Hart, Carter, Conn, Dingle & Wearing Teacher Morale and Organisational Climate DEVELOPMENT OF THE SCHOOL ORGANISATIONAL HEALTH QUESTIONNAIRE: A MEASURE FOR ASSESSING TEACHER MORALE AND SCHOOL ORGANISATIONAL CLIMATE * Peter M. Hart, Nicholas L. Carter, Michael Conn, Roger K. Dingle and Alexander J. Wearing Department of Psychology University of Melbourne Parkville Vic 3052 Australia Paper presented at the 1993 National Conference of the Australian Association for Research in Education, Freemantle, November 22 to 25. Correspondence should be addressed to Peter M. Hart, Department of Psychology, University of Melbourne, Parkville, Victoria, 3052, Australia. Phone 61-3-344-6353, Fax 61-3-347-6618 * Preparation of this article was supported by the Victorian Health Promotion Foundation. The cooperation and support of the Victorian Department of School Education is gratefully acknowledged. ABSTRACT This study reports on the development of the School Organisational Health Questionnaire; a measure for assessing teacher morale and school organisational climate. As part of an organisational health program conducted by the Victorian Department of School Education, there was a need to develop a psychometrically sound instrument that would provide the basis for organisational change and provide a means for evaluating the effectiveness of organisational development programs. Data were provided by 1,520 Victorian teachers in 18 primary and 26 secondary schools. A series of exploratory and confirmatory factor analyses resulted in a 54 item questionnaire that measures teacher morale and 11 separate dimensions of school organisational climate: curriculum consultation, discipline policy, excessive work demands, feedback, goal congruence, participative decision making, professional development, professional interaction, student orientation, supportive leadership, and role clarity. Correlations with job satisfaction and quality of work life indices supported the questionnaire's construct validity. Comparisons between primary and secondary school teachers showed that primary teachers reported higher levels of morale and tended to have more favourable perceptions of their schools' organisation. The way in which the School Organisational Health Questionnaire has been used to facilitate school improvement is also discussed. DEVELOPMENT OF THE SCHOOL ORGANISATIONAL HEALTH QUESTIONNAIRE: A MEASURE FOR ASSESSING TEACHER MORALE AND SCHOOL ORGANISATIONAL CLIMATE During 1991 the Victorian Department of School Education conducted several organisational development programs that aimed to enhance teacher quality of work life (Carter, Hart & Wearing, 1993). In order to evaluate these programs and provide participating schools with diagnostic information about their schools' organisation, we developed a questionnaire that assesses teacher morale and various dimensions of organisational climate. This paper reports three studies conducted during the development of the School Organisational Health Questionnaire. As a result of a major study into teacher stress (Wearing, Bell, McMurray, Conn & Dudgeon, 1990) and the initiative of a statewide network of organisational consultants, the Department of School Education recognised the link between organisational health and teacher well-being (Cox, 1988, 1992; Cox, Boot, Cox & Harrison, 1988). This led to the development of various intervention programs designed primarily to improve organisational behaviour and human resource management (e.g., leadership, decision-making, policy formation and professional development) within individual schools (e.g., Carter et al., 1993; Hart, Wearing & Conn, 1993). It was expected that these programs would influence teachers' affective response to their work by reducing psychological distress and improving morale (cf. Hart, 1992). In turn, these changes were expected to result in a reduction of stress related WorkCare claims (Wearing et al., 1990) and more effective teaching (e.g., Cheng, 1993; Conley, Bacharach & Bauer, 1989; Rosenholtz, 1989). Teacher affect is an important determinant of student learning outcomes (Rowe, 1990; 1992) and, more recently, Hill, Holmes-Smith and Rowe (1993) have proposed a model linking school organisational climate and teacher affect to student academic achievement. Moreover, through the use of multilevel analyses, Hill et al. have demonstrated that after controlling for student-level intake factors, between 29 and 46 percent of the variance in students' literacy and numeracy skills can be accounted for by teacher level effects. It is therefore important to establish the determinants of teachers' affective responses to their work.. When investigating the determinants of teacher affect, educational researchers have tended to focus on the relationship between negative experiences (stressors) and psychological distress (e.g., Wearing et al., 1990). This approach fails to recognise that positive experiences and morale operate independently from negative experiences and psychological distress (Hart et al., in press, 1993; cf. Headey & Wearing, 1992), and that psychological distress and morale make equal contributions to teacher quality of work life (Hart, 1992; cf. Hart & Wearing, 1993). As noted by Evans (1992), however, teacher morale has attracted little systematic empirical research. Smith (1966) has referred to morale as a group phenomena that exists when there is persistence and energy, cohesion and cooperation, and enthusiastic striving among teachers. Although morale is often viewed as a group phenomenon (Milton, Entrekin & Stening, 1984; Smith, 1966; 1976), a growing number of researchers recognise that the individual experience of morale is psychologically more meaningful (Doherty, 1988; Evans, 1992). This recognition, however, has not yet led to the development of adequate measures for assessing teacher morale. Although researchers are beginning to view morale as an individual psychological state (Doherty, 1988; Evans, 1992) somewhat akin to positive affect (Watson, 1988; cf. Hart, 1992), measures of morale have continued to assess teachers' perceptions of their environment and its effect upon them. For example, the Staff Morale Questionnaire (Meaney & Smith, 1988; Smith, 1976; Wong, 1991) employs items such as "Every teacher on this staff contributes towards the achievement of the school's aims", "There is no complaining, arguing and taking of sides among my colleagues", and "Keeping up to date professionally is too much of a burden" to assess three dimensions of morale referred to as cohesive pride, personal challenge and leadership synergy. These items might assess factors contributing to morale, but they say nothing about the psychological state of individual teachers. Moreover, Smith (1987) has noted that whilst the questionnaire has been widely used, a major shortcoming of this work has been a failure to differentiate the causes from the manifestations of morale. More recently, Doherty (1988) developed a measure of teacher morale that was referred to as "a number of heterogeneous items related to adjustment and mental health" (p. 72). This measure was based on the notion that morale is a unitary concept which reflects how people feel at a particular point in time. Many items, however, appear to be similar to those used in measures of neuroticism (e.g., "People don't take much notice of me", "Sometimes I feel that life is pretty meaningless", and "I'm not really a very happy kind of person")(cf. Costa & McCrae, 1985) and are more likely to tap teachers' dispositional levels of negative affect (Watson, 1988), rather than a positive psychological state. This is not surprising, given that the scale was development from an item pool that measured, among other things, anxiety, stability of self-concept, experience of psychosomatic symptoms and depression. We agree that morale is a unitary construct, but believe that it is more accurately associated with the energy, enthusiasm, team spirit and pride teachers experience in their school (Hart & Conn, 1992). These adjectives mirror Smith's (1966) description of morale, and are similar to those used by Watson (1988) in defining positive affect. In order to identify the causes of teacher morale, it is necessary to assess separately morale and its determinants. There is considerable evidence to suggest that organisational factors (e.g., leadership and communication) are equally, if not more important than classroom specific issues (e.g., student misbehaviour) in determining psychological outcomes among teachers (e.g., Borg, 1990; Kyriacou, 1987; cf. Hart, 1992; Hart, Conn, Carter & Wearing, 1993; Wearing et al., 1990). Despite this evidence, however, there were no measures available to assess teachers' perceptions of the organisational factors being addressed by the Department of School Education's organisational development programs. These programs assumed that schools were formal organisations, and therefore needed to concern themselves with organisational behaviour and human resource management issues such as decision-making, provision of feedback to staff, goal congruency, peer and leadership support, policy formation, professional development and role clarity (e.g., Milton et al., 1984; Schuler, Dowling, Smart & Huber, 1992). Although a large number of scales have been developed to assess various aspects of school organisational climate (Owens, 1991; Rentoul & Fraser, 1983), none measure this range of issues. One of the most widely used measures of school organisational climate is the Organizational Climate Description Questionnaire (Haplin & Croft, 1963; cf. Kottkamp, Mulhern & Hoy, 1987). Unfortunately, this questionnaire is restricted to the assessment of social interaction between teachers and the principal. Similarly, the School Level Environment Questionnaire developed in Australia by Rentoul and Fraser (1983) and recently reported by Fisher and Fraser (1990), was designed to measure eight dimensions of schools' organisational climate. Although the items that make up these dimensions appear to assess interpersonal relationships (e.g., staff affiliation and student support) and some aspects of organisational behaviour (e.g, participatory decision making and resource adequacy), they do not assess important aspects of human resource management such as role clarity, the provision of feedback and professional development. More importantly, there was little evidence to support the questionnaire's construct validity. Whilst the internal consistency coefficients and correlations between different dimensions were encouraging, these were based on three small samples of teachers. Moreover, the 56 items had not been factor analysed and there was no evidence to show that the dimensions measured the constructs they purported to assess. Given the absence of appropriate measures, the aim of the three studies reported in this paper was to develop a psychometrically sound questionnaire that assessed teacher morale and various dimensions of school organisational climate. Although some school specific issues had to be assessed (e.g., curriculum consultation, effective discipline policy and student orientation), it was important that the questionnaire taped general issues related to organisational behaviour and human resource management. As noted by DeFrank (1988), occupational measures should assess conditions peculiar to the occupational group under investigation, as well as factors that are common to all organisations. In order to successfully develop this questionnaire, it was necessary to achieve the following five goals: (a) to provide a measure of teacher morale and various dimensions of school organisational climate that enabled assessment of important organisational behaviour and human resource management issues; (b) to demonstrate that morale and the various dimensions of organisational climate could be distinguished from one another when subjected to factor analysis; (c) to demonstrate through internal consistency coefficients that items within each dimension measured a single construct; (d) to demonstrate that each dimension measured the construct that it purported to assess, by correlating these dimensions with other theoretically related scales; and, (e) to demonstrate that the questionnaire could effectively be used to facilitate and evaluate school improvement programs. Study 1: Initial Development of the School Organisational Health Questionnaire Method Participants The data were derived from responses to a pretest survey used to evaluate an organisational development program conducted by the Victorian (Australia) Department of School Education during early 1991. All teachers working in the schools that took part in this program were invited to participate in the evaluation. Completed questionnaires were provided by 615 teachers (response rate: 92%)1 from 5 primary (N = 97) and 11 secondary (N = 518) schools in both country and metropolitan regions. Teachers' ages ranged from 22 to 63 years (M = 38.3, SD = 8.51), and 50.5% were female. The age and gender profiles were similar to that found in the Department as a whole. Measures Morale and School Organisational Climate. The School Organisational Health Questionnaire was developed to provide a measure of teacher morale and school organisational climate. A large number items was generated from a review of the education literature and the experience of organisational health practitioners working in the Victorian Department of School Education. These items were divided into conceptual groups on the basis of their content (e.g., morale, decision-making, leadership), and each group of items was examined for clarity and conceptual overlap. This resulted in several items being either reworded or discarded. A pool of 80 items was used during the development of the questionnaire. Given that the School Organisational Health Questionnaire was designed to facilitate school improvement, there was a need to develop 'Actual' and 'Preferred' forms (cf. Fisher & Fraser, 1990). The instructions for the Actual form read as follows: "Listed below are a number of statements that could be used to describe some aspects of your school. Please read each statement carefully, and indicate the extent to which the statement actually applies to your school". Teachers were asked to respond on a 5-point scale, anchored by "strongly disagree" and "strongly agree". The same response scale was used for the Preferred form, but the instructions read as follows: "We would now like you to indicate how the following statements describe the way you would prefer your school to be. The statements listed below are very similar to the statements you have just completed. Please do not worry about this, but simply indicate the extent to which the statement describes the way you would prefer your school to be." Each of the 80 items was reworded for the Preferred form, generally by inserting the word 'should'. For example, the item "Teachers go about their work with enthusiasm" became "Teachers should go about their work with enthusiasm". The results reported in these studies are for the Actual items only. It was not possible to test all 80 items in this study. The data were collected primarily for the purpose of program evaluation and to identify issues relevant to the content area of the organisational development program. Consequently, 55 items were used that on face validity were believed to measure Morale and the organisational climate dimensions of Feedback, Goal Congruence, Ministry Demands, Participative Decision-Making, Professional Development, Professional Interaction, Role Clarity, Social Cohesion and Supportive Leadership. The items from each dimensions were spread evenly throughout the questionnaire. Results and Discussion The 55 items were subjected to a series of exploratory principal components analyses, using the SPSS/PC+ statistical package. Due to limitations on the number of variables that can be analysed with the SPSS/PC+ factor procedure, the 55 items were divided into two groups. Those items that were thought to measure Feedback, Morale, Professional Interaction, Social Cohesion and Supportive Leadership were placed in the first group, and the remaining items were placed in the second. Principal components analyses were conducted on the two groups of items, using an oblimin rotation to allow for the expected correlation between different dimensions. Throughout the analyses Kaiser's criterion was used to extract an appropriate number of factors, and loadings below .30 were considered 'nonsignificant'. Items were progressively deleted from each group on the basis of either nonsignificant or conceptually unsound loadings (e.g., some items loaded on factors where the relationship could not be supported theoretically). This led to 38 items being retained in the final two solutions; 20 items in the first group and 18 in the second. There were four factors in each group with an eigenvalue greater than unity. The four factors extracted in the first group (ordered according to their eigenvalues) were Morale, Feedback, Professional Interaction and Supportive Leadership. This solution accounted for 64.4% of the variance, and the correlations between the factors ranged from .28 to .44 (M = .38, SD = .06) in absolute value. Only one item significantly crossloaded, with the secondary loading being smaller than the primary loading. The four factors extracted in the second group were Goal Congruence, Professional Development, Participative Decision-Making and Role Clarity. This solution accounted for 54.3% of the variance, and the correlations between the factors ranged from .19 to .37 (M = .29, SD = .06) in absolute value. Two items from the Goal Congruence dimension significantly crossloaded on the Role Clarity dimension. An overall analysis, constrained to extract eight factors, failed to replicate the factor structure suggested by the two separate solutions. We should also note that the items for the Ministry Demands and Social Cohesion dimensions failed to load on their respective factors, and these dimensions were excluded from subsequent investigations of the questionnaire's structure. The Linear Structural Relations (LISREL VII) program (Joreskog & Sorbom, 1989a) was used to further examine the factor structure of the 38 items identified through the exploratory analyses. Confirmatory factor analyses were conducted on the two groups of items separately, in an attempt to replicate our earlier findings. The 38 items were also combined into a single analysis, to ensure that there was no crossover between the two sets of items. In all analyses, it was assumed that the factors were correlated. The three confirmatory analyses were conducted using four different estimation procedures. This enabled assumptions about the distributional properties of the items to be varied, and provided a comparison between the most common estimation techniques used in structural equation modeling. The four estimation procedures used in these analyses were: (a) maximum likelihood estimation based on a Pearson product- moment correlation matrix, (b) maximum likelihood estimation based on a variance-covariance matrix, (c) generalised least squares estimation based on a variance-covariance matrix, and (d) weighted least squares estimation based on the asymptotic covariances of a polychoric correlation matrix. The first three procedures are based on normal theory estimators, and assume that the items were measured on interval level scales. Although the skewness and kurtosis for all items was less than one in absolute value, multivariate normality cannot easily be demonstrated. Moreover, the items were measured on 5-point ordinal scales. It was therefore appropriate, in terms of the distributional assumptions being made, to examine the asymptotic covariances of a polychoric correlation matrix using the generally weighted least-squares method of estimation. The goodness-of-fit statistics for the confirmatory analyses are shown in Table 1. ______________________ Insert Table 1 about here ______________________ The goodness-of-fit statistics suggested a reasonable fit between the observed data and theoretical factor structure. Although the likelihood ratio test statistic provides the only true parametric test of a model's fit (Cuttance, 1987), this statistic is strongly influenced by sample size and departures from multivariate normality. Anderson and Gerbing (1984) have also noted that the goodness-of-fit index (GFI) is biased downwards as the number of indicators and latent constructs in a particular model increases. This can be seen in the current analyses by comparing the GFI for the four and eight factor solutions, and was particularly evident when maximum likelihood estimation was used. More recently, Gerbing and Anderson (1992) have argued that the relative noncentrality index (RNI)(McDonald and Marsh, 1990) provides the best incremental fit index. The RNI is unaffected by sample size, and compares the model under investigation with a null model that assumes no relationship between the observed variables. As shown by these analyses, however, the RNI also appears to biased downwards as the number of indicators and latent constructs increases. Moreover the GFI and RNI seem to be influenced by the estimation procedure used in a particular analysis. The factor loadings and standardised item reliabilities derived from the exploratory and confirmatory analyses are shown in Table 2. The parameter estimates for the confirmatory analyses are based on the combined analysis of all 38 items. These estimates were very similar to those obtained in the separate analyses of the Group 1 and 2 items. This was a little surprising for the weighted least squares estimation procedure. The estimation of asymptotic covariances requires very large samples. According to Joreskog and Sorbom (1989a), at least 630 and 513 teachers respectively was needed for the analysis of the Group 1 and 2 items, whilst 2,223 teachers was required for the analysis of all 38 items. The similarity in the estimates obtained for the weighted least squares solutions, however, suggests that this procedure may yield stable estimates when used with smaller than ideal samples. ______________________ Insert Table 2 about here ______________________ All parameter estimates in the confirmatory analyses were significant at the .001 level, and the standard errors were all less than or equal to .05. The factor loadings tended to be slightly higher in the weighted least squares solution, with 89% of the item reliabilities being greater than .55 (i.e., at least 30% shared variance between the item and its latent construct). Between 79% and 82% of the factor loadings were greater than .55 for the other solutions. The overall pattern of factor loadings, however, was similar in all solutions. Moreover, the estimated true correlations between the 8 factors also tended to be similar across solutions. The correlations ranged from .48 to .85 (M = .66, SD = .08), suggesting that there was a moderate to strong relationship between Morale and the various dimensions of organisational climate. Despite the strength of these correlations, an eight factor model provided a better explanation of the data than other models in which a smaller number of factors was estimated. The results of this study suggest that 38 items of the School Organisational Health Questionnaire adequately measure Morale and seven different dimensions of organisational climate. These analyses, however, were based on just 55 of the 80 items in the total item pool. Moreover, these items were concerned with general organisational issues, rather than those issues specifically associated with the nature of schooling. The aim of the second study, therefore, was to examine the items associated with school specific issues (i.e, school's discipline policy, curriculum consultation and teachers orientation toward students). Study 2: School Specific Dimensions of the School Organisational Health Questionnaire Method Participants The data were derived from responses to a pretest survey used to evaluate an organisational development program conducted by the Victorian (Australia) Department of School Education during early 1991. All teachers working in the schools that took part in this program were invited to participate in the evaluation. Completed questionnaires were provided by 342 teachers (response rate: 91%) from 1 primary (N = 18) and 6 secondary (N = 324) schools in the metropolitan area. Teachers' ages ranged from 21 to 61 years (M = 38.4, SD = 8.77), and 53.9% were female. Measures Morale and School Organisational Climate. Thirty-four items from the item pool of the School Organisational Health Questionnaire were used to measure Morale and seven dimensions of organisational climate. The 5 items measuring Morale were the same as those used during Study 1, as were 12 of the 13 items used to assess Participative Decision-Making, Professional Interaction, and Supportive Leadership. A change was made to one item from the Supportive Leadership dimension because of the small factor loading obtained for this item during Study 1, and to make this item read in the same direction as other items in the dimension. The item was changed from "The school's administrators don't really know the problems faced by teachers" to "The school's administrators know the problems faced by teachers in this school". A further 16 items, which had not been examined during Study 1, were used to assess Curriculum Consultation, Effective Discipline Policy, Excessive Work Demands, and Student Orientation. Since the data were collected as part of an organisational development program, it was not possible to examine all 80 items from the item pool. Teachers were asked to respond to the 34 items used in this study according to the instructions and response format outlined in Study 1. Results and Discussion The 34 items of the School Organisational Health Questionnaire were subjected to a series of exploratory principal components and confirmatory factor analyses. In all analyses, it was assumed that the factors were correlated. Given that listwise deletion of missing data resulted in 287 teachers remaining in the analyses, the items were divided into two groups to maintain at least 10 subjects per item. Group 1 included those items associated (in terms of theoretical content) with Morale, Professional Interaction, Supportive Leadership and Participative Decision-Making. With the exception of one item in Supportive Leadership, these items were identical to those used during Study 1. The items in Group 2 were associated with Excessive Work Demands, Curriculum Consultation, Effective Discipline Policy and Student Orientation. The principal components analyses, conducted separately on the Group 1 and 2 items, failed to produce a solution that was consistent with the expected factor structure. Although the eigenvalues for each analysis suggested a four factor solution, the fourth factor in each analysis was based on a single item ("The school's administrators know the problems faced by teachers" in Group 1, and "There is time for teachers to relax in this school" in Group 2). After the removal of these items, both analyses suggested a three factor solution. For Group 1, the analysis failed to separate the Participative Decision-Making and Supportive Leadership dimensions, while for Group 2, the analysis failed to separate the Effective Discipline Policy and Student Orientation dimensions. The failure to separate these dimensions may have resulted from limitations associated with principal components analysis, which has difficulty assigning highly correlated items to separate theoretically related factors. Confirmatory factor analyses were therefore conducted, using the four estimation procedures outlined in Study 1. Separate analyses were conducted on the Group 1 and 2 items (four factors were estimated in each analysis), and these were combined in an overall analysis to ensure that there was no overlap between the eight factors. The goodness-of-fit statistics for these analyses are shown in Table 3. ______________________ Insert Table 3 about here ______________________ The goodness-of-fit statistics suggest a reasonable fit between the observed data and theoretical factor structure. As shown in Study 1, the GFI seemed to be biased downwards as the number of indicators and latent constructs increased. The separate analyses of Group 1 and 2 items yielded a GFI that was well within acceptable limits for all estimated solutions (Cuttance, 1987). This compares with the overall analysis which, for solutions based on a Pearson product-moment and variance- covariance matrix, yielded a GFI that would ordinarily be questionable. These results are consistent with Anderson and Gerbing's (1984) findings, derived through Monte Carlo studies, that suggest large structural equation models should not be rejected merely because their GFI is less than .90. Although further Monte Carlo studies are needed before accurate guidelines can be set, these analyses suggest that GFIs between .80 and .90 can be expected for adequately fitting models which have a large number of indicators and latent constructs. This argument is supported by the likelihood ratio test statistics, RMSR's and RNI's, which tend to support the fit of the reported models (see Cuttance & Ecob, 1987; Loehlin, 1992). Given Gerbing and Anderson's (1992) recommendation that the RNI provides the best measure of relative fit, it is worth noting that the RNI also appeared to be biased downwards as the number of indicators and latent constructs increased. The standardised item reliabilities (factor loadings) for the confirmatory factor analysis conducted on all 34 items are shown in Table 4. These were very similar to those obtained during the analyses conducted separately on the Group 1 and 2 items. Generally speaking, the factor loadings obtained from the four factor solutions were either identical or up to .03 higher than those reported in Table 4. ______________________ Insert Table 4 about here ______________________ All factor loadings, except those for one item in the Excessive Work Demands and Supportive Leadership dimensions, were significant at the .001 level. Interestingly, the factor loading for the reworded item in the Supportive Leadership dimension ("The school's administrators know the problems faced by teachers in this school") suggested that the original wording employed during Study 1 was more effective. It might be that teachers are used to thinking about the content of this item in the negative. The factor loadings reported here showed that there was less than 4% shared variance between this item and the latent construct of Supportive Leadership. Two other items also failed to relate adequately to their respective latent constructs. More than 95% unique variance was found in one of the Excessive Work Demands items ("There is time for teachers to relax in this school") and there was over 78% unique variance in one of the Effective Discipline Policy items ("The rules and sanctions relating to discipline are enforced in a consistent fashion in this school"). It will be shown in Study 3, however, that these items work effectively when worded in the opposite direction. As a rule of thumb, items should have less than 70% unique variance to be considered adequate (cf. Cuttance (1987) who recommends less than 90%). Although the factor loadings varied between the different solutions, the overall pattern remained consistent and demonstrated that the vast majority of items were good indicators of their respective latent constructs. The estimated true correlations suggested that Excessive Work Demands was not related to the other dimensions. The vast majority of correlations involving Excessive Work Demands were nonsignificant (p > .05), with just a few being marginally significant in the generalised least squares and weighted least squares solutions. There was a moderate to strong relationship between the other dimensions, with correlations ranging from .51 to .85 (M = .69, SD = .09). The results of Studies 1 and 2 suggest that 54 items from the School Organisational Health Questionnaire provide good indicators of teacher Morale and 11 separate dimensions of school organisational climate. Whilst these studies supported the questionnaire's hypothesised factor structure, all items had not been used in a single study. Moreover, there was still a need to examine the relationships between these dimensions and other theoretically related constructs. A third study, therefore, aimed to simultaneously examine these 54 items, and relate the different dimensions of the School Organisational Health Questionnaire to job satisfaction and quality of work life indices. Study 3: Combining Organisational and School Specific Dimensions of the School Organisational Health Questionnaire Method Participants The data were derived from responses to a pretest survey used to evaluate an organisational development program conducted by the Victorian (Australia) Department of School Education during late 1991. All teachers working in the schools that took part in this program were invited to participate in the evaluation. Completed questionnaires were provided by 563 teachers (response rate: 92%) from 12 primary (N = 209) and 9 secondary (N = 343) schools in both country and metropolitan regions. Teachers' ages ranged from 21 to 62 years (M = 39.7, SD = 8.30), and 56.2% were female. Measures Quality of Work Life. Teacher quality of work life refers to the judgements that teachers make about the extent to which their work is satisfying and meeting their needs (cf. Efraty & Sirgy, 1990). It reflects the overall impression that teachers have about their work, and is determined equally by teachers' level of psychological distress and morale (Hart, 1992). Quality of work life was measured with a 5-item scale adapted from Diener, Emmons, Larsen & Griffin's (1985) Satisfaction With Life Scale by replacing the word 'life' in each item with the phrase 'life at work'. For example, the item "I am satisfied with my life" became "I am satisfied with my life at work". Teachers were asked to rate their level of agreement with each of the five items on a 7-point scale ranging from 'strongly disagree' to 'strongly agree' (alpha = .87). Psychological Distress. The General Strain Index (Tellenback, Brenner & Lofgren, 1983) was used to measure the aversive feelings that teachers experienced as a result of their work. Teachers were asked to indicate the extent to which they experienced various emotional states during the past month, on a 5-point scale ranging from 'rarely or never' to 'very often'. These emotional states are similar to the most frequent stress symptoms reported by teachers (Kyriacou & Pratt, 1985). Four of the five items in this scale refer to aversive feelings associated with work in general (e.g., "Worry or feeling of discomfort before work"), whereas one item refers to feelings associated with a specific class (e.g., "Worry or feeling of discomfort before a particular class"). Only those items associated with work in general were used, as internal consistency and confirmatory factor analyses suggested that the items referring to a specific class did not form part of the overall scale (alpha = .82). Global Job Satisfaction. A single item was used to assess teachers' overall level of job satisfaction. Teachers were asked to indicate how satisfied they were with their job on a 9-point scale ranging from 'extremely dissatisfied' to 'extremely satisfied'. Facet Job Satisfaction. Teachers were asked to rate their level of satisfaction with 10 specific aspects of their job on a 7-point scale ranging from 'extremely dissatisfied' to 'extremely satisfied'. For example, teachers were asked to rate their level of satisfaction with "The amount of work that is expected of you", "The discipline policy in your school", and "The people management skills of your school administration". Principal components analysis suggested that these items measure a single factor, accounting for 47% of the variance (alpha = .87). The unit weighted composite score for this factor correlated .57 (N = 514, p < .001) with the measure of global job satisfaction. The 10 items were used individually for the purpose of this study. Mood State. The two mood states of stress and arousal were assessed by the 30-item Stress Arousal Checklist (Gotts & Cox, 1988). The 18-item Stress subscale (SACL-S) reflects the bipolar dimension of pleasantness-unpleasantness, while the 12-item Arousal subscale (SACL-A) provides a measure of the physiological and behavioural state of attentiveness. A 4-point Likert scale was used in place of the dichotomised asymmetric response scale to enhance differentiation among teachers. Teachers were asked to rate the extent to which they were experiencing each of 30 mood adjectives (e.g., alert, calm, relaxed, tense) 'at this moment in time' on a scale ranging from 'not at all' to 'very much' (alpha = .92 & .86 respectively). Consequently, these scales provide a measure of teachers' mood at the time of completing the survey. Moreover, it has been shown that SACL-S is an indicator of ill-being (also known as psychological distress), whilst SACL-A is an indicator of well-being (Hart, Wearing & Headey, in press, 1993). Teacher Morale and School Organisational Climate. During Studies 1 and 2, 54 items from the item pool of the School Organisational Health Questionnaire were identified as measuring teacher morale and 11 separate dimensions of school organisational climate. These items were employed in this study, together with 3 additional items whose content appeared to assess the quality of communication between teachers and school leadership. An item from the Supportive Leadership dimension, which had been worded differently in Studies 1 and 2, was worded the same as for Study 1 on the basis of previous results. Two other items that were found to be poor indicators of their latent constructs in Study 2 were reworded. The item "The rules and sanctions relating to discipline are enforced in a consistent fashion in this school" became "The rules and sanctions relating to discipline are not enforced in a consistent fashion in this school", while the item "There is time for teachers to relax in this school" became "There is no time for teachers to relax in this school". It was possible that teachers were used to thinking about the content of these items in the negative, as might be the case for one of the Supportive Leadership items that was worded differently in Studies 1 and 2. Result and Discussion Due to limitations of the SPSS/PC+ factor procedure, it was not possible to simultaneously analyse all 57 items of the School Organisational Health Questionnaire. Moreover, the asymptotic covariances of the polychoric correlation matrix for these items could not be computed (over nine megabytes of computer memory is required to compute asymptotic covariances for 57 items). Consequently, it was necessary to divide the items into two groups. Group 1 (45 items) consisted of those items for the Communication, Curriculum Consultation, Effective Discipline Policy, Feedback, Goal Congruence, Participative Decision-Making, Professional Development, Professional Interaction, Role Clarity, and Supportive Leadership dimensions, whereas Group 2 consisted of those items for the Excessive Work Demands, Morale and Student Orientation dimensions (12 items). The items were divided in this way so that as many items as possible could be simultaneously examined. Morale and Student Orientation were placed in Group 2 as these were thought to be outcomes of other organisational climate dimensions (cf. Hart, Conn, Carter & Wearing, 1993). Excessive Work Demands was also included in this group, because Study 2 showed that it was not related to any other dimension. Principal components analyses, using an oblimin rotation, were conducted on both groups of items. The results for Group 1 showed that two of the Curriculum Consultation items and one of the Student Orientation items failed to load on their respective dimensions. These items were therefore deleted in subsequent analyses. The principal components analysis for the remaining 42 items showed that 9 factors, accounting for 61.7% of the variance, had an eigenvalue greater than unity. Although the factor loadings in this solution were generally consistent with the questionnaire's theoretical structure, nine items loaded significantly on more than one dimension. In all but two cases, the primary loadings were on the expected dimension, and the secondary loadings ranged from .31 to .47 (M = .38, SD = .06) in absolute value. Interestingly, the Communication items failed to form a separate dimension. One item ("There is good communication between teachers and the administration in this school") loaded (.59) on the Supportive Leadership dimension, and two others ("There is good communication between staff members in this school" and "There is good communication between groups in this school") loaded (.66 and .50 respectively) on the Professional Interaction dimension. A solution constrained to extract 10 factors also failed to separate the Communication items. Moreover, confirmatory factor analyses showed that the 9 factor solution was a significantly better fit, when compared with a 10 factor solution in which Communication was estimated as a separate dimension (p < .001). The principal components solution for the Group 2 items showed that 3 factors had an eigenvalue greater than unity, accounting for 63.0% of the variance. These three factors reflected the Excessive Work Demands, Morale and Student Orientation dimensions, and no items significantly crossloaded. The two groups of items were also subjected to confirmatory factor analyses, using the estimation procedures outlined in the previous two studies. The goodness-of-fit indices for these analyses are shown in Table 5. ______________________ Insert Table 5 about here ______________________ As can be seen by the goodness-of-fit indices, there was an acceptable fit between the theoretical factor structure and observed data. Although the likelihood ratio test statistic, relative to the degrees of freedom, was high for the Group 2 analysis, this was not unexpected given this statistics sensitivity to sample size. Once again the GFI and RNI varied across the different estimation techniques, suggesting that these indices are influenced by the estimation procedure used in a particular analyses. The estimated factor loadings derived from the confirmatory analyses are shown in Table 6, together with the factor loadings obtained from the principal components analyses. ______________________ Insert Table 6 about here ______________________ Although the pattern of factor loadings was similar for the different confirmatory solutions, these varied from those obtained during the principal components analyses. The differences were most evident for those items which had significantly crossloaded during the exploratory analysis of Group 1 items. Examination of the estimated loadings derived from the confirmatory factor analyses showed that the vast majority (89%) were greater than .55, reflecting at least 30% shared variance between the item and its respective dimension. Moreover, a reworded item from the Excessive Work Demands dimension ("There is no time for teachers to relax in this school") loaded adequately in this study (estimated loading ranged from .60 to .74). Two other items ("The rules and sanctions relating to discipline are not enforced in a consistent manner" and "The school's administrators don't really know the problems faced by teachers") that were reworded as a result of the findings in Study 2, were found to have better loadings in this study, but were still considered marginal (estimated loadings ranged from .48 to .62 and from .52 to .60 respectively). The summary statistics and internal consistency coefficients for the twelve dimensions of the School Organisational Health Questionnaire are shown in Table 7. The internal consistency coefficients were based on the combined data for Primary and Secondary teachers, whilst the summary statistics were calculated separately. Two internal consistency statistics are reported; Cronbach's alpha and the total coefficient of determination obtained from the single factor congeneric models estimated by LISREL VII. Although it is inappropriate to use single factor congeneric models to support the structure of a questionnaire, when individual constructs are known to be correlated with one another (Joreskog & Sorbom, 1989b), the total coefficients of determination derived from these models indicate the degree of internal consistency associated with each dimension. The single factor congeneric models were estimated with the weighted least squares method of estimation based on the asymptotic covariances of the polychoric correlation matrix. ______________________ Insert Table 7 about here ______________________ The alpha coefficients and the total coefficients of determination supported the internal consistency of the twelve dimensions. Moreover, the goodness-of-fit statistics for the single factor congeneric models were all satisfactory, with the GFI's being at least .994. Examination of the summary statistics suggested that primary teachers reported higher levels of Morale and more favourable perceptions of their schools' organisation. Analyses of variance confirmed this view, and showed that primary teachers reported significantly more favourable levels of Morale, Feedback, Supportive Leadership, Goal Congruence, Professional Development, Participative Decision-Making, Excessive Work Demands, Effective Discipline Policy and Student Orientation (p < .05) when compared to secondary teachers. There were no significant differences between primary and secondary teachers' reports of Professional Interaction, Role Clarity and Curriculum Consultation (p > .10). Analyses of variance also showed that there were no significant gender differences, except for Professional Development. Male teachers tended to report more favourable levels of Professional Development than did female teachers, although this difference was only marginal (p = .03). Cochrans C and Bartlett-Box F tests also showed that the variances for each dimension generally were similar (p > .10) for primary and secondary teachers, and for males and females. There were just three exceptions to this trend; primary teachers varied more on their reports of Professional Interaction and less on their reports of Student Orientation, while females varied more on their perceptions of Professional Interaction. Although it has been shown that the School Organisational Health Questionnaire has good psychometric properties in terms of its factor structure and internal consistency, it was still necessary to examine whether the questionnaire assessed the constructs it purported to measure. A first step was to correlate the twelve dimensions of the School Organisational Health Questionnaire with measures of teachers' current mood state, global job satisfaction, work strain and quality of work life. The Pearson product-moment correlations for these analyses are shown in Table 8. ______________________ Insert Table 8 about here ______________________ All correlations were in the expected direction. The School Organisational Health Questionnaire's dimensions tended to correlate more strongly with quality of work life and job satisfaction indices, than with general work strain. The correlations for Excessive Work Demands were the only exception, suggesting that this dimension correlated more strongly with teachers' level of psychological strain. This pattern of results was not surprising. It has often been found, in work (e.g., Hart, 1992; Hart, Wearing & Headey, 1993; in press) and other community settings (e.g., Headey & Wearing, 1992), that positive experiences tend to contribute more to positive psychological outcomes (e.g., psychological well-being, positive affect, satisfaction), whereas negative experiences tend to contribute more to negative psychological outcomes (e.g., psychological distress, negative affect, psychosomatic symptoms). Whilst the theoretical content of most items in the School Organisational Health Questionnaire is related to positive aspects of a school's environment, the items for Excessive Work Demands tap into a negative aspect. It was, therefore, a little surprising that the SACL-S subscale correlated more strongly than the SACL-A subscale with all 12 dimensions. Consistent with the negative affectivity bias, which is known to be a confounding factor in self-report questionnaires of this nature (e.g., Parkes, 1990; Schaubroeck, Ganster & Fox, 1992; Watson, 1990; Watson & Penneback, 1989; cf. Dohrenwend & Shrout, 1985), these correlations suggest that teachers' negative mood at the time of completing the questionnaire influences their responses. Measures of neuroticism or trait negative affect (e.g., Costa & McCrae, 1985) should be included in the further assessment of this questionnaire, in order to test this possibility. A second step in establishing the extent to which the School Organisational Health Questionnaire assessed the constructs in purported to measure, was to correlate the separate dimensions of this questionnaire with facets of job satisfaction that were similar in nature. The Pearson product-moment correlations between the dimensions of the School Organisational Health Questionnaire and measures of facet job satisfaction are shown in Table 9. ______________________ Insert Table 9 about here ______________________ The pattern of correlations tended to support the notion that each dimension of the School Organisational Health Questionnaire assessed what it purported to measure. It can be seen that each dimension correlated more strongly with the facet of job satisfaction that was most similar in nature. For example, satisfaction with People Management Skills was more strongly related to Supportive Leadership than any other dimension of the School Organisational Health Questionnaire. These correlations, along with those reported in Table 8, tend to support the construct validity of the School Organisational Health Questionnaire. General Discussion Three large studies have demonstrated that the School Organisational Health Questionnaire has good psychometric properties. Confirmatory factor analyses showed that the 54 items retained in the final version of this questionnaire measured teacher morale and 11 separate dimensions of school organisational climate. These dimensions were both theoretically and empirically distinct. Moreover, internal consistency coefficients suggested that each dimension measured a unitary construct. The questionnaire's construct validity was also supported by the pattern of correlations found between the various dimensions of the School Organisational Health Questionnaire and other theoretically related constructs. The confirmatory factor analyses reported in this paper also helped to clarify the merits of the GFI as a measure of fit in structural equation modeling. As suggested by Anderson and Gerbing (1984), it was found that the GFI was biased downwards as the number of indicators and latent constructs in a given solution increased. This brings into question Cuttance's (1987) blanket assertion that most adequate models will have a GFI greater than .90. As suggested by the data reported here, a GFI between .80 and .85 may be quite adequate when the model under investigation contains a large number of indicators (say > 30). It is also interesting to note that different values of the GFI and the RNI were obtained from different estimation techniques. Both statistics appeared to biased downwards when maximum likelihood estimation was used with a Pearson product-moment correlation matrix or a variance-covariance matrix. This finding is particularly important, given Gerbing and Anderson's (1992) recommendation that the RNI provides the best measure of relative fit. Although these studies demonstrated that the School Organisational Health Questionnaire has good psychometric properties, it was also necessary for the questionnaire to be of practical value to schools (Rentoul & Fraser, 1983). During the questionnaire's development there was a continual interplay between theory, research and practice. The questionnaire was used to facilitate school improvement, as part of the organisational development programs which gave rise to the data reported in these studies (Carter et al., 1993). Schools' Actual and Preferred mean scores were graphically represented for each dimension of the questionnaire (see Appendix 1). These scores were transformed to a common metric (0 - 100), so that discrepancies between the various dimensions could be compared. This information was used, together with a frequency distribution for each item in the Actual form and the qualitative responses obtained from teachers after the questionnaire's administration, to identify the areas that a schools' staff believed were in most need of improvement. Teachers found this information very useful, and generally spent a great deal of time discussing its implications. The quantitative and qualitative data were also used to generate alternative strategies for dealing with identified issues. Carter et al. (1993) have also found the School Organisational Health Questionnaire to be effective in assessing the outcomes of school organisational health interventions. The results of these studies offer promising measures of teacher morale and school organisational climate, and will enable research into the determinants and consequences of teacher morale to proceed in a more systematic manner (Evans, 1992; cf. Hill et al., 1993). It is still necessary, however, to obtain test- retest information on the School Organisational Health Questionnaire, and replicate the current findings with teachers from different educational settings. Meanwhile, this questionnaire provides the only validated measure of teacher morale and school organisational climate that assesses a range of important organisational behaviour and human resource management issues. NOTES 1. The response rates reported in this paper were based on the staff establishment size for each school during 1990. This may have resulted in slightly higher response rates being reported than were actually obtained, as the staff establishment refers to full-time equivalent positions rather than the number of teachers in a particular school. REFERENCES Anderson, J.C. & Gerbing, D.W. (1984). The effect of sampling error on convergence, improper solutions, and goodness-of-fit indices for maximum likelihood confirmatory factor analysis. Psychometrika, 49, 155-173. Borg, M.G. (1990). Occupational stress in British educational settings: a review. Educational Psychology, 10, 103-126. Carter, N.L., Hart, P.M. & Wearing, A.J. (1993). Teacher Well- Being and Organisational Health: Development and Evaluation of a School Based Program to Enhance Organisational Climate. Manuscript in preparation, Department of Psychology, University of Melbourne. Cheng, Y.C. (1993). Profiles of organizational culture and effective schools. School Effectiveness and School Improvement, 4, 1-26. Conley, S.C., Bacharach, S.B. & Bauer, S. (1989). The school work environment and teacher career dissatisfaction. Educational Administration Quarterly, 25, 58-81. Costa, P.T., Jr., & McCrae, R.R. (1985). The NEO Personality Inventory, Manual. Odessa, Florida: Psychological Assessment Resources. Cox, T. (1988). Organizational health. Work and Stress, 2, 1- 2. Cox, T. (1992). Occupational health: past, present and future. Work and Stress, 6, 99-102. Cox, T., Boot, N., Cox, S. & Harrison, S. (1988). Stress in schools: an organizational perspective. Work and Stress, 2, 353- 362. Cuttance, P. (1987). Issues and problems in the application of structural equation models. In Structural Modeling by Example, Applications in Educational, Sociological and Behavioral Research, ed. P. Cuttance & R. Ecob, 241-279. New York: Cambridge. Cuttance, P. & Ecob, R. (Eds). (1987). Structural Modeling by Example, Applications in Educational, Sociological and Behavioral Research. New York: Cambridge. DeFrank, R.S. (1988). Psychometric measurement of occupational stress: current concerns and future directions. In Occupational Stress: Issues and Developments in Research, ed. J.J. Hurrell, Jr., L.R. Murphy, S.L. Sauter & C.L. Cooper, 54-65. London: Taylor & Francis. Diener, E., Emmons, R.L.A., Larsen, R.J. & Griffin, S. (1985). The Satisfaction With Life Scale. Journal of Personality Assessment, 49, 71-75. Doherty, J. (1988). Psychological morale: its conceptualisation and measurement. The Doherty Inventory of Psychological Morale (DIPM). Educational Studies, 14, 65-75. Dohrenwend, B.P. & Shrout, P.E. (1985). "Hassles" in the conceptualization and measurement of life stress variables. American Psychologist, 40, 780-785. Efraty, D. & Sirgy, M.J. (1990). The effects of quality of working life (QWL) on employee behavioral responss. Social Indicators Research, 22, 31-47. Evans, L. (1992). Teacher morale: an individual perspective. Educational Studies, 18, 161-171. Fisher, D. & Fraser, B. (1990). School climate, assessing and improving school environments. SET, Research Information for Teachers, No.2, item 4. Gerbing, D.W. & Anderson, J.C. (1992). Monte Carlo Evaluations of Goodness of Fit Indices for Structural Equation Models. Sociological Methods & Research, 21, 132-159. Gotts, G. & Cox, T. (1988). Stress Arousal Checklist, A Manual for its Administration, Scoring and Interpretation. Melbourne: Swinburne. Hart, P.M. (1992). Stress and Morale: Their Independence in Determining Teacher Quality of Work Life. Paper presented to the 1992 joint conference of the Australian and New Zealand Association's for Research in Education, Deakin University, 22-26 November. Hart, P.M. & Conn, M. (1992). Stress, morale and teachers. Education Quarterly, No.8, 38-40. Hart, P.M., Conn, M., Carter, N.L. & Wearing, A.J. (1993). Understanding teacher quality of work life: a dynamic model of organisational climate, psychological distress and morale. Paper presented to the 1993 National Conference of the Australian Association for Research in Education, Freemantle, Western Australian, 22-25 November. Hart, P.M. & Wearing, A.J. (1993). Problems of Stability and Change in Quality of Life Research: Implications for a Longitudinal Model of Personality, Health and Well-Being. Manuscript in preparation, Department of Psychology, University of Melbourne. Hart, P.M., Wearing, A.J. & Headey, B. (1993). Determinants of police well-being and ill-being: a dynamic model of personality, coping and work experiences. Paper presented to the 1993 National Conference of the Australasian Society for Traumatic Stress Studies, Adelaide, South Australia, 23-25 April. Hart, P.M., Wearing, A.J. & Headey, B. (in press). Assessing police work experiences: development of the Police Daily Hassles and Uplifts Scales. Journal of Criminal Justice, 21. Haplin, A.W. & Croft, D.B. (1963). The Organizational Climate of Schools. Chicago: Midwest Administration Center, University of Chicago. Headey, B. & Wearing, A.J. (1992). Understanding Happiness: A Theory of Subjective Well-Being. Melbourne: Longman Cheshire. Hill, P.W., Holmes-Smith, P.H. & Rowe, K.J. (1993). A Portrait of Schools in Victoria, Summary of Results from Phase 1 of the Victorian Quality Schools Project. Unpublished manuscript, Centre for Applied Educational Research, University of Melbourne. Joreskog, K.G. & Sorbom, D. (1989a). LISREL 7, User's Reference Guide. Mooresville, IN: Scientific Software. Joreskog, K.G. & Sorbom, D. (1989b). LISREL 7, A Guide to the Program and Applications (2nd Ed.). Chicago: SPSS Inc. Kottkamp, R.B., Mulhern, J.A. & Hoy, W.K. (1987). Secondary school climate: a revision of the OCDQ. Educational Administration Quarterly, 23, 31-48. Kyriacou, C. (1987). Teacher stress and burnout: an international review. Educational Research, 29, 146-152. Kyriacou, C. & Pratt, J. (1985). Teacher stress and psychoneurotic symptoms. British Journal of Educational Psychology, 55, 61-64. Loehlin, J.C. (1992). Latent Variable Models, An Introduction to Factor, Path, and Structural Analysis (2nd Ed.). Hillsdale, NJ: Lawrence Erlbaum. McDonald, R.P. & Marsh, H.W. (1990). Choosing a multivariate model: noncentrality and goodness of fit. Psychological Bulletin, 107, 247-255. Meaney, D. & Smith, K.R. (1988). The assessment and development of staff morale. The Practising Administrator, 10, 14-17. Milton, C.R., Entrekin, L. & Stening, B.W. (1984). Organizational Behaviour in Australia. Sydney: Prentice-Hall. Owens, R.G. (1991). Organizational Behavior in Education (4th Ed.). Boston: Allyn and Bacon. Parkes, K.R. (1990). Coping, Negative Affectivity, and the Work Environment: Additive and Interactive Predictors of Mental Health. Journal of Applied Psychology, 75, 399-409. Rentoul, A.J. & Fraser, B.J. (1983). Development of a school- level environment questionnaire. The Journal of Educational Administration, XXI, 21-39. Rosenholtz, S.J. (1989). Workplace conditions that affect teacher quality and commitment: Implications for teacher induction programs. The Elementary School Journal, 89, 421-439. Rowe, K.J. (1990). The Impact of Teacher Professional Development on Students' Reading Achievement: A Multilevel and Structural Equation Modelling Approach. Paper presented to the Annual Conference of the Australian Association for Research in Education, University of Sydney, November 27 to December 2. Rowe, K.J. (1992). Identifying Type I Errors in Educational and Social Research: Comparisons of Results from Fitting OLS and Multilevel Regression Models to Hierarchically Structured Data. Paper presented to the Third National Social Research Conference, University of Western Australia, June 29 to July 2. Schaubroeck, J., Ganster, D.C. & Fox, M.L. (1992). Dispositional affect and work-related stress. Journal of Applied Psychology, 77, 322-335. Schuler, R.S., Dowling, P.J., Smart, J.P. & Huber, V.L. (1992). Human Resource Management in Australia (2nd Ed.). Artarmon, NSW: Harper Educational. Smith, K.R. (1966). A proposed model for the investigation of teacher morale. Journal of Educational Administration, 4, 143- 148. Smith, K.R. (1976). Morale: A refinement of Stogdill's model. Journal of Educational Administration, 14, 87-93. Smith, K.R. (1987). One hundred dissertations: a review of the University of New England morale research project. Journal of Educational Administration, 25, 101-125. Tellenback, S., Brenner, S.O. & Lofgren, H. (1983). Teacher Stress: exploratory model building. Journal of Occupational Psychology, 56, 19-33. Watson, D. (1988). Intraindividual and interindividual analyses of positive and negative affect: their relation to health complaints, perceived stress, and daily activities. Journal of Personality and Social Psychology, 54, 1020-1030. Watson, D. (1990). On the dispositional nature of stress measures: stable and nonspecific influences on self-reported hassles. Psychological Inquiry, An International Journal of Peer Commentary and Review, 1, 34-37. Watson, D. & Pennebaker, J.W. (1989). Health complaints, stress and distress: exploring the central role of negative affectivity. Psychological Review, 96, 234-254. Wearing, A.J., Bell, R., McMurray, N.E., Conn, M. & Dudgeon, P. (1990). Teacher Stress in Victoria, A Survey of Teacher's Views, The Report. Melbourne: Education Shop. Table 1 Goodness-of-Fit Statistics for the Confirmatory Factor Analyses of 38 Items From the School Organisational Health Questionnaire. ____________________________________________________________________________________ Solution x2 df p GFI AGFI RMSR RNI ____________________________________________________________________________________ Group 1 Analysis (20 items, 4 factors) ML/PPM 510 163 < .001 .91 .89 .06 .94 ML/COV 510 163 < .001 .91 .89 .07 .94 GLS/COV 400 163 < .001 .93 .91 .09 .98 WLS/ASC/PCC 459 163 < .001 .97 .96 .06 .96 Group 2 Analysis (18 items, 4 factors) ML/PPM 382 127 < .001 .93 .90 .05 .91 ML/COV 382 127 < .001 .93 .90 .06 .91 GLS/COV 338 127 < .001 .93 .91 .07 .92 WLS/ASC/PCC 265 127 < .001 .97 .96 .05 .96 Combined Analysis (38 items, 8 factors) ML/PPM 1691 634 < .001 .85 .83 .06 .89 ML/COV 1691 634 < .001 .85 .83 .07 .89 GLS/COV 1254 634 < .001 .88 .86 .10 .97 WLS/ASC/PCC 1345 634 < .001 .93 .92 .06 .92 ____________________________________________________________________________________ Note. N = 543 listwise. The Group 1 analysis was based on items measuring Morale, Feedback, Professional Interaction and Supportive Leadership, and the Group 2 analysis was based on items measuring Goal Congruence, Professional Development, Participative Decision-Making and Role Clarity. Key to solutions: ML/PPM = maximum likelihood estimation based on a Pearson product moment correlation matrix; ML/COV = maximum likelihood estimation based on a variance-covariance matrix; GLS/COV = generalised least squares estimation based on a variance-covariance matrix; WLS/ASC/PCC = weighted least squares estimation based on the asymptotic covariances of the polychoric correlation matrix. Table 2 Principal Components Factor Loadings and Standardised Item Reliabilities Obtained from Confirmatory Factor Analyses Conducted on 38 Items from the School Organisational Health Questionnaire. __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ Morale 61. The morale in this school is high. 80 83 96 82 85 38. There is good team spirit in this school. 69 79 90 78 83 80. Teachers take pride in this school. 84 78 82 72 79 71. Teachers go about their work with enthusiasm. 83 76 70 68 78 50. There is a lot of energy in this school. 80 75 82 70 75 Feedback 58. There is a structure and ongoing process that provides feedback on my work performance. 88 86 81 75 89 69. I am happy with the quality of feedback I receive on my work performance. 84 84 87 78 84 47. I am regularly given feedback on how I am performing my role. 84 81 81 75 82 34. I have the opportunity to discuss and receive feedback on my work performance. 79 75 86 81 76 78. Teachers receive recognition for good work.a 45 64 70 74 74 22. I am encouraged in my work by praise, thanks or other recognition. 44 61 76 74 70 Professional Interaction 74. I receive support from my colleagues. 74 76 71 64 75 29. Teachers in this school can rely on their colleagues for support and assistance when needed. 66 73 71 64 75 53. Teachers frequently discuss and share teaching methods and strategies with each other. 59 59 66 65 62 6. I feel accepted by other staff in this school. 71 56 53 43 59 (Table 2 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ 10. I have the opportunity to be involved in cooperative work with other members of staff. 61 53 54 52 57 Supportive Leadership 62. The administration in this school can be relied upon when things get tough. 74 86 101 92 87 41. There is support from the administration in this school. 75 85 96 86 85 3. I am able to approach the administration in this school to discuss concerns or grievances. 77 67 82 71 70 15. The school's administrators don't really know the problems faced by teachers. 59 42 50 44 45 Goal Congruence 23. The staff are committed to the school's goals. 69 71 71 61 72 59. The school has a clearly stated set of objectives and goals.b 47 64 70 67 67 79. There is agreement in the teaching philosophy of this school. 80 63 65 57 61 70. My personal goals are in agreement with the goals of this school. 52 53 52 55 56 37. The goals of this school are not easily understood. and values.c 42 42 49 45 43 Professional Development 9. Others in the school take an active interest in my career development and professional growth. 72 68 75 68 72 27. There are opportunities in this school for developing new skills. 56 65 69 66 69 2. I am encouraged to pursue further professional development. 81 67 79 76 65 (Table 2 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ 14. The professional development planning in the school takes into account my individual needs and interests. 66 64 67 68 63 39. It is not difficult to gain access to inservice courses. 62 49 61 60 48 Participative Decision-Making 57. I am happy with the decision-making processes used in this school. 57 77 88 80 79 33. There are forums in this school where I can express my views and opinions. 72 62 67 56 63 77. There is opportunity for staff to participate in school policy and decision-making. 82 66 63 54 65 67. Teachers are frequently asked to participate in decisions concerning administrative policies and procedures in this school. 69 54 54 51 58 Role Clarity 5. I am always clear about what others at school expect of me.d 54 66 69 62 69 63. I am clear about my professional responsibilities. 75 60 59 58 59 18. I always know how much authority I have in this school. 52 56 62 58 57 11. My work objectives are always well defined. 71 51 51 50 51 __________________________________________________________________________________ N = 543 listwise. Note. Decimal points omitted. Key to solutions: (1) principal components analysis; (2) maximum likelihood estimation based on a Pearson product-moment correlation matrix; (3) maximum likelihood estimation based on a variance-covariance matrix; (4) generalised least squares estimation based on a variance-covariance matrix; (5) weighted least squares estimation based on the asymptotic covariances of the polychoric correlation matrix. All item reliabilities derived from the confirmatory analyses were significant at the .001 level. Results for the first solution (principal components analysis) were based on two separate analyses. The Morale, Feedback, Professional Interaction and Supportive Leadership items were used in the first analysis, and the Goal Congruence, Professional Development, Participative Decision-Making and Role Clarity items were used in the second analysis. a This item crossloaded (-.34) on Supportive Leadership in the principal components analysis. b This item crossloaded (-.33) on Role Clarity in the principal components analysis. c This item crossloaded (-.42) on Role Clarity in the principal components analysis. d This item crossloaded (-.32) on Professional Development in the principal components analysis. Table 3 Goodness-of-Fit Statistics for the Confirmatory Factor Analyses Conducted on 34 items from the School Organisational Health Questionnaire. ____________________________________________________________________________________ Solution x2 df p GFI AGFI RMSR RNI ____________________________________________________________________________________ Group 1 Analysis (18 items, 4 factors) ML/PPM 257 129 < .001 .91 .88 .05 .94 ML/COV 257 129 < .001 .91 .88 .06 .94 GLS/COV 230 129 < .001 .91 .88 .09 .98 WLS/ASC/PCC 157 129 > .05 .97 .96 .05 .99 Group 2 Analysis (16 items, 4 factors) ML/PPM 214 98 < .001 .92 .89 .07 .90 ML/COV 214 98 < .001 .92 .89 .09 .91 GLS/COV 191 98 < .001 .92 .88 .13 .94 WLS/ASC/PCC 181 98 < .001 .96 .94 .07 .95 Combined Analysis (34 items, 8 factors) ML/PPM 988 499 < .001 .84 .81 .06 .87 ML/COV 988 499 < .001 .84 .81 .08 .88 GLS/COV 781 499 < .001 .84 .81 .13 .97 WLS/ASC/PCC 710 499 < .001 .92 .90 .07 .94 ____________________________________________________________________________________ N = 287 listwise Note. The Group 1 analysis was based on items measuring Morale, Professional Interaction, Supportive Leadership and Participative Decision-Making, and the Group 2 analysis was based on items measuring Excessive Work Demands, Curriculum Consultation, Effective Discipline Policy and Student Orientation. Key to solutions: ML/PPM = maximum likelihood estimation based on a Pearson product moment correlation matrix; ML/COV = maximum likelihood estimation based on a variance-covariance matrix; GLS/COV = generalised least squares estimation based on a variance-covariance matrix; WLS/ASC/PCC = weighted least squares estimation based on the asymptotic covariances of the polychoric correlation matrix. Table 4 Standardised Item Reliabilities Obtained from Confirmatory Factor Analyses Conducted on 34 Items from the School Organisational Health Questionnaire. __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 __________________________________________________________________________________ Morale 38. There is good team spirit in this school. 79 86 75 79 61. The morale in this school is high. 78 86 79 77 50. There is a lot of energy in this school. 70 80 78 71 71. Teachers go about their work with enthusiasm. 70 70 72 70 80. Teachers take pride in this school. 68 75 71 69 Professional Interaction 74. I receive support from my colleagues. 78 72 65 78 29. Teachers in this school can rely on their colleagues for support and assistance when needed. 70 61 50 72 6. I feel accepted by other staff in this school. 61 60 56 67 53. Teachers frequently discuss and share teaching methods and strategies with each other. 55 60 58 58 10. I have the opportunity to be involved in cooperative work with other members of staff. 52 51 47 53 Supportive Leadership 41. There is support from the administration in this school. 86 91 80 86 62. The administration in this school can be relied upon when things get tough. 82 90 72 82 3. I am able to approach the administration in this school to discuss concerns or grievances. 68 80 66 71 15. The school's administrators know the problems faced by teachers. 16* 20* 09NS 20* (Table 4 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 __________________________________________________________________________________ Participative Decision-Making 57. I am happy with the decision-making processes used in this school. 77 83 71 80 77. There is opportunity for staff to participate in school policy and decision-making. 70 67 64 70 33. There are forums in this school where I can express my views and opinions. 66 76 66 68 67. Teachers are frequently asked to participate in decisions concerning administrative policies and procedures in this school. 63 66 63 63 Excessive Work Demands 26. Teachers are overloaded with work in this school. 85 95 74 91 7. There is too much expected of teachers in this school. 77 94 65 75 65. There is constant pressure for teachers to keep working. 58 62 52 60 45. There is time for teachers to relax in this school. 16* 20* 22* 23* Curriculum Consultation 28. There is sufficient contact between different sections of the school in curriculum planning. 58 63 63 60 44. There is effective coordination of the curriculum in this school. 58 56 43 62 12. Teachers consult with each other about their teaching and curriculum. 69 74 69 67 56. Teachers consult with area/subject coordinators about their teaching. 57 62 55 58 Effective Discipline Policy 4. There is an agreed philosophy on discipline in this school. 73 89 81 73 48. The rules and sanctions relating to discipline in this school are well understood by both staff and students. 74 81 72 72 (Table 4 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 __________________________________________________________________________________ 32. My own expectations about discipline are the same as most other teachers at this school. 57 62 60 63 60. The rules and sanctions relating to discipline are enforced in a consistent fashion in this school. 37 47 34 38 Student Orientation 8. Students are treated as responsible people in this school. 71 69 62 72 64. Students in this school are encouraged to experience success. 66 69 61 67 40. Students have access to advice and counselling when needed. 62 64 63 64 52. This school promotes the concept of students being individuals. 62 62 61 61 __________________________________________________________________________________ N = 543 listwise. Note. Decimal points omitted. All item reliabilities significant at the .001 level, except those marked (*), which were significant at the .05 level. Key to solutions: (1) maximum likelihood estimation based on a Pearson product-moment correlation matrix; (2) maximum likelihood estimation based on a variance-covariance matrix; (3) generalised least squares estimation based on a variance-covariance matrix; (4) weighted least squares estimation based on the asymptotic covariances of the polychoric correlation matrix. Table 5 Goodness-of-Fit Statistics for the Confirmatory Factor Analyses Conducted on 54 items from the School Organisational Health Questionnaire. ____________________________________________________________________________________ Solution x2 df p GFI AGFI RMSR RNI ____________________________________________________________________________________ Group 1 Analysis (42 items, 9 factors) ML/PPM 1754 781 < .001 .84 .82 .05 .90 ML/COV 1754 781 < .001 .84 .82 .06 .90 GLS/COV 1275 781 < .001 .87 .85 .08 .98 WLS/ASC/PCC 1292 781 < .001 .93 .92 .06 .93 Group 2 Analysis (12 items, 3 factors) ML/PPM 197 51 < .001 .93 .90 .06 .93 ML/COV 197 51 < .001 .93 .90 .07 .93 GLS/COV 166 51 < .001 .94 .91 .11 .97 WLS/ASC/PCC 186 51 < .001 .97 .96 .07 .96 ____________________________________________________________________________________ Note. N = 468 for the ML/PPM, ML/COV and GLS/COV solutions; N = 480 and 536 for the EPC and WLS/ASC/PPM solutions for Group 1 and 2 respectively. The Group 1 analysis was based on items measuring Feedback, Professional Interaction, Supportive Leadership, Goal Congruence, Professional Development, Participative Decision-Making, Role Clarity, Curriculum Consultation, and Effective Discipline Policy, and the Group 2 analysis was based on items measuring Morale, Excessive Work Demands and Student Orientation. Key to solutions: ML/PPM = maximum likelihood estimation based on a Pearson product moment correlation matrix; ML/COV = maximum likelihood estimation based on a variance-covariance matrix; GLS/COV = generalised least squares estimation based on a variance-covariance matrix; WLS/ASC/PCC = weighted least squares estimation based on the asymptotic covariances of the polychoric correlation matrix. Table 6 Principal Components Factor Loadings and Standardised Item Reliabilities Obtained from Confirmatory Factor Analyses Conducted on 54 Items of the School Organisational Health Questionnaire. __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ Group 1 Items Feedback 47. I am regularly given feedback on how I am performing my role. 85 84 91 85 86 69. I am happy with the quality of feedback I receive on my work performance. 79 84 92 88 85 58. There is a structure and ongoing process that provides feedback on my work performance. 81 78 77 74 81 34. I have the opportunity to discuss and receive feedback on my work performance. 72 77 90 81 79 78. Teachers receive recognition for good work. 65 72 78 74 75 22. I am encouraged in my work by praise, thanks or other recognition. 62 72 89 86 75 Professional Interaction 74. I receive support from my colleagues. 72 76 72 68 74 29. Teachers in this school can rely on their colleagues for support and assistance when needed. 71 73 72 66 72 72. There is good communication between staff members in this school. 66 71 67 63 71 53. Teachers frequently discuss and share teaching methods and strategies with each other.* 40 67 74 67 70 17. There is good communication between groups in this school.* 50 63 61 59 64 10. I have the opportunity to be involved in cooperative work with other members of staff. 34 57 62 58 59 6. I feel accepted by other staff in this school. 71 59 52 47 58 (Table 6 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ Supportive Leadership 62. The administration in this school can be relied upon when things get tough. 70 81 95 86 81 41. There is support from the administration in this school. 67 85 88 83 86 3. I am able to approach the administration in this school to discuss concerns or grievances. 56 63 66 65 64 15. The school's administrators don't really know the problems faced by teachers. 65 52 60 53 53 54. There is good communication between teachers and the administration in this school. 59 83 91 86 84 Goal Congruence 79. There is agreement in the teaching philosophy of this school.* 43 74 73 63 71 23. The staff are committed to the school's goals.* 37 69 68 61 70 70. My personal goals are in agreement with the goals of this school.* 57 58 56 59 64 59. The school has a clearly stated set of objectives and goals. 63 61 63 60 59 37. The goals of this school are not easily understood. and values. 70 47 52 48 47 Professional Development 9. Others in the school take an active interest in my career development and professional growth.* 40 77 90 81 77 2. I am encouraged to pursue further professional development. 66 68 76 72 66 14. The professional development planning in the school takes into account my individual needs and interests. 45 64 68 65 63 27. There are opportunities in this school for developing new skills. 49 57 55 52 62 (Table 6 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ 39. It is not difficult to gain access to inservice courses. 68 38 45 44 42 Participative Decision-Making 57. I am happy with the decision-making processes used in this school.* 28 74 72 70 76 77. There is opportunity for staff to participate in school policy and decision-making. 61 67 59 53 69 33. There are forums in this school where I can express my views and opinions.* 39 61 64 61 64 67. Teachers are frequently asked to participate in decisions concerning administrative policies and procedures in this school. 68 59 61 51 59 Role Clarity 5. I am always clear about what others at school expect of me. 56 69 71 72 70 11. My work objectives are always well defined. 74 63 62 61 61 18. I always know how much authority I have in this school.* 44 58 62 60 61 63. I am clear about my professional responsibilities. 70 56 53 50 59 Curriculum Consultation 28. There is sufficient contact between different sections of the school in curriculum planning. 76 76 78 69 78 44. There is effective coordination of the curriculum in this school. 70 73 67 59 74 Effective Discipline Policy 48. The rules and sanctions relating to discipline in this school are well understood by both staff and students. 75 80 92 80 76 (Table 6 cont.) __________________________________________________________________________________ Solution ____________________________ Subscale and Items 1 2 3 4 5 __________________________________________________________________________________ 4. There is an agreed philosophy on discipline in this school. 71 72 85 79 73 32. My own expectations about discipline are the same as most other teachers at this school. 66 50 56 56 60 60. The rules and sanctions relating to discipline are not enforced in a consistent fashion in this school. 55 48 62 54 49 Group 2 Items Morale 61. The morale in this school is high. 85 78 86 87 82 38. There is good team spirit in this school. 86 79 86 85 82 71. Teachers go about their work with enthusiasm. 82 78 72 68 81 80. Teachers take pride in this school. 77 76 79 75 79 50. There is a lot of energy in this school. 59 62 63 58 66 Excessive Work Demands 26. Teachers are overloaded with work in this school. 80 75 85 78 76 7. There is too much expected of teachers in this school. 75 72 84 78 72 45. There is no time for teachers to relax in this school. 72 60 74 66 62 65. There is constant pressure for teachers to keep working. 76 61 63 51 62 Student Orientation 52. This school promotes the concept of students being individuals. 86 74 69 63 76 8. Students are treated as responsible people in this school. 86 67 62 59 70 64. Students in this school are encouraged to experience success. 58 60 54 48 73 __________________________________________________________________________________ Note. N = 468 for the ML/PPM, ML/COV and GLS/COV solutions; N = 480 and 536 for the EPC and WLS/ASC/PPM solutions for Group 1 and 2 respectively. Decimal points omitted. All item reliabilities significant at the .001 level. (*) denotes those items which significantly crossloaded in the principal components solution. Key to solutions: (1) principal components analysis; (2) maximum likelihood estimation based on a Pearson product-moment correlation matrix; (3) maximum likelihood estimation based on a variance-covariance matrix; (4) generalised least squares estimation based on a variance-covariance matrix; (5) weighted least squares estimation based on the asymptotic covariances of the polychoric correlation matrix. Table 7 Summary Statisticsa and Internal Consistency Coefficientsb for SOHQ Subscales. __________________________________________________________________________________ Primary Secondary Subscale Items M SD M SD Alpha TCD __________________________________________________________________________________ Morale 5 17.6 3.82 15.6 4.17 .86 .90 Feedback 6 16.9 5.64 15.2 5.24 .90 .94 Professional Interaction 7 24.9 5.44 24.5 4.69 .84 .89 Support Leadership 5 18.8 4.10 17.3 4.43 .84 .91 Goal Congruence 5 16.8 3.51 16.0 3.76 .77 .82 Professional Development 5 17.1 3.74 15.8 3.91 .74 .82 Participative Decision Making 4 14.9 2.82 13.9 3.07 .75 .83 Role Clarity 4 13.9 3.07 13.8 2.87 .71 .77 Excessive Work Demands 4 11.8 3.26 13.7 3.48 .76 .81 Curriculum Consultationc 2 6.1 1.69 6.3 1.69 .71 .96 Effective Discipline Policy 4 14.0 3.29 12.0 3.35 .71 .81 Student Orientationc 3 11.8 1.91 11.3 2.32 .71 .96 __________________________________________________________________________________ a N=172 for Primary Schools; N=282 for Secondary Schools b N=468 c The total coefficient of determination for Curriculum Consultation and Student Orientation was based on the combined items for these dimensions, since there were insufficient items to estimate these separately. Table 8 Correlations Between SOHQ (Actual) Subscales, SACL-S, SACL-A, GWS, SWL, and GJS. __________________________________________________________________________________ Subscale SACL-S SACL-A GWS SWL GJS __________________________________________________________________________________ Morale -23 09 -21 46 49 Student Orientation -22 09 -16 35 35 Feedback -23 13 -29 45 47 Work Demands 26 -16 38 -24 -27 Discipline Policy -21 13 -21 34 35 Administrative Support -21 11 -22 39 40 Professional Interaction -31 10 -29 47 47 Role Clarity -32 18 -34 43 46 Curriculum Coordination -16 12 -18 25 31 Professional Development -22 09 -17 44 47 Goal Congruence -25 11 -23 42 48 Decision-Making -21 04 -20 36 40 __________________________________________________________________________________ Note. Pairwise deletion of missing data (Minimum N = 420). Decimal points omitted. Critical r (two tailed, N = 420): .10 at .05 level; .13 at .01 level; .16 at .001 level. Table 9 Correlations Between SOHQ (Actual) Subscales and Facet Job Satisfaction Measures. __________________________________________________________________________________ Subscale 1 2 3 4 5 __________________________________________________________________________________ Morale 76 46 24 46 50 Student Orientation 34 33 16 36 30 Feedback 48 63 32 37 44 Work Demands -28 -23 -45 -15 -23 Discipline Policy 44 35 22 64 35 Administrative Support 55 57 30 41 72 Professional Interaction 54 47 29 35 47 Role Clarity 42 42 33 30 36 Curriculum Coordination 33 31 14 28 26 Professional Development 39 45 18 40 39 Goal Congruence 52 47 24 50 42 Decision-Making 45 46 20 37 50 6 7 8 9 10 _______________________________________________ Morale 43 31 39 52 47 Student Orientation 27 26 33 43 37 Feedback 45 34 45 44 43 Work Demands -12 -08 -10 -09 -24 Discipline Policy 24 19 35 44 31 Administrative Support 34 27 39 45 61 Professional Interaction 70 38 47 46 48 Role Clarity 40 32 33 41 39 Curriculum Coordination 35 61 29 40 35 (Table 9 cont.) __________________________________________________________________________________ Subscale 6 7 8 9 10 __________________________________________________________________________________ Professional Development 42 28 60 40 42 Goal Congruence 34 39 38 70 42 Decision-Making 32 30 46 43 60 __________________________________________________________________________________ Note. Pairwise deletion of missing data (minimum N = 536). Decimal points omitted. Critical r (two tailed, N = 536): .08 at .05 level; .11 at .01 level; .14 at .001 level. Key to satisfaction measures: "The extent to which you are satisfied with the following aspects of your work": 1 The level of morale at your school. 2 The structures and processes used in your school to provide you with feedback on your work. 3. The amount of work that is expected of you. 4 The discipline policy in your school. 5 The people management skills of your school administration. 6 The professional support and interaction you receive from other staff at your school. 7 The coordination of curriculum in your school. 8 The opportunities within your school for career and professional development. 9 The clarity and relevance of your school's goals. 10 The decision-making processes operating in your school. Table 7 Likelihood Ratio Test Statistics for Models Comparing the Similarity in Measurement Properties for 38 items of the School Organisational Health Questionnaire. ____________________________________________________________________________________ Change in x2 __________________ Model x2 df p x2 df p Decision ____________________________________________________________________________________ Unconstrained 3144 1268 <.001 H 1 3190 1306 < .001 46 38 > .10 Accept H 2 3246 1347 < .001 102 79 > .05 Accept H 3 3284 1375 < .001 140 107 > . 05 Accept H 4 3188 1296 < .001 44 28 > .10 Accept ____________________________________________________________________________________ Note. H 1: that the factor loadings were invariant in Studies 1 & 2; H 2: that the factor loadings and unique variances were invariant in Studies 1 & 2; H 3: that the factor loadings, unique variances and factor intercorrelations were invariant in Studies 1 & 2; and, H 4: that whilst the factor loadings and unique variance may differ, the factor intercorrelations were invariant in Studies 1 & 2.