BRITISH EDUCATIONAL RESEARCH ASSOCIATION (BERA)
2004 CONFERENCE
UMIST, MANCHESTER
16-18 September 2004
INTERNATIONAL PATTERNS OF TEACHER SATISFACTION AND MOTIVATION IN AUSTRALIA,
ENGLAND, NEW ZEALAND, UNITED STATES OF AMERICA, CANADA, CYPRUS AND MALTA:
THE ROLE OF CONTEXT AND THE ‘THIRD DOMAIN’
Steve Dinham & Catherine Scott
University of New England, Australia
INTERNATIONAL PATTERNS OF TEACHER SATISFACTION AND MOTIVATION IN AUSTRALIA, ENGLAND, NEW ZEALAND, UNITED
STATES OF AMERICA, CANADA, CYPRUS AND MALTA: THE ROLE OF CONTEXT AND THE ‘THIRD DOMAIN’.
Steve Dinham & Catherine Scott
Abstract
This paper reports findings of the International Teacher 2000 Project. The International Teacher 2000 Project was launched in 1995 to investigate teachers’ occupational satisfaction, motivation and mental health in the context of educational change and ‘reform’.
Using data derived from common instruments administered to samples of teachers and school executive in Australia, New Zealand, England, Canada, the USA, Malta and Cyprus – over 4,500 educators in total - the paper presents evidence for a ‘Third Domain’ of teacher career satisfaction grounded in societal context, pressures, trends and forces. Earlier phases of the project conducted in Australia, New Zealand and England revealed that ‘two factor’ models of occupational satisfaction based on the work of Herzberg and Sergiovanni are not adequate to describe/predict teachers’ occupational satisfaction in the current social and political climate. This ‘Third Domain’ is increasingly influencing the two traditionally recognised factors/domains of career satisfaction: the nature of the job itself, and the conditions of work that surround and support it.
As noted, the ‘Third Domain’ is derived from the educational and socio-political context and comprises such issues and forces as the status of teachers, imposed responsibilities placed on schools, recognition of teachers by systems and employers, and the pace and nature of educational change. Differences in general patterns of occupational satisfaction, motivation and mental health across the sample countries can be explained by differences in educational context and thus the relative difference of the influence of the ‘Third Domain’.
In the paper an argument is presented for teachers, school executive and system and government officials to engage with ‘Third Domain’ elements and forces if teacher professionalism and standing are to be enhanced, or even maintained at present levels, and if quality applicants are to be attracted to teaching and retained.
1 The Context Of Teaching
Although ‘people are always wanting teachers to change’ (Hargreaves, 1994: 5), the pressure for and pace of educational change have increased considerably over the past two decades. Education systems have experienced change in the areas of teaching practice and curricula; greater involvement of stakeholders in education; attempts to streamline and ‘restructure’ educational bureaucracies and organisations with a greater emphasis upon accountability, ‘rationality’ and self-management; and the increased politicisation and ‘reform’ of educational systems.
With unresolved social problems of unemployment, family break-up, crime, poverty, and poor health for many, schools have been looked to for solutions, with the result that they have, in many respects, become the ‘wastebaskets of society’ (Halsey, cited
1
by Hargreaves, 1994: 5), being expected to solve the problems that society appears unwilling or unable to deal with (Dinham, 1994).
In Australia, for example, an ageing teaching workforce and an ageing population generally reflected in a lack of growth in student numbers, means that overall, until recently, it has been difficult for ‘new blood’ or younger, more recently trained teachers to enter teaching. Yet the demand and supply equation is very varied, and there is likely to be considerable demand for teachers with rising retirements of older teachers in the near future. Teacher shortages, both for permanent and casual staff, are already becoming critical in some areas. In addition, teacher mobility in educational systems has declined markedly, fewer males are entering teaching, and there has been concern over the ‘quality’ of those entering teacher training (Dinham, 1996).
Given this context of increased expectations, an ageing, increasingly immobile teaching force and a turbulent educational environment, a key question relates to the issue of the satisfaction, motivation and health of teachers and school executive charged with the task of educating today’s and tomorrow’s youth and meeting society’s increasingly complex demands for schools.
2 How the Project Arose
The Teacher 2000 Project arose because of the researchers’ desire to test and extend the findings of earlier qualitative research into teacher resignation and persistence (Dinham, 1992), the impact of teaching on teachers’ families (Dinham, 1997), and the ageing of the Australian teaching population (Dinham, 1996).
It was hoped that a survey of a relatively large number of teachers and school executive would lead to better informed decision making and policy formation in the areas of teacher satisfaction, motivation and health and educational change generally.
Because levels and sources of teacher morale, teacher stress, and the like are often contentious, it was also hoped that examination and measurement of these and other aspects of teachers’ work and lives would provide benchmark data for longitudinal and comparative studies to enable meaningful comparisons and conclusions and better informed decision making.
Once the project had been implemented in Australia, replications were launched in England and then New Zealand, later the USA, and more recently Malta, Cyprus and Canada.
Further background on the project, including relevant literature may be found in previously published work (see Dinham & Scott, 1996, 1998, 2000).
2
3 Project Aims
The International Teacher 2000 Project sought to achieve the following:
1. To build upon and validate understandings of teacher satisfaction, teacher dissatisfaction, orientation to teaching, teachers’ values and teacher health, revealed by prior research.
2. To develop an instrument suitable for identifying and quantifying the sources and relative strength of factors contributing to teacher satisfaction and dissatisfaction.
3. To refine the instrument and to apply it more widely in the future, both inside and outside Australia.
4. To obtain benchmark information on matters relating to teacher welfare which can be used for purposes of tracking, explanation, planning and prediction at school, system and other levels.
5. Where established instruments are used, to compare the findings of the study with previous research.
4 Method
4.1 Instrument
The instrument was a machine readable self-report questionnaire. Minor changes were made to the original, Australian version, with wording of items made consistent with terminology employed in the other national contexts. As noted, detailed description of the methodology may be found elsewhere, but briefly, the instrument contained the following:
demographic items
seven orientation to teaching and two preparedness to teach items (true/false)
75 satisfaction/dissatisfaction with teaching items (seven point Likert scales)
two Likert scale items measuring overall satisfaction and change in satisfaction since beginning teaching
the Commitments scale1 (40 item version)
the 12 item form of the General Health Questionnaire used as a measure of mental distress/stress, and
several opportunities for open ended comments.
1 Not used with the Malta, Canada, Cyprus samples
3
4.2 Data Analysis
Data from completed surveys were computer scanned and analysed using SPSS, while open-ended responses were subject to content analysis and grounded theory techniques (Strauss & Corbin, 1990)), using NUD*IST (QSR, 1994).
5 Participants
5.1 Sampling
Separate teams undertook the project in each countryi and sampling methods employed by the three teams were varied to suit each local context.
5.1.1 Australia
The Metropolitan West Region was one of the largest of the 10 regions in the then New South Wales Department of School Education public school system, which in total employs over 50,000 teachers.2
Western Sydney was chosen because of convenience of access and because of its heterogeneity, ranging from small rural primary schools to large urban high schools, and from schools with large proportions of students with languages other than English to schools with negligible numbers of students with this background. Economically, the region covers a wide spectrum, from areas of high unemployment and poverty, to pockets of affluence. However, the region is usually portrayed as being ‘disadvantaged’ in comparison with the rest of Sydney.
An invitation to participate was made to the principals of over one third of government schools in the region, ensuring a representative sample of schools, given the heterogeneity of the region. Data were collected in a pilot study and a second data collection step, both of which comprised the first phase of the project . In the second phase, responses to the instrument from additional schools were obtained and fuller analysis of the data took place (Dinham & Scott, 1996).
Overall, 47 of the region’s 185 primary schools, 19 of 54 secondary schools, and 5 of the region’s 16 Schools for Specific Purposes (SSP) took part with 2,336 surveys distributed to the 71 schools.
5.1.2 England
Schools were selected by taking every fifth school appearing in the lists of eight local education authorities (LEAs). The decision was made to over-sample Grant Maintained schools and so all GMS in each LEA area were also included. In all, 661 schools were approached to participate with 114 Head Teachers [principals] consenting. Of these, two were from Brighton-Hove, 12 from Cornwall, 34 from Kent, 12 from Leeds, 12 from Cheshire, two from Tower Hamlets, none from Richmond and 40 from Nottinghamshire. In all, 2,384 questionnaires were posted to participating schools.2 The DSE is now known as the NSW Department of Education and Training [DET].
4
The Head Teachers of a sample of schools were contacted to inquire why they had declined to participate. The most common reason given was that the school was soon to have its OFSTED inspection.3 Other Head Teachers explained that their staff were already too heavily burdened and they, the Head Teachers, did not wish to ask them to commit any more time to work or work related activities (see Scott, Cox & Dinham, 1999).
5.1.3 New Zealand
A stratified, random cluster sample was drawn with the objective of obtaining at least 300 primary and at least 300 secondary teachers. Sampling frames were lists of all state and integrated primary schools (including Intermediate schools), and all state and integrated secondary schools in the southern half of the North Island of New Zealand. Two separate random samplings were carried out, one from each frame. The number of schools was determined by staffing numbers at each school drawn, and sampling was continued until a target of 500 possible respondents was reached in each sector.
Once schools were identified, approaches were be made to arrange a visit by a team member to a regular staff meeting to introduce the project and seek co-operation of staff. Questionnaires were distributed at the meeting and arrangements made for their return when completed (see Harker, et. al., 1998).
5.1.4 USA
US data were initially collected from public schools in the state of New Jersey. New Jersey was chosen because it has 609 school districts (more than any other state), yet it is one of the smallest states. It also contains a wide range of schools and whilst many school districts appear to be suburban-like, these suffer many of the problems of inner-city school districts. There are also many ‘one building’ school districts with as few as 100 students in the entire district.
The study was designed as a systematic random sample, stratified by county, and weighted according to gender. The sample included a total of 1056 individuals, including 89 administrators/supervisors (55 male and 34 female) and 968 classroom teachers (252 male and 716 female). Each of the 21 counties in the state was represented in the study in proportion to its contribution of administrators/supervisors to the whole. In all, individuals from 117 public schools were included in the sample.
A second US sample was obtained from St Louis, where 200 questionnaires were distributed to teachers in several locations within and outside Missouri, using convenience sampling. Sixty one were returned completed. In addition, 20 teachers were interviewed concerning their career motivation, satisfaction and dissatisfaction.4
3 The Office for Standards in Education (OFSTED), officially the Office of Her Majesty’s Chief Inspector of Schools in England, was set up on 1 September 1992. It is a non-ministerial government department, independent from the Department for Education. One of OFSTED’s major tasks has been to set up the new independent system of school inspection defined by the Education (Schools) Act 1992. This provides inspection on a four-year cycle for all schools in England which are wholly or mainly state-funded. [Source: OFSTED web site]
5
5.1.5 Canada
For cost and convenience reasons, Canadian teachers were not formally sampled: all teaching members of provincial and territorial teachers’ federations/unions were simply invited to participate.
Over ten months, between September 2001 and June 2002, teachers from all provinces - other than Québec - and all territories were invited to complete a slightly modified (stylistic and context) questionnaire very similar to that used elsewhere. With the questionnaire available either on-line, or in paper format, teachers showed little preference for one or the other. One or two reminder letters or notices about participating were forwarded to teachers by their federations/ unions in a subset of provinces/ territories. There were 627 completed surveys.
Assessing the survey’s completion rate for is somewhat challenging, since no ‘sample’ was involved. Nevertheless, in 1999-2000, Canada had 216,423 full-time teachers in all provinces - other than Québec - and territories (Statistics Canada, 2003).
5.1.6 Malta
Sampling details are not available for Malta. There were however 612 participants.
5.1.7 Cyprus
A preliminary version of the instrument was piloted in September 2002, while the final version was administered in October 2002. A total of 461 teachers from all parts of the Republic of Cyprus participated in the study (see Zembylas & Papanastasiou, 2004: 364-365).
5.2 Sample Description
Table 1: Mean Age, Length of Service and Time in Current School, by Country.
YearsAge Service Current School
Australia 40 15 6England 42 16 6New Zealand 42 15 7USA 45 18 13Malta 45 N/A 6Canada 42 16 8Cyprus 37 13 3
4 Our friend and colleague who conducted the St Louis replication, Dr Barbara Stone, died before the St Louis data were fully analysed and reported.
6
5.2.1 Australia
Of the 2,336 surveys distributed there were 892 respondents (38%), 65% of whom were women and 35% men.
The mean age of respondents was 40 years (women = 39, men = 42), with a range of 20 to 66. These means, while lower than the NSW DSE/DET overall, are close to those for the then Metropolitan West Region which has tended to be younger in profile than more ‘favoured’ regions in the state.
Of women respondents, 56% were primary trained and 44% high school trained, while 32% of men were primary trained and 68% high school trained. Mean length of service as a teacher was 15 years (range less than one to 37 years), and mean length of time in current school was 6 years (range less than one to 31 years).
Including the position of Advanced Skills Teacher5, 44% of the women were in promotions positions, and 55% of the men (48% of the total sample).
The great majority (84%) of the 892 respondents were born in Australia. Only 9% of participants had a first language other than English.
5.2.2 England
Of 2,384 surveys distributed to 114 schools, 609 were returned (26%).6 Mean age of participants was 42 years (range 22 to 62 years). Mean length of service was 16 years (range less than one year to 41 years) and mean length of time in current school was 6 years (range less than one year to 29 years). Only 4% of the sample reported having a first language other than English.
In all 70% of participants were women and 30% men, whilst 40% (24% of men, 47% of women) taught in the infants-primary range, 2% in middle schools, 54% in secondary schools (69% of men and 48% of women) and 5% in schools classed as ‘other’, chiefly special schools. (Rounding of percentages may mean that figures do not total 100%.).
Eleven percent of participants in the English study were Head Teachers (13% of men and 9% of women), 7% were deputy heads (8% of men and 7% of women) and 23% were heads of faculties, years or departments (32% of men, 19% of women). A further 25% described themselves as classroom teachers with extra responsibilities and salary (22% of men and 26% of women), 30% as classroom teachers (23% of men and 33% of women), only 1% were supply teachers (no men and 1% of women) whilst 4% (3% of men and 4% of women) described themselves as ‘other’ (specialists of various sorts including librarians).
5 Introduced in the early 1990s to recognise and reward the exemplary classroom teacher, the AST classification is in the process of being removed from most Australian educational systems (see Dinham & Scott, 1997).6 In both England and Australia, schools were over supplied by mail with surveys (5% or more) to enable casual/emergency and part-time staff to participate, thus the response rates for those two countries are deflated. In New Zealand, where response rates were highest, researchers were able to visit the sample schools to explain the project and procedures.
7
5.2.3 New Zealand
A total of 1,002 surveys were distributed to 71 schools with 565 returned (56%). The mean age of New Zealand participants was 42 years (range 21 to 66), mean length of service 15 years (range less than one year to 45 years) and average service in current school was 7 years (range less than one year to 35 years). Six percent reported having a first language other than English.
Of the 565 participants, 71% were women and 29% were men. Eighteen percent of the men and 4% of the women were school principals (total = 8%), 9% of the men and 10% of the women were deputies (total = 10%), whilst 30% of the men and 23% of the women described themselves as senior teachers or heads of departments (total = 25%). Thirty seven percent of the men and 47% of the women were classroom teachers (total = 44%), whilst 5% of the sample described themselves as relieving teachers (men = 25, women = 6%) and 9% as ‘other’ (men = 4%, women = 11%).
Primary school teachers accounted for 53% of the sample (men = 32%, women = 62%), teachers in intermediate schools 8% (men = 7%, women = 8%), secondary teachers 37% (men = 58%, women = 29%) and 2% taught in mixed schools (men 3%, women = 2%).
Table 2: Percentage Teaching in Each Type of School by Country
Type of School_______________________________________________
Primary* Middle Secondary Other
Australia 48 N/A 52 N/AEngland 40 2 54 5New Zealand 53 8 37 2USA 22 16 43 N/ACyprus 73 16 12 N/AMalta 61 39 (mid/high combined) N/A
* Includes nursery schools. Canada not available
5.2.4 USA - New Jersey Phase
1,056 surveys were distributed and 668 were returned, giving a response rate of 63%. A check of selected demographics (gender, age, length of service, and academic qualifications) suggests that the respondents do not differ from the population in a systematic way.
Of the 668 respondents 66% of were women and 34% men. The mean age of respondents was 45 years (women = 44, men = 46). Mean length of tenure in current school was 13 years (women = 12, men = 15) and mean length of time in the teaching service was 18 years (women = 17, men = 22).
Twenty two percent of respondents were primary teachers (women = 46%, men = 30%), 16% taught in middle schools/junior highs (women = 16% men = 14%) and 43% in high schools, (women 37%, men = 53%).
8
Fourteen percent of the women were in promotions positions, and 39% of the men, with 41% of the sample overall holding some promotion position. The great majority (98%) were born in USA and 98% were also of English speaking background.
5.2.5 Canada
Given that the potential sample was more than 216,000 teachers, with 627 completed questionnaires, the completion rate is 0.3%. Though low, completion rates such as these for large populations and this type of sampling are not uncommon. The average age of participants was 42, and average length of service was 16 years. Teachers had been in their current schools for an average of 8 years.
5.2.6 Malta
Of the 612 participants, 61% were primary teachers, and the remaining 39% taught in middle schools or high schools. Seventy one percent were women, and 29% were men. Age of participants ranged from 21 to 79 years, with a median of 44.5. The age distribution was marked bimodal with peaks in the 30s and 50s. Both length of time in current position and in current setting were markedly skewed to the left. Time in current position ranged from1 to 42 years, with a median of 7.5 years. Time in current setting ranged from 1 to 40 years, with a median of 6 years.
5.2.7 Cyprus
Of the 461 participants, 29.4% were male and 70.6% were female. The average age of the sample was 37.0 years. Average years of teaching experience was 13.0 years, with an average of 2.7 years in the same school. In addition, 80.8% of the sample were teachers, 13.1% vice-principals and 6.0% of the sample were principals.
6 Results
6.1 Development of the Scales; The Australian Model7
The 75 satisfaction items in the questionnaire were based on results of a previous qualitative study (Dinham, 1992). This work had suggested 16 broad aspects of teachers’ satisfaction with teaching and its context (see Scott & Dinham, 2003, for more detail on the development of the scales).
Given an hypothesised structure, the appropriate analytical method is Confirmatory Factor Analysis (CFA) and LISREL 8 (Joreskog & Sorbom, 1993) was used to conduct the analysis. The first analysis on the Australian data, based on the 16 factors, failed to run and further examination revealed that the model was poorly specified.
7 The invaluable assistance provided by Dr Robert Brooks during the model building and testing phases is gratefully acknowledged.
9
Exploratory Factor Analysis (EFA) was conducted using SPSS and CFA was used to explore solutions suggested by EFA. An advantage of using CFA in this context is that it allows the error of measurement associated with items to be correlated. A correlated error is extra covariance between items that is not explained by the existing factor structure. A pattern of correlated errors suggests the existence of another factor not currently in the model.
Initial EFA, using all 75 items, maximum likelihood estimation and Oblimin rotation suggested 17 factors on the eigen value criteria (>1) and inspection of the scree plot that the final solution would contain between 8 and 18 factors. The 17 factor model explained 56% of variance in the data and was overall not a good fit to the data 2χ = 3761 df = 1636 p < 0.001. There were 76 (2%) non-redundant residuals in the reproduced correlation matrix.
Given that one of the objectives of the analysis was to produce a coherent set of scales describing teacher satisfaction, items were removed using a number of criteria. First, one item of items that correlated very highly (0.85) (suggesting they were not measuring anything additional) were removed. Second, items that were highly skewed, and lacked variation (‘motherhood’ items that were ‘automatically’ given the same answer by the vast majority) were removed. Finally, items that loaded on a large number of factors were also removed. All decisions were made in the context of the qualitative understanding of items and their relationships with each other.
The final model contained eight factors composed of 42 of the original 75 items. The overall 2χ of 2151.29 (df = 772 p<0.001), is not unusual given the large number of parameters being estimated. The RSMEA = 0.047, (it is recommended that RMSEA be below 0.05 and 90% confidence interval not calculated due to large 2χ ), GFI = 0.88 (recommended > 0.9) NFI = 0.87 (recommended >0.9). Although the model did not reach acceptable scores on all criteria, the RMSEA was acceptable and the solution conceptually acceptable.
6.2 Cross Validation of the Satisfaction Model
The model has been examined in three subsequent studies. Cross validation is considered an essential component of model development, yet rarely conducted.
6.2.1 British Study
CFA of the 8 factor model in the British sample produced a similarly fitting model 2χ =1588.11, df=772, RMSEA = 0.045, GFI= 0.87, AGFI=0.84, NFI = 0.83.
However, formal testing of the equivalence of the two models using multiple sample analysis (Joreskog & Sorbom, 1993) failed. Errors in the output suggested that the model was misspecified.
10
To facilitate the comparison of the British and Australian data one-factor congeneric modelling of the scales was undertaken. Six of the eight factors were found invariant. Two of the factors were considered to consist of two factors each.
To examine the consistency with the original model derived on the Australian sample, it was re-examined using ten factors. The ten factor model using same 42 items produced a slightly better fitting model ( 2χ =2015.64, df =763 p<0.001, RMSEA = 0.045, GFI 0.89, AGFI = 0.88, NFI 0.88) than the original 8 factor model.
6.2.2 New Zealand Study
Examination of the data from New Zealand on the original eight factor model produced a fit similar to the original Australian 8-factor model 2χ = 1576.58 df = 770 p<0.001, RMSEA = 0.045 GFI = 0.87 AGFI 0.85 NFI = 0.84. Examination of the 10 factor model, ( 2χ = 1718.15 df = 763 p<0.001, RMSEA = 0.049 GFI = 0.87 AGFI 0.84 NFI = 0.82) produced a slightly worse fitting overall model. Thus, the ten-factor model in the NZ sample produced a marginal fit due to the RMSEA of 0.049. One factor congeneric modelling of each of the ten factors however, indicated the construct validity of each of the factors was good to excellent. Although the ten-factor model was not as good as the eight-factor model, the ten factors were all valid.
The ten factors were named (highest loading item with factor loadings included in brackets):
1. School Leadership, Climate, Decision-Making (item with the highest loading ‘Satisfaction with Leadership in your school’, .86);
2. Promotion Procedures and Opportunities (‘ ... current criteria for promotion ‘, .83);
3. School Infrastructure (‘ ... your school’s material resources’, .67); 4. School Reputation (‘ .. pupil behaviour in your school’, .72); 5. Status and Image of Teachers (‘ ... status of teachers in society’, .83); 6. Student Achievement (‘ ... your capacity to influence pupil achievement’, .78);7. Pastoral Care (‘ ... your capacity to change student behaviour’, .62 ); 8. Workload (‘ ... your current workload overall’, .86);9. Change and Change Management (‘ .... degree of support to implement change’,
.71);10. Professional Self-growth (‘ ... your acquisition/development of professional
skills’, .91).
6.2.3 United States of America Study
Examination of the 10-factor model in the US sample failed to produce a fit. ( 2χ = 2383.76 df = 763 p<0.001, RMSEA = 0.059, GFI = 0.85, AGFI 0.82, NFI = 0.79). There was no ‘simple’ modification providing a better fitting model than the originally tested model. Testing the original 8-factor model produced an even worse
11
fit. Further, one-factor congeneric modelling failed to confirm the 10 factors. Thus, exploratory procedures were implemented to gain an understanding of the data from US teachers.
Following similar procedures as previously used, 42 items were found to produce 16 factors. These items overlapped somewhat with the original 42 used in the modelling of the Australian, New Zealand and British data, but also included some different items. The 16 factors produced an acceptable model on a number of the goodness of fit criteria, ( 2χ = 1543.95, df = 680, p<0.001, RMSEA = 0.045, GFI = 0.90 ,AGFI = 0.86, NFI = 0.91). The 16 factors produced from the US data were very similar to the originally hypothesised 16 factors (Dinham, 1992).
The sixteen factors were named (highest loading item with factor loadings included in brackets):
1. Student Welfare (item with the highest loading ‘Satisfaction with support structures for students with special needs in your school’, .99);
2. Teacher Welfare (‘ ... support structures for teachers in your school’, .96); 3. Teachers’ Associations (‘ ... the way teacher unions work for the betterment of
education’, .98); 4. Extra Curricula (‘ ... your involvement in extracurricular activities’, .95); 5. Community Involvement (‘ ...the degree of community involvement in your
school’, .88); 6. Student Attitudes and Behaviour (‘ ... pupil behaviour in your school’, .90);7. Collegiality (‘ ... your dealings with teachers in your school’, .84); 8. Inservice (‘ .... the range of professional inservice courses offered to
teachers’, .82);9. Professional Self-growth (‘ ... your mastery of teaching content’, .82).10. Educational Change (‘ .... the amount of educational change in recent
years’, .81)11. School Infrastructure (‘ ... your school’s financial resources’, .90)12. Professional Efficacy (‘ ... your capacity to change student attitudes’, .89)13. Recognition (‘ ... the amount of recognition you receive for your efforts from
people in your school’, .82)14. Student and Community Interaction (‘ ... your dealings with parents’, .87)15. Empowerment {‘ ... opportunity for involvement in school decision
making’, .81)16. Workload (‘ ... your current workload overall’, .81)
12
6.2.4 Malta and Cyprus
Neither the Anglo-Australian model nor the USA model fitted the Maltese and Cypriot data, due, it is suspected, to contextual differences. Exploratory factor analysis revealed scale structures that were similar to Herzberg’s original model, with ‘core business’ and ‘conditions of work’ items in separate clusters, but no indication of the existence of discrete ‘Third Domain’ factors.
6.3 Motivation
Motivation was measured in two ways – first, participants were invited to agree or disagree that seven orientation to teaching and two preparedness to teach items were true of them at the time they commenced their careers. Second, participants completed the modified version of the Novacek and Lazarus Commitments scales (1990).
6.3.1 Orientation to Teaching
Table 3 contains the percentages of teachers from each country who agreed with the nine orientation to teaching/preparedness to teach items. Across the five countries, patterns of motivations for entering teaching were very similar. In all three countries ‘always wanting ‘ to be a teacher was the most frequently endorsed reason for entering the profession.
Table 3: Orientation to and Preparedness for Teaching, by Country
Item___________________________
Aust Engl NZ Malta USA Can Cyp % True
49 45 46 66 56 42 63 I always wanted to teach40 40 43 41 36 42 29 not my first choice of career20 18 24 16 6 12 N/A lack of other options13 6 10 6 7 6 15 family pressure34 31 32 36 28 29 55 hours and holidays44 32 40 50 49 39 52 family commitments 10 7 6 7 3 12 41 salary54 65 60 69 68 40 58 realistic view of teaching38 39 44 69 64 38 69 adequately prepared for teaching
6.2.2 Commitments
Table 4 reports the sample means and standard deviations for the six Commitments scales.8 For each of the four countries, Affiliation is the most strongly endorsed personal commitment, followed closely by Altruism and Personal Growth. The
8 The 40 item Commitments Scale was used as a measure of motivation/commitment. Novacek and Lazarus’ instrument (1990) yields scale scores for six components of commitment - Affiliation, Power and Achievement, Stress Avoidance, Sensation Seeking, Personal Growth, and Altruism. The scales indicate one’s broad ‘life values’, rather than actual commitment to one’s career.
13
commitment which received the lowest endorsement in all three countries was the need for Power and Achievement.
Table 4: Sample means and Standard Deviations for the Commitments Scales
Country*______________________________________________________
Australia England N Z USA
Scale
Affiliation 5.53 (1.04) 5.46 (1.05) 5.54 (1.03) 5.86 (.77)Altruism 5.39 ( .89) 5.34 ( .87) 5.44 ( .84) 5.45 (.87)Personal Growth 5.38 ( .90) 5.39 ( .82) 5.48 ( .81) 5.66 (.76)Stress Avoidance 5.14 (1.03) 5.15 ( .97) 5.03 (1.01) 5.02 (.94)Sensation Seeking 5.12 (1.04) 5.11 (1.02) 5.24 ( .99) 5.17 (.89)Power and Achievement 4.93 ( .94) 4.94 ( .92) 4.90 ( .88) 5.09 (.83)
* Commitments Scales were not administered to Canadian, Maltese and Cypriot samples
6.3 Satisfaction Items and Scales
Table 5 presents the means for self-rating of overall satisfaction, and change in satisfaction for the five countries surveyed to date.
Table 5: Mean Overall Satisfaction, Change in Satisfaction and GHQ Scores by Country.
.Aust Eng N Z USA Malta Canada Cyprus_________________________________________________
Overall 4.07 3.90 4.25 5.12 5.27 4.32 5.33Satisfaction (1.67) (1.66) (1.54) (1.66) (1.16) (1.60) (1.73)
Change in 3.41 3.22 3.66 4.67 4.91 3.51 5.38Satisfaction (1.78) (1.71) (1.70) (1.70) (1.37) (1.83) (1.03)
GHQ 2.11 2.20 2.11 1.85 2.7 2.7 N/A(mental (0.50) (0.52) (0.47) (0.36) ( 43) ( .36)health)
14
Table 6 contains the mean level of satisfaction each of the 10 aspects of satisfaction scales for the three countries where these apply - (1 = Highly dissatisfied, 4 = Neutral, 7 = Highly satisfied), while Table 6 shows the rankings for the 10 satisfaction scales in each country.
Table 6: Means and Deviations for Satisfaction Scales - Australia, New Zealand and England.
Country_____________________________________Australia England New Zealand
Scale
Leadership 4.37 (1.20) 4.25 (1.16) 4.63 (1.12)Promotion 3.09 (1.54) 3.64 (1.20) 4.29 (1.16)Infrastructure 3.69 (1.18) 3.44 (1.13) 4.07 (1.16)Reputation 4.31 (1.17) 4.55 ( .97) 4.68 ( .99)Status 2.18 ( .94) 2.34 ( .82) 2.76 ( .91)Student Achieve 5.36 (1.11) 5.53 ( .93) 5.45 (.91)Pastoral Care 4.78 (1.08) 4.99 ( .93) 5.02 (.91)Workload 3.46 (1.37) 2.80 (1.26) 3.44 (1.29)Change and Change Management 2.68 (1.09) 2.80 (1.01) 3.34 (1.10)Self-Growth 5.31 (1.06) 5.22 (1.01) 5.31 ( .97)
Table 7: Rankings of Anglo-Australian Scale Means by Country and Total Sample
Country_____________________________________Australia England New Total
ZealandScale
Student Achieve 1 1 1 1Self-Growth 2 2 2 2Pastoral Care 3 3 3 3Sc Reputation 5 4 4 4Sc Leadership 4 5 5 5Promotion 8 6 6 6Sc Infrastructure 6 7 7 7Workload 7 8= 8 8Change/Chnge Man 9 8= 9 9Status 10 10 10 10
15
In all seven countries the intrinsic rewards of teaching - working with students and seeing them achieve, helping them change their behaviour for the better and increasing one’s own level of professional skills and knowledge - were the most satisfying aspects of teaching. In contrast, in Australia, England and New Zealand, the systemic/societal level factors of the pace of educational change and its management, related workload, and the status and image of teaching were the least satisfying. In between these two bands of factors lay school level aspects including school leadership and decision making, school climate, communication, resources and reputation of the school in the local community.
6.4 Malta, Cyprus and the Absence of the ‘Third Domain’
Maltese and Cypriot teachers have not experienced many of the difficulties that have beset their Anglophone colleagues. The profession continues to enjoy higher levels of respect and recognition and the debilitating waves of educational change have not been experienced. It is predictable that matters that are outside practitioners’ control will be experienced as less satisfying than those aspects that can be controlled. Results from Malta and Cyprus demonstrate this phenomenon but also show that satisfaction is much higher on ‘Third Domain’ aspects than in countries marked by more educational change compounded with increased criticism of teachers and erosion of their professional status.
Table 8 compares the percentages of Maltese, Cypriot and English teachers satisfied with a selection of ‘Third Domain’ aspects. In most cases English teachers report dramatically lower levels of satisfaction.
Table 8: Percentages of Maltese (M), Cypriot (C) and English (E) teachers satisfied with selected ‘Third Domain’ aspects
Country
M C E% satisfied63 56 10 the effects of teaching on your personal/family life?53 -- 07 the pace of educational change?53 56 14 your current workload overall?50 62 14 the amount of recognition you receive for your efforts from education officials? 48 -- 15 recent changes to curricula?46 54 05 the status of teachers in society?38 29 04 the way that the politicians work for the betterment of education? 31 -- 06 the degree of support provided by education officials to implement change?31 15 01 the image of teachers portrayed in the media?13 28 02 the community’s opinion of the “official” working hours and holidays in teaching?
M = Malta, C = Cyprus, E = England
16
7 Discussion
7.1 The Three Domains of Teacher and School Executive Satisfaction
In earlier times, career satisfaction had been conceptualised as a simple continuum or single measure. However, the work of Herzberg and others who built on this confirmed the existence of mutually exclusive and different satisfiers and dissatisfiers - in other words, there were two broad measures or aspects of peoples’ career satisfaction.
As predicted from this previous research (Herzberg, Mausner & Snyderman, 1959; Sergiovanni, 1967; Dinham, 1992) and confirmed by the earlier Australian study (Dinham & Scott, 1996), it was found that teachers and those holding promotions positions in schools are most satisfied by matters intrinsic to the role of teaching. Student achievement, helping students to modify their attitudes and behaviour, positive relationships with students and others, self-growth, mastery of professional skills, and feeling part of a collegial, supportive environment are powerful satisfiers across the six samples. This finding was consistent with the results of the Commitments scale, which revealed that teachers’ strongest commitments are to affiliation, altruism and personal growth values.
Conversely, with the Australian, England, New Zealand, US and Canada samples, the major sources of teacher and executive dissatisfaction were matters extrinsic to the task of teaching children and working with other staff. These dissatisfiers are largely out of the control of teachers and schools, and found within the wider domain of society, governments, and the employing body. The nature and pace of educational change, increased expectations and responsibilities being placed on schools with resultant increases to teacher and executive workloads were also found to have contributed to the most strongly felt dissatisfiers, which included the community’s apparent poor opinion of teachers and their ‘easy’ working conditions, the negative image of teachers portrayed in the media, problems associated with change and change management, coping with added responsibilities, the perceived low level of support provided to implement change, lack of support services for teachers, and promotion opportunities and procedures which many found problematic.
Thus, the ‘two factor’ theory of teacher satisfaction derived from the work of Herzberg, Mausner and Snyderman with engineers and accountants (1959), and Sergiovanni with teachers (1967) was confirmed - in so much as the factors giving rise to teacher satisfaction and teacher dissatisfaction are considered different and thus not arranged at opposite ends of the same continuum. What had not been previously recognised, however, was the fact that the major dissatisfiers are located not within the school - i.e., within the ‘conditions of work’ - but within the broader societal context and environment each school is part of.
Thus, the study showed that teachers across five samples (Australia, New Zealand, England, USA, Canada) are most satisfied with the intrinsic rewards of facilitating student achievement and their own self-growth, somewhat ambivalent about school based factors which, in part, are a product of the leadership and decision-making processes and styles existing in their particular school, the school’s resources and its
17
relationship with its local community, and most dissatisfied with the largely extrinsic societal and employer based factors such as the status and image of teachers and imposed educational change over which they little control or influence.
Maltese and Cypriot teachers were more satisfied overall than their counterparts in the five other countries sampled. The Maltese and Cypriot data failed to reveal a set of the ‘Third Domain’ factors in Maltese and Cypriot teachers’ thinking about their occupation, a phenomenon that proved problematic for teachers in the other countries sampled. It appears from analysis of the Maltese and Cypriot data to date, that the aspects of the context of education that in the other five countries have led teachers to develop a well-articulated model of the socio-political realm, have failed to apply. In this case the ‘classic’ two-factor model of occupational satisfaction developed by Herzberg describes the Maltese and Cypriot data well.
7.2.1 Some Implications
The study has confirmed that teachers and school executive surveyed in the various countries sampled want to perform what they perceive to be the central part of their role, the facilitation of pupil achievement - something they find highly satisfying.
However, it is also apparent that school teaching staff are increasingly feeling inadequate and over burdened in the face of the rising expectations and greater responsibilities being placed upon them, particularly in the Anglophone nations sampled. There are elements of what Biddle (in Biddle, Good & Goodson, 1997: 534) and others have identified as ‘role overload’ in this situation, in that there are too many responsibilities for teachers to meet within their expected roles, (such as too many academic subjects to master) and also ‘role malintegration’, in that teachers have expectations placed upon them to solve or accomplish a variety of roles (academic, social, administrative, basic needs) that are difficult to accommodate simultaneously and which may even be counterproductive to some degree. A further concept, ‘role ambiguity’ or lack of clarity of one’s responsibilities, might also be problematic for some.
What is probably more critical is a concept not identified by Biddle in his review of role theory, that of what could be termed role resistance, whereby teachers and school executive refuse to accept or attempt to resist additional responsibilities being placed upon them such as performing social welfare roles and marketing their school, thus leading to ‘role dissensus’, where expectations for the position are not shared by all concerned.
Commensurate with this situation of role overload, role ambiguity, role malintegration, role resistance and role dissensus, is the perception that the general community does not value or appreciate - in both senses of the word - what teachers and schools do, something reflected in the poor way teachers in five of the countries believe they are regarded by society. At the same time however, it is also apparent that teachers feel that there has been an unreasonable shift of society’s and the family’s responsibilities to schools.
18
It is equally clear that the increased politicisation of education and the increasing pressure for and rate of educational change have been problematic for teachers and school executive.
It is unrealistic to believe that the clock can be turned back to a time when things were easier and simpler for teachers and schools – although the data for Cyprus and Malta hint at what this might be like - and, in any case, there never was a ‘golden age’ in education. Teaching has always been a challenging occupation and we need to look forward, not back.
What does need to happen is for all those involved and concerned with education to rethink and agree on what they expect of teachers and schools today, and in the future, and to provide the support needed, both material and moral, to ensure that these expectations have the best chance of being achieved. ‘Teacher bashing’ and talk of education being in ‘crisis’ in the media and community does little to aid in this process, nor does continuous pressure for change and the ongoing restructuring of educational systems.
Looming teacher shortages in many countries will force the issue of rethinking and reconceptualising teachers’ and school executives’ work if this issue is not tackled now. Teachers want to teach and school executive want to support the teaching and learning process. Currently, there are too many extraneous pressures and additional responsibilities interfering with this. An audit of teachers’ current work is needed, with a view to removing from teachers ‘non-core business’ such as routine administrative and clerical responsibilities, matters which can be taken up by non-teaching staff. Where specialist expertise is required which is beyond the average teacher, e.g., in the health area, technology, languages, etc., staff need to be provided to schools, particularly primary schools, where the range of current subjects and expertise required is beyond all but the ‘super teacher’.
There also needs to be a reassessment of the social responsibilities being placed on schools. If, after this reassessment, society wants schools to be de facto social welfare agencies, then once again, trained specialist staff need to be provided.
It is important to recognise that every teacher may define his or her core business and what he or she finds most rewarding in different ways. Thus, ‘one size fits all’ may not be appropriate in the sense that common staffing formulae and job specifications will not necessarily apply to all teachers. What is needed is the flexibility for schools to consult with their staff to determine how best to meet the needs of students and teachers. For example, some teachers might like to spend more time on curriculum development, or school policy development, or sport, others less. Some might like to not teach at all, but to support others to do so through developing curriculum resource materials, teaching resources, becoming more involved in pupil welfare, taking a leading role in information technology or contributing more to school administration. Additionally, some teachers and school executive might vary their workload over the course of a year, being released from some of their duties on a full or part time basis for a time in order to free others to concentrate more on teaching, for example.
19
The present context provides the ideal opportunity to reconceptualise teachers’ work, in fact, it demands it.9 If this can be done in such a way as to let teachers and school executive concentrate more on professional aspects of their roles, it will also help to limit the negative effects of the looming teacher shortage, remove or reduce many of the present sources of teacher dissatisfaction, promote teacher professionalism and ultimately contribute to the reversal of the current downward trend of teachers’ status in the community.
On the other hand, if nothing is done to rethink teachers’ work and the expectations and responsibilities being placed on our schools, the current situation will only worsen. The general community and the various educational stakeholders need to be more prepared to listen to the concerns of teachers and teachers need to be less defensive and more open with the community. If not, then the worrying trends evident in the England, Australian and New Zealand samples in particular, are likely to continue and worsen.
References
Biddle, B.J., ‘Recent Research on the Role of the Teacher’, in Biddle, B.J., Good, T.L. & Goodson, I. (1997). International Handbook of Teachers and Teaching. Boston: Kluwer.
9 See Dinham, S. & Scott, C. ‘Reconceptualising Teachers’ Work’, paper presented to the Australian College of Education, National Conference, Canberra, September 28th, 1998.
i International Teams:Australia, Steve Dinham, Catherine Scott (then University of Western Sydney, now
University of New England)England, Catherine Scott, Sue Cox (Nottingham-Trent University)Malta Gino Stagno Mule (University of Malta)New Zealand, Richard Harker, Colin Gibbs, Kama Weir, Heather Ryan (Massey University),
David Adams (Okato College, Taranaki)USA: Ron Capasso and Tom Monaghan (, Rowan University, New Jersey); Barbara Stone
(Imagination Group, St Louis)Cyprus, Michalinos Zembylas (Intercollege, Nicosia, Cyprus and Michigan State University,
USA), Elena Papanastasiou (Department of Education, University of Cyprus, Nicosia, Cyprus)
Canada, Heather Ryan, University of Regina.
Authors
Steve Dinham, PhD, is Professor of Teacher Education, Pedagogy and Professional Development, School of Education, University of New England, Australia.Email: [email protected]
Catherine Scott, PhD, Senior Lecturer in Educational and Developmental Psychology, School of Education, University of New England, Australia.Email: [email protected]
20
Dinham, S. (1992). ‘Human Perspectives on the Resignation of Teachers from the New South Wales Department of School Education: Towards a Model of Teacher Persistence’, Doctor of Philosophy thesis, University of New England, Armidale.
Dinham, S. ‘Societal Pressures and Teaching’, paper presented to the Australian Association for Research in Education, Annual Conference, Newcastle, 27th November - 1st December, 1994.
Dinham, S. (1996). ‘In Loco Grandparentis?: The Challenge of Australia’s Ageing Teacher Population’, International Studies in Educational Administration, 24(1), pp. 16-30.
Dinham, S. (1997). ‘Teaching and Teachers’ Families’, Australian Educational Researcher, AARE, August, 24(2), pp. 59-88.
Dinham, S. & Scott, C. (1996). The Teacher 2000 Project: A Study of Teacher Satisfaction, Motivation and Health. Penrith: University of Western Sydney, Nepean.
Dinham, S. & Scott, C. (1997). ‘The Advanced Skills Teacher: An Opportunity Missed?’, Unicorn, Australian College of Education, November, 23(3), pp. 36-49.
Dinham, S. & Scott, C. (1998). ‘A Three Domain Model of Teacher and School Executive Satisfaction’, Journal of Educational Administration, 36(4), pp. 362-378.
Dinham, S. & Scott, C. (2000). ‘Moving Into The Third, Outer Domain Of Teacher Satisfaction’, Journal of Educational Administration, 38(4), pp. 379-396.
Hargreaves, A. (1994). Changing Teachers, Changing Times. London: Cassell.Harker, R.; et.al. (1998). Teacher 2000 Project: an international study of teacher
satisfaction, motivation and health : the New Zealand report. Palmerston North, N.Z.: Massey University College of Education.
Herzberg, F.; Mausner, B. & Snyderman, B. (1959). The Motivation to Work. New York: John Wiley & Sons.
Joreskog, K. & Sorbom, D. 1993). LISREL 8, Structural Equation Modelling with the SIMPLIS Command Language. Hillsdale, NJ: Lawrence Erlbaum.
Novacek, J. & Lazarus, R.S. (1990). The Structure of Personal Commitments, Journal of Personality, 58(4), pp. 693-715.
Qualitative Solutions And Research Pty Ltd. (1994). QSR NUD*IST. La Trobe University: QSR.
Scott, C.; Cox, S. & Dinham, S. (1999). ‘The Occupational Motivation, Satisfaction and Health of English School Teachers’, Educational Psychology, 19(3), pp. 287-308.
Scott, C., & Dinham, S. (2003). ‘The Development of Scales to Measure Teacher and School Executive Occupational Satisfaction’, Journal of Educational Administration, 41(1), pp. 74-86.
Sergiovanni T. (1967). ‘Factors Which Affect Satisfaction and Dissatisfaction of Teachers’, Journal of Educational Administration, 5(1), pp. 66-81.
Strauss, A. & Corbin, J. (1990). Basics of Qualitative Research – Grounded Theory Procedures and Techniques. Newbury Park, Ca. : Sage.
Zembylas, M. & Papanastasiou, E. (200). ‘Job Satisfaction Among Teachers in Cyprus’, Journal of Educational Administration, 42(3), pp. 357-374.
21
