertise. Deeper knowledge in these areas is critical when considering the pressure school districts are under to improve, and the relatively few examples of districts that consistently make teaching and learning better at scale (Bryk, Gomez, Grunow, & LeMahieu, 2015; Fullan, 2010). Unlike descriptive accounts of improving districts (Firestone, 2009; Waters & Marzano, 2006), trust explores the relational connections that either ignite positive change or become barriers to turning good ideals into effective practices. It is the human and social side of school districts that ultimately determines how school systems adapt to a changing environment (A. Hargreaves & Shirley, 2009; Sharrat & Fullan, 2009), and additional research on district trust can bring this knowledge to the surface.

Third, evidence on district trust can be used by practitioners to gauge capacity of the school system to continuously improve. Indeed, there is more to capacity than trust, but it is hard to envision a high-capacity system without similarly strong levels of trust between teachers and district leaders. High trust signals an open, cooperative, and cohesive relational network (Forsyth et al., 2011; Bryk & Schneider, 2002; Tschannen-Moran, 2004, 2014), whereas low trust points to relational problems that constrain information exchange and knowledge creation (Adams, 2013). Accurate information on district trust allows central office leaders to formatively assess the strength of relational ties with teachers.

As a whole, this study makes a strong case for developing a line of research on teacher trust in district administration. It establishes a good measure to use in future research, and it provides initial evidence showing that teacher beliefs are sensitive to the actions of district administrators. Looking ahead, additional knowledge on district trust can be used to map the process by which district leaders work through teachers to create high functioning school systems.

Footnotes

Appendix A

Exploratory Factor Analysis Results of the Teacher Trust in District Administration and Teacher Evaluation Scales.

	Factor 1	Factor 2
TTDA1	.74	−.21
TTDA2	.79	−.22
TTDA3	.82	−.25
TTDA4	.83	−.23
TTDA5	.78	−.27
TE1	.35	.65
TE2	.25	.72
TE3	.21	.77

Note. N = 849 teachers. Principal axis factoring with no rotation was used. Two factors were extracted with Eigen values over one. Factor 1 explained 56% of the variance, and Factor 2 explained 20%.

Appendix B

Exploratory Factor Analysis Results of the Teacher Trust in District Administration, Teacher Trust in Principal, and Teacher Commitment.

	Factor 1	Factor 2	Factor 3
TTDA1	.73	.21	−.08
TTDA2	.78	.32	−.01
TTDA3	.72	.35	−.09
TTDA4	.80	.31	−.17
TTDA5	.77	.37	−.18
TTP1	.29	.67	−.15
TTP2	.22	.70	−.29
TTP3	.37	.69	−.28
TTP4	.39	.68	−.23
TTP5	.34	.65	−.24
TC1	−.31	.27	.69
TC2	−.24	.21	.75
TC3	−.19	.30	.67
TC4	−.29	.29	.71
TC5	−.31	.29	.72

Note. N = 785 teachers. Principal axis factoring with no rotation was used. Three factors were extracted with Eigen values over one. Factor 1 explained 33% of the variance, Factor 2 explained 24%, and Factor 3 explained 13%. TTDA = Teacher Trust in District Administration items; TTP = Teacher Trust in Principal items; TC = Teacher Commitment items.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Curt M. Adams is the Linda Clarke Anderson Presidential Professor at the University of Oklahoma and co-director of the Oklahoma Center for Educational Policy. He conducts research on the social-psychology of school systems, performance measurement, leadership, accountability, and improvement science.

Ryan C. Miskell is a research associate with WestEd where he conducts research related to issues and factors affecting teaching, learning, and school improvement.

References

Adams

C. M.

(2010). Social determinants of student trust in high poverty elementary schools. In Hoy

W. K.

&M. D. (Eds.), Analyzing school contexts: Influences of principals and teachers in the service of students (pp. 65-85). Charlotte, NC: Information Age.

Adams

C. M.

(2013). Collective trust: A social indicator of instructional capacity. Journal of Educational Administration, 51(3), 1-36.

Adams

C. M.

Forsyth

P. B.

(2009). Conceptualizing and validating a measure of student trust. In Hoy

DiPaola

(Eds.), Improving schools: Studies in leadership and culture (pp. 186-206). Charlotte, NC: Information Age.

Angle

H. L.

Perry

J. L.

(1981). An empirical assessment of organizational commitment and organizational effectiveness. Administrative Science Quarterly, 26, 296-319.

Baier

A. C.

(1986). Trust and antitrust. Ethics, 96, 231-260.

Barber

(1983). The logic and limits of trust. New Brunswick, NJ: Rutgers University Press.

Bryk

A. S.

Gomez

L. M.

Grunow

LeMahieu

P. G.

(2015). Learning to improve. How America’s schools can get better and getting better. Cambridge, MA: Harvard Education Press.

Bryk

A. S.

Schneider

(2002). Trust in schools: A core resource for improvement. New York, NY: Russell Sage Foundation.

Butler

J. K.

Cantrell

S. R.

(1984). A behavioral decision theory approach to modeling dyadic trust in superiors and subordinates. Psychological Reports, 55, 19-28.

10.

Chenowith

(2007). It’s being done: Academic success in unexpected schools. Cambridge, MA: Harvard Education Press.

11.

Cohen

Doveh

Eick

(2001). Statistical properties of the rwg(i) index of agreement. Psychological Methods, 6, 297-310.

12.

Cosner

(2009). Building organizational capacity through trust. Educational Administration Quarterly, 45, 248-291.

13.

Cronbach

L. J.

(1971). Test validation. In Thorndike

R. L.

(Ed.), Educational measurement (pp. 443-507). Washington, DC: American Council on Education.

14.

Daly

A. J.

Liou

Moolenaar

N. M.

(2014). The principal connection: Trust and innovative climate in a network of reform. In Van Maele

Forsyth

P. B.

Van Houte

(Eds.), Trust and school life: The role of trust for learning, teaching, leading, and bridging (pp. 285-312). New York, NY: Springer.

15.

Daly

A. J.

Moolenaar

N. M.

Bolivar

J. M.

Burke

(2010). Relationships in reform: The role of teachers’ social networks. Journal of Educational Administration, 48, 359-391.

16.

Darling-Hammond

(2005). Policy and change: Getting beyond bureaucracy. In Hargreaves

(Ed.), Extending educational change: International handbook of educational change (pp. 362-387). Dordrecht, Netherlands: Springer.

17.

Darling-Hammond

Hightower

A. M.

Husbands

J. L.

LaFors

J. R.

Young

V. M.

Christopher

(2006). Building instructional quality: Inside-out and outside-in: Perspective on San Diego’s school reform. In Harris

Chrispeels

(Eds.), Improving schools and educational systems (pp.129-185). New York, NY: Routledge.

18.

Firestone

W. A.

(2009). Culture and process in effective school districts. In Hoy

W. K.

DiPaola

M. F.

(Eds.), Studies in school improvement (pp. 177-204). Raleigh-Durham, NC: Information Age.

19.

Firestone

W. A.

Pennell

J. R.

(1993). Teacher commitment, working conditions, and differential inventive policies. Review of Educational Research, 63, 489-525.

20.

Forsyth

P. B.

Adams

C. M.

(2014). Organizational predictability, the school principal, and achievement. In Van Maele

Van Houtte

Forsyth

(Eds.), Trust and school life: The role of trust for learning, teaching, leading, and bridging (pp. 83-98). New York, NY: Springer.

21.

Forsyth

P. B.

Adams

C. M.

Hoy

W. K.

(2011). Collective trust: Why schools can’t improve without it. New York, NY: Teachers College Press.

22.

Fullan

(2008). Six secrets of change: What the best leaders do to help their organizations survive and thrive. San Francisco, CA: Wiley.

23.

Fullan

(2010). All systems go: The change imperative for whole system reform. Thousand Oaks, CA: Corwin.

24.

Fullan

(2014). The principal: Three keys to maximizing impact. San Francisco, CA: Jossey-Bass.

25.

Gabarro

J. J.

(1987). The development of working relationships. In Lorsch

(Ed.), Handbook of organizational behavior (pp.172-189). Engle-wood Cliffs, NJ: Prentice-Hall.

26.

Garvin

D. A.

Edmondson

A. C.

Gino

(2008, March). Is yours a learning organization? Harvard Business Review, 3-11.

27.

Glisson

James

L. R.

(2002). The cross-level effects of culture and climate in human service teams. Journal of Organizational Behavior, 23, 767-794.

28.

Hargreaves

Shirley

(2009). The fourth way: The inspiring future for educational change. Thousand Oaks, CA: Corwin Press.

29.

Hargreaves

D. H.

(2011). System redesign for system capacity building. Journal of Educational Administration, 49, 685-700.

30.

Harris

(2011). System improvement through collective capacity building. Journal of Educational Administration, 49, 624-636.

31.

Harris

Chrispeels

J. H.

(2006). Improving schools and educational systems: International perspective s. New York, NY: Routledge.

32.

Henkin

A. B.

Holliman

S. L.

(2009). Urban teacher commitment: Exploring associations with organizational conflict, support for innovation, and participation. Urban Education, 44, 160-180.

33.

Hoffman

J. D.

Sabo

Bliss

Hoy

W. K.

(1994). Building a culture of trust. Journal of School Leadership, 4, 484-501.

34.

Honig

M. I.

Hatch

T. C.

(2004). Crafting coherence: How schools strategically manage multiple, external demands. Educational Researcher, 33(8), 16-30.

35.

Hoy

W. K.

Tarter

C. J.

(2004). Organizational justice in schools: No justice without trust. International Journal of Educational Management, 18, 250-279.

36.

Hoy

W. K.

Tschannen-Moran

(1999). Five faces of trust: An empirical confirmation in urban elementary schools. Journal of School Leadership, 9, 184-208.

37.

King

B. M.

Bouchard

(2011). The capacity to build organizational capacity in schools. Journal of Educational Administration, 49, 653-669.

38.

Kolenikov

Bollen

K. A.

(2012). Testing negative error variances: Is a Heywood case a symptom of misspecification? Sociological Methods & Research, 41, 124-167.

39.

Kushman

J. W.

(1992). The organizational dynamics of teacher workplace commitment: A study of urban elementary and middle schools. Educational Administration Quarterly, 28, 5-42.

40.

Lee

J. C.

Zhang

Yin

(2011). A multilevel analysis of the impact of a professional learning community, faculty trust in colleagues, and collective efficacy on teacher commitment to students. Teaching and Teacher Education, 27, 820-830.

41.

Leithwood

Louis

K. S.

Anderson

Wahlstrom

(2004). How leadership influences student learning. New York, NY: The Wallace Foundation. Available at: http://www.wallacefoundation.org/knowledge-center/school-leadership/key-research/Documents/How-Leadership-Influences-Student-Learning.pdf

42.

Levin

(2008). Thinking about knowledge mobilization. Paper presented at the Canadian Council on Learning and the Social Sciences and Humanities Research Council, Vancouver, Canada.

43.

Makiewicz

Mitchell

(2014). Teacher trust in the principal: Factor structure and effects. In VanMaele

Forsyth

P. B.

Van Houte

(Eds.), Trust and school life: The role of trust for learning, teaching, leading, and bridging (pp. 99-120). New York, NY: Springer.

44.

McLaughlin

M. W.

(1990). The rand change agent study revisited: Macro perspectives and micro realities. Educational Researcher, 19(6), 11-16.

45.

Meade

A. W.

Watson

A. M.

Kroustalis

C. M.

(2007). Assessing common methods bias in organizational research. Paper presented at the 22nd Annual Meeting of the Society for Industrial and Organizational Psychology, New York, NY.

46.

Messick

(1995). Validity of psychological assessment: Validation of inferences from persons’ responses and performances as scientific inquiry into school meaning. American Psychologist, 50, 741-749.

47.

Milanowski

A. T.

Heneman

H. G.

(2001). Assessment of teacher reactions to a standards-based teacher evaluation system: A pilot study. Journal of Personnel Evaluation in Education, 15, 193-212.

48.

Miller

D. M.

(2008). Data for school improvement and educational accountability: reliability and validity in practice. In Ryan

Shepard

(Eds.), The future of test-based educational accountability (pp. 249-262). New York, NY: Routledge.

49.

Mishra

A. K.

(1996). Organizational responses to crises: The centrality of trust. In Kramer

Tyler

(Eds.), Trust in organizations: Frontiers of theory and research (pp. 261-287). Thousand Oaks, CA: Sage.

50.

Moss

P. A.

(1995). Themes and variations in validity theory. Educational Measurement: Issues and Practice, 14(2), 5-12.

51.

Mourshed

Chinezi

Barber

(2010). How the world’s most improved school systems keep getting better. New York, NY: McKinsey.

52.

Mowday

R. T.

Porter

L. W.

Steers

R. M.

(1982). Employee-organization linkages: The psychology of commitment, absenteeism and turnover. New York, NY: Academic Press.

53.

Nordin

A. R.

Darmawan

G. N.

Keeves

J. P.

(2009). Teacher commitment. In Saha

Dworkin

(Eds.), International handbook of research on teachers and teaching (pp. 343-360) New York, NY: Springer.

54.

Ouchi

W. G.

(1981). Theory Z: How American business can meet the Japanese challenge. Reading, MA: Addison-Wesley.

55.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J. Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903.

56.

Porter

L. W.

Steers

R. M.

Mowday

R. T.

Bouhan

P. V.

(1974). Organizational commitment, job satisfaction, and turnover among psychiatric technicians. Journal of Applied Psychology, 59, 603-609.

57.

Reyes

(1992). Preliminary models of teacher organizational commitment: Implications for restructuring the workplace. Madison, WI: Center on Organization and Restructuring of Schools.

58.

Romero

(2010). Student trust: Impacting high school outcomes (Unpublished doctoral dissertation). University of California Riverside.

59.

Ross

J. A.

Gray

(2006). Transformational leadership and teacher commitment to organizational values: The mediating effects of collective teacher efficacy. School Effectiveness and School Improvement, 17, 179-199.

60.

Rotter

J. B.

(1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35, 651-665.

61.

Schein

E. H.

(1996). Culture: The missing concept in organization studies. Administrative Science Quarterly, 41, 229-240.

62.

Schneider

Judy

Ebmeye

Broda

(2014). Trust in elementary and secondary urban schools: A pathway for student success and college ambition. In VanMaele

Forsyth

P. B.

Van Houte

(Eds.), Trust and school life: The role of trust for learning, teaching, leading, and bridging (pp. 37-56). New York, NY: Springer.

63.

Schreiber

J. B.

Nora

Stage

F. K.

Barlow

E. A.

King

(2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. Journal of Educational Research, 99, 323-338.

64.

Shamir

Lapidot

(2003). Trust in organizational superiors: Systemic and collective considerations. Organization Studies, 24, 463-491.

65.

Sharrat

Fullan

(2009). Realization: The change imperative for deepening district wide reform. San Francisco, CA: Corwin Press.

66.

Spillane

J. P.

Reiser

B. J.

Reimer

(2002). Policy implementation and cognition: Reframing and refocusing implementation research. Review of Educational Research, 72, 387-431.

67.

Thompson

Daniel

L. G.

(1996). Factor analytic evidence for the construct validity of scores: A historical overview and some guidelines. Educational and Psychological Measurement, 56, 213-224.

68.

Tschannen-Moran

(2004). Trust matters: Leadership for successful schools. San Francisco, CA: Jossey-Bass.

69.

Tschannen-Moran

(2014). Trust matters: Leadership for successful schools (2nd ed.). San Francisco, CA: Jossey-Bass.

70.

Tschannen-Moran

Hoy

W. K.

(1998). A conceptual and empirical analysis of trust in schools. Journal of Educational Administration, 36, 334-352.

71.

Tsui

K. T.

Cheng

Y. C.

(1999). School organizational health and teacher commitment: A contingency study with multi-level analysis. Educational Research and Evaluation: An International Journal on Theory and Practice, 5, 249-268.

72.

Van Maele

Van Houtte

Forsyth

P. B.

(2014). Introduction: Trust as a matter of equity and excellence in education. In VanMaele

Forsyth

P. B.

Van Houte

(Eds.), Trust and school life: The role of trust for learning, teaching, leading, and bridging (pp. 1-36). New York, NY: Springer.

73.

Ware

H. W.

Kitsantas

(2011). Predicting teacher commitment using principal and teacher efficacy variables: An HLM approach. Journal of Educational Research, 104, 183-193.

74.

Waters

T. J.

Marzano

R. J.

(2006). School district leadership that works: The effect of superintendent leadership on student achievement. Denver, CO: MCREL.

75.

Zand

(1972). Trust and managerial problem solving. Administrative Science Quarterly, 17(2), 229-239.

76.

Zand

(1997). The leadership triad: Knowledge, trust, and power. New York, NY: Oxford University Press.