Quality of School Life and Intentions for Further Education: The Case of a Rural High School Sid Bourke & Max Smith University of Newcastle, NSW A paper presented at the Annual Conference of the Australian Association for Research in Education at Adelaide, South Australia, November-December 1989. ABSTRACT The quality of school life (QSL) as perceived by students in a NSW rural high school was found to be comparable with those in a Victorian study of 50 secondary schools. The patterns of sex and year level differences were very similar with the exceptions that mean scores on the Teacher and Identity scales were less favourable in the NSW rural school, while the Negative Affect score was more favourable. Using students as the units of analysis, the independent variables year level, sex and measures of ability (background variables) and the various QSL scales (intervening variables) were entered into regression equations with measures of the students' educational aspirations and intentions as successive dependent variables (outcomes). A causal model was developed and tested. In descending order of importance, ability (assessed by the school), year level, QSL, self perception of ability and sex were important for the students' educational aspirations and intentions. Of the seven QSL scales, the General Satisfaction and the Teacher scales were the ones most frequently related to retention at school and plans for further study. Other QSL scales related to some outcomes were the Achievement and the Identity measures. INTRODUCTION Student perceptions of their life at school are important both in themselves and because of potential effects they may have on significant educational outcomes. Young people spend a significant amount of time at school, much of it compulsory, and the quality of their lives during this time is important. In addition, the proportion of Australian youth spending time beyond the compulsory years is increasing in line with expressed national priorities and targets for a more educated population. Possible relationships of student satisfaction with achievement and educational aspirations are relevant. For these reasons it is important that teachers, schools and school systems take into account concern about the quality of school life when formulating curricular and administrative policies and practices. A conceptualization of the ways in which students perceive the quality of their school lives, developed by Williams and Batten (1981), formed the basis of a quality of school life instrument developed for use with secondary students (Ainley, Reed and Miller, 1986). In a major study of almost 8500 students attending 50 government schools in Victoria in 1984, Ainley et al (1986) examined the relationships of students' perceptions of the quality of school life with school organizational patterns. The present study has used the same form of the questionnaire to examine relationships of quality of school life with the intentions of students to stay on to complete their secondary schooling and to go on to tertiary education or further studies of some kind. In June 1989 all students in all year levels of a government high school in a rural area responded to the quality of school life questionnaire. The school, being in the Hunter Region of NSW, was not remote from other schools or tertiary institutions. The staff of Dunn High School (DHS), which is not its real name, were primarily interested in using the questionnaire results to evaluate aspects of the school program, especially in the student welfare area. The work described in this paper was subsequent to the initial use of the data, and relates the data to other information concerning the students' educational aspirations and intentions collected at the same time. MEASURING THE QUALITY OF SCHOOL LIFE Although both are based on the original work by Williams and Batten (1981), measurement of the quality of school life (QSL) has followed slightly different paths at primary and secondary levels (see Bourke, 1986 and Ainley & Bourke, 1988 for descriptions of the primary development, and Ainley, Reed and Miller, 1986 for work at the secondary level). The secondary conceptualization used here divides the QSL instrument into seven different areas. Two areas relate to general feelings of satisfaction with school life and the other five domains are more specific. The areas are as follows. General Satisfaction - concerns general positive feelings about school. (e.g. My school is a place where I really like to go each day.) Negative Affect - concerns general negative personal reactions to school. (e.g. My school is a place where I feel restless.) Teachers - concerns the adequacy of interaction between teachers and students. (e.g. My school is a place where teachers are fair and just.) Status - concerns students sense of worth and prestige within the school. (e.g. My school is a place where I am treated with respect.) Identity - concerns a sense of learning about other people and getting along with other people. (e.g. My school is a place where I learn to get along with other people.) Opportunity - relates to a belief in the relevance of schooling. (e.g. My school is a place where the things I learn are important to me.) Achievement - relates to a sense of being successful in school work. (e.g. My school is a place where I am a success as a student.) The domains are not independent of each other but together they provide a view of the various aspects of school life for students who recorded their answers on a four point Likert scale. Responses were coded 4 = 'Definitely Agree', 3 = 'Mostly Agree', 2 = 'Mostly Disagree', and 1 = 'Definitely Disagree'. In the case of the Negative Affect items this means that higher scores represented a worse situation of stronger negative attitudes, and lower scores indicated a better situation. The reverse was true for items in the six other domains where higher scores represented more favourable attitudes. The questionnaire was administered at Dunn High School under supervision during the roll call period. The method of responding was explained and each question was read aloud by the teacher to keep the students moving through the task and to obviate any potential reading problems. One follow up administration was carried out to include students absent on the first occasion. In all 534 students responded providing a response rate of almost 98 per cent of the school enrolment. The QSL instrument was designed to assess student attitudes in the seven areas described above by using scales developed by factor and reliability analyses. Each item was intended to contribute towards one of the general or specific scales, and the analyses were intended to confirm the scale membership of items. Factor analysis of the QSL data from DHS indicated that the internal structure of the responses was consistent with the results obtained by Ainley et al (1986). As had been necessary in the earlier study, Items 7 and 17 were not included in subscales as they proved to be inconsistent with the proposed factor structure. Scores for the seven subscales were then calculated by adding the individual item responses to provide scale scores which may be compared directly across the two studies (shown in Table 1). Some differences were revealed. DIFFERENCES IN PERCEIVED QUALITY OF SCHOOL LIFE General Scale Differences The means of four of the scales (General Satisfaction, Status, Achievement and Opportunity) were not significantly different between the Victorian and DHS studies. The means of the other three scales were different. The Teachers and the Identity scale means were lower for DHS. Students at DHS, however, responded more favourably to the Negative Affect scale than the Victorian students had. At least three factors must be taken into consideration when making such comparisons: the fact that a rural school was being compared with a statewide sample inevitably dominated by city schools; the different state systems which could have led to differences in student perceptions; and a difference of five years in data collection could all be important in a time of rapid change. Table 1. Comparison of QSL Scale Scores Between DHS and Victoria (DHS N=534, Victoria N=8464) Scales DHS VIC Sig Mean SD Mean SD Diff General 13.5 3.1 13.7 3.2 No Satisfaction Teachers 17.0 3.5 17.6 3.4 Yes1 Status 12.8 2.8 12.7 2.8 No Identity 18.3 2.9 18.8 2.9 Yes1 Opportunity 18.8 3.2 18.7 3.5 No Achievement 15.2 2.6 15.4 2.5 No Negative 9.3 2.9 9.7 3.0 Yes2 Affect Notes 1. DHS students responded less favourably than the Victorian students on the Teachers and Identity scales. 2. In this case higher scores indicate stronger negative feelings. Hence DHS students responded more favourably to the Negative affect scale. The Hunter Region of NSW is one where unemployment is relatively high and educational retention is relatively low. It is therefore of interest that students at DHS saw the same level of opportunity arising from their education as the students in Victoria. It might have been expected that DHS students would have responded less favourably to the Opportunity scale, but this proved not to be the case. The Negative Affect scale is a very general one for which responses clearly would be affected by the everyday lives of students, although the questionnaire items are linked with school experiences. The existence of lower general negative affect at DHS may be related to what is perceived as lesser general stress in rural communities. No explanation can be offered for the lower levels of student response to the Teachers and Identity scales at DHS which may be related to specific circumstances existing at that school. Student responses to the QSL questionnaire were used by staff as a basis for the formulation of policy to improve student welfare within the school. The areas of greatest difference between both the scale and single item results obtained at DHS and the large sample of 50 Victorian schools formed the focus of within school in-service discussions and intentions for subsequent action in 1990. As suggested above, the staff recognized student perceptions of the quality of their lives at school are important in themselves as well as for the students' futures, including their retention in education to the end of secondary schooling and beyond. Year Level Differences As might be expected, views of school life were not the same for students in different year levels. The pattern of results for DHS parallelled those for the Victorian study, with some variations of interest. A series of analyses of variance was conducted with year level and sex as the independent variables and the mean QSL scale scores as successive dependent variables. Table 2 shows the way in which the scales varied with year level. Scores on the General Satisfaction and Teachers scales declined until Year 9 and then rose. Opportunity and Achievement fell steadily while Negative Affect was higher in the later years, that is it became worse. Status and Identity displayed less variation between year levels than the other scales. The mean differences between year levels in the present study were significant on all but the Status and Identity scales. The eta squared statistic indicates the proportion of variance in the scale which can be attributed to differences in year level. The highest eta squared value was for the Teacher scale indicating that 9.0 per cent of the variation in the Teacher scale can be attributed to differences in year level. The Opportunity scale with 7 per cent and the Negative Affect scale with 5 per cent are also worthy of note. While these effects are considered reasonably high within the context of the model, it is clear that other factors exerted considerable influence. Eta squared values related to year level differences for the other scales were lower and the value was almost zero for the Identity scale. Table 2. Dunn High School QSL Scale Scores by Year Level1 Year Gen. Teach. Status Ident. Opport. Ach. Neg. Satis. Affect 7 14.1 18.5 12.9 18.4 19.7 15.8 9.1 8 14.2 17.7 12.8 18.2 19.7 15.6 8.9 9 13.2 17.1 13.3 18.5 19.3 15.2 8.9 10 13.1 15.5 12.7 18.4 18.0 15.2 9.0 11 12.8 16.4 12.6 18.5 17.6 14.4 10.5 12 13.3 17.3 12.2 18.0 17.7 14.0 10.9 F2 4.5* 10.0* 0.7 0.4 7.6* 3.3* 3.9* Eta sq 0.03 0.09 0.01 0.00 0.07 0.04 0.05 Notes 1. As different numbers of items were involved in calculating each scale, score comparisons in this table may be considered between year levels but not between scales. 2. F values marked * were significant at the 0.05 level. Perhaps a clearer indication of the variation by year level of scale scores from the grand mean for all students at DHS is provided in Table 3. A multiple classification analysis was used to determine these deviation scores from the grand mean. These scores have also been illustrated in Figures 1 to 7, where they have been compared with the Victorian results. When comparing DHS year level results with those of the Victorian scores, differences were considered to be significant if the DHS estimates fell outside two standard errors of the estimates for the much larger Victorian study. This was true only for two points on the Opportunity scale and for one point on the General Satisfaction scale. A detailed consideration of year level responses to each scale follows. Table 3. Deviations from the Grand Mean for All Students at DHS by Year Level for the QSL Scales Scale Grand Yr.7 Yr.8 Yr.9 Yr.10 Yr.11 Yr.12 Mean General 13.5 0.7 0.7 0.3 -0.5 -0.8 -0.2 Satis. Teachers 17.1 1.4 0.7 0 -1.6 -0.7 0.2 Status 12.8 0.1 0.1 0.4 -0.2 -0.1 -0.3 Opport. 18.9 0.8 0.8 0.5 -1.0 -1.1 -1.0 Identity 18.4 0.1 0.1 0.2 0 0.1 -0.6 Achieve. 15.2 0.6 0.4 -0.1 -0.2 -0.7 -0.8 Negative 9.3 -0.2 -0.4 -0.4 -0.2 1.27 1.6 Affect1 Note 1. In this case, negative deviations indicate more favourable attitudes. General Satisfaction (Figure 1). The DHS mean scores for this scale declined considerably from Year 8 to a minimum in Year 11, and then rose again in Year 12. Whereas it might have been expected that general satisfaction would be higher in later years, the DHS scores did not cease falling until Year 12, well after the transition to the senior school. It was claimed by some staff that students at DHS are ill-prepared for the transition and a delayed rise in general satisfaction could reflect this. Within DHS, Years 10, 11 and 12 had significantly lower scores than the more junior years. Students at DHS had a significantly higher mean scale score than the Victorian students in Year 9. The overall Victorian pattern was similar to DHS, but with larger falls in satisfaction in Years 8 and 9, and with minima in Years 9 and 11. Teachers (Figure 2). The Victorian and DHS graphs share the same basic pattern for this scale, however the DHS minimum mean score occurs in Year 10, again one year later than in the Victorian study. The pattern of decline in Years 9 and 10 and rise through Years 11 and 12 was consistent with the changes in maturity, interests and critical outlook of students in this age band. For this scale, Years 9 and 10 at DHS were significantly lower than the other junior years. This was the expected pattern. Status (Figure 3). There was no definite pattern to the variation in this scale. The DHS Year 9 students recorded the highest mean score for Status but there is no clear reason for this. The Victorian Status scale scores fell between Year 8 and 9. The DHS Status scale score in Year 9 was significantly higher than Years 10, 11 or 12. The noticeably lower level on the Status scale for Year 12 DHS students was unusual when the expected result was for increased feelings of Status among seniors. Identity (Figure 4). There was no definite pattern of mean scores for the Identity scale of either sample. The sharp fall for DHS Year 12 students could possibly be attributed to their growing sense of alienation from the much larger junior school (according to teachers, a characteristic which was quite noticeable in the school). Opportunity (Figure 5). Student responses to this scale were significantly different between the two studies at the Year 9 and Year 10 levels, although the scale means for both samples progressively declined with rising year levels. Students are faced with considering their post-school prospects in the later years and, as a result, are probably more inclined to challenge the relevance of the curriculum. The sharp decline at DHS between Years 9 and 10 is particularly worthy of note. DHS Opportunity scale scores show that Years 10, 11 and 12 saw their opportunity as significantly lower than students in other years. Achievement (Figure 6). The pattern of mean scores for Achievement by year level was similar to that for Opportunity for the Victorian study. DHS scores also showed a considerable decline. More challenging courses and growing doubts about ability may have been the dominant influences here, and it was notable that self assessment of ability scores was also lower in Years 11 and 12. The difference between any two consecutive years was not statistically significant for the responses at DHS but they were different over broader year ranges. Negative Affect (Figure 7). The pattern of mean scores for the Negative Affect scale was generally the same in the two studies. Students' negative feelings increased quite markedly as the year level increased, at DHS particularly between Years 10 and 11. Again, the reasons are most likely linked to maturity and the increasing demands of the senior years. The responses at DHS to the Negative Affect scale indicated that Year 11 and Year 12 students had significantly worse negative feelings than students in the lower year levels. Sex Differences Sex of student was related to three of the QSL scales: General Satisfaction, Teachers and Opportunity, with the girls having the more positive attitudes. The most outstanding features when the scale scores were considered jointly by sex and by year level were the poor attitudes of the boys beyond Year 7, and the very positive attitude of senior girls (especially in Year 12) on scales other than Opportunity. Ability Estimates Two estimates of student ability were made in order to compare scores on the QSL scales for students of different ability and to allow ability to be related to educational aspirations. First there was a student self-report of ability for which each student was asked to locate themselves on a five-point scale of ability in comparison with other students. The categories ranged from 'a lot above average' to 'a lot below average'. The second measure was one made by the school in allocating students to mathematics classes, the only department in the school operating with graded classes. Grade classifications were averaged across year levels to take account of the different number of classes existing at different levels. No similar measures were available for other subject areas. As might be expected, the students' self estimate of ability was related most strongly to the Achievement scale (r=0.35, n=492). It was also correlated significantly with the General Satisfaction scale (r=0.15, n=492), to Status (r=0.15), to Opportunity (r=0.11) and to the Teacher scale (r=0.09). The school ability estimate described above was also related most strongly to scores on the Achievement scale (r=0.22, n=415), and with Negative Affect (r=-0.15), Identity (r=0.11) and the General Satisfaction scale (r=0.08). Although some of these significant correlations were small, the relationships of ability with the scales was as might be predicted, higher levels of estimated ability were associated with more positive attitudes. Other Comparisons As well as the results for Victoria, it was possible to compare the DHS scale scores with those from a sample of four schools in the Hunter Valley Region of NSW (Manning, 1988). Table 4 sets out the mean scale scores for Year 11 and 12 students combined for three samples of students. Although caution should be exercised when comparing the results for one individual school with the larger samples in the other two studies, in general it would seem that the DHS scale scores were below those for both the other samples. There were two exceptions: for the Opportunity scale the DHS result was similar to that for the Hunter Region with both of these below the Victorian result, and for Negative Affect the DHS result was similar to that for Victoria, being below that for the Hunter Region. There were differences in the time of the year when the QSL questionnaire was administered in the studies. The Hunter Region data used here were collected in August for the Year 11 students and in April for Year 12, the Victorian information was collected in October, and the DHS data in June. It is possible that these differences affected the results obtained. Other data from the Hunter Valley study strongly suggest this, with a considerable decline in attitudes by Year 11 students on several of the scales between the beginning of the school year and August, and particularly on the Opportunity scale (Manning, 1988: 5). With the DHS results, one area of concern for a single school study is the possibility that there may be chance differences in the group ethos of the different year levels which would be averaged out in a larger sample. If this were so, observed differences in scale scores between year levels could be attributed to the character of particular groups of students. Year 9 at DHS reported significantly different feelings on several scales, for example, the scale score for the Opportunity domain was different for Year 9 compared with other years at DHS and was also different from the Victorian study. It is possible that this year group was unique in some ways not recorded in the study. The patterns discovered for Year 9 at DHS may not recur with future Year 9 groups and may move with this year group into Year 10. A longitudinal study would be required to answer questions of this type. Other concerns would include the different socioeconomic areas served by an individual school, and the possibility of a freak occurrence in the few days preceding the administration of the questionnaire. Any such occurrence was not obvious in the views of teachers at DHS. Table 4. Hunter Region, DHS and Victorian Students: Mean Scores on QSL Scales Years 11 and 12 Combined Scales Hunter Region1 DHS Victoria General Satisfaction4 13.6 13.1 13.5 Teachers 17.7 16.9 17.8 Status 13.72 12.4 12.8 Opportunity 17.5 17.7 18.8 Identity 19.0 18.3 18.8 Achievement 15.02 14.2 14.7 Negative Affect 9.73 10.7 10.6 Notes 1. Obtained from Manning (1988), by averaging the scores for Years 11 and 12. 2. The two scales indicated each included an additional item not in the form of the questionnaire used in the other two studies. With the assistance of Manning, these scales have been adjusted to provide comparable data. 3. Higher scores indicate less strong negative attitudes. 4. Referred to as Positive Affect in the other two studies. MEASURES OF RETENTION IN EDUCATION The relationships of student personal and background factors and the quality of school life variables with retention in education at DHS were explored. Four criteria of retention in education were investigated. These were: * the year level when students intended to leave school. Five categories ranging from before the end of Year 10 to the completion of Year 12 were offered. Given the post-secondary focus of the other questions on educational intentions and the fact that many students were already in the senior school, for most analyses this information was recoded to indicate whether the students intended to complete Year 12 or not. * whether the students planned to do any further study after leaving school. * the relative emphasis on work and study intended. The students were asked whether they intended full time work, part time study or full time study. * for those students intending some further study, the nature of the course they thought they would do. Three categories covered courses at University, College, and Technical and Further Education (including apprenticeships). The distributions of the variables used to measure these four criteria are shown in Tables 5 to 7. Table 5. Distribution of School Retention: Numbers of Students by Year Level Intention to Before end At end At end At end TOTAL Leave School/ of Year 10 Yr 10 Yr 11 Yr 12 YEAR LEVEL Year 7 1 8 4 58 71 Year 8 0 19 4 65 88 Year 9 4 22 1 66 93 Year 10 2 39 2 50 93 Year 11 . . 5 44 49 Year 12 . . . 36 36 TOTAL 7 88 16 319 430 School Retention. From Table 5 it will be noted that slightly more than 80 per cent of the total of 534 students who responded to the questionnaire gave the year level they intended to leave school, the remaining students stating that they had not decided. Of the 430 students providing a definite response, almost three-quarters indicated they intended to stay on to complete Year 12, with one-fifth intending to leave at the end of Year 10. Very few intended to leave before the end of Year 10 or at the end of Year 11. Intentions for Further Study. Of the 392 students (73 per cent) who provided a definite answer to whether they intended to undertake further study after leaving school, almost 70 per cent said that they did (see the first two columns in Table 6). However, it can be seen in column 5 of the same table that a little more than half the students intended to undertake full-time work rather than any study. These two results suggest some uncertainty on the part of a fifth of the students about the relationships possible between full-time work and part-time study. Of the other half of the students who stated that they intended some study after leaving school, about 80 per cent intended full-time rather than part-time study. Table 6. Distribution of Intentions for Study and the Mix of Study and Work: Numbers of Students by Year Level Study and/or Yes TOTAL F/T P/T F/T TOTAL Work Study Study Study Work YEAR LEVEL Year 7 40 77 13 7 34 54 Year 8 59 86 25 9 39 73 Year 9 53 78 33 3 44 80 Year 10 66 86 27 8 45 80 Year 11 33 39 23 3 12 38 Year 12 21 26 16 3 8 27 TOTAL 272 392 137 33 182 352 Type of Course. Approximately 57 per cent of the students indicated what type of educational institution they preferred to attend after leaving school (see Table 7). Almost equal numbers (about 40 per cent of those responding to this question) answered that they would do TAFE or University courses. Less than 20 per cent stated they would do a CAE course. Table 7. Distribution of Type of Course Intended: Numbers of Students by Year Level Type of Course/ TAFE1 CAE Univ Total YEAR LEVEL Year 7 17 7 20 44 Year 8 27 10 26 63 Year 9 36 8 30 74 Year 10 38 14 20 72 Year 11 7 9 13 29 Year 12 4 6 11 21 TOTAL 129 54 120 303 Note 1. TAFE includes apprenticeships and other courses. THE CAUSAL MODEL Relationships between student background information, QSL and the four measures of educational aspiration and intention were examined. A causal model, outlined in Figure 8, was proposed and the relationships between the variables in the model were examined using multiple linear regression analysis. In summary, the model hypothesised that the background measures (year level, sex and the two measures of ability) directly influenced both the students' feelings about the quality of their lives at school and their future educational intentions or aspirations. The students' perceived quality of school life also affected their educational intentions. In addition, the model proposed that the background variables had an indirect effect on the students' educational intentions, working through their perceptions of the quality of school life. The results of analyzing the model shown in Figure 8 for the four separate outcomes are displayed in Table 8. The numbers of students involved in each analysis are also shown in this table, in each case being smaller than those indicated in Tables 5 to 7 because of some missing data for the independent variables in the regression analyses. In all, 376 students were included in the analysis where the dependent variable was whether the students intended to complete Year 12, while fewer students were included in the other analyses of their educational intentions after leaving school. Only students intending to continue with study after leaving school could respond to the question about the nature of the post-secondary course they intended, and 262 were included in that analysis. The proportion of variance in the outcome variable explained by the combination of the five independent variables in the causal model shown in Figure 8 ranged from 19 per cent for the intention to continue with study after completing school, through 21 and 24 per cent for completion of Year 12 and plans for work and study (respectively), to a high of 36 per cent for the nature of the post- secondary course intended. From Table 8, it is evident that year level and sex were more strongly related to intentions for further study and to plans for balancing work and study than to the other two outcomes, although all results were in the same direction and almost all were significant. The students' perceptions of their own ability were more closely related to whether they intended full time, part time or no study and, if any, the type of course they intended. The school's assessment of each student's ability was strongly related to the type of post-secondary course intended. Again results for the measures of ability were consistently in the same direction, and all were significant. Quality of school life was most strongly related to whether the student intended to complete Year 12, and also strongly related to the type of course intended. Results were in the same direction and all were significant. Table 8. Path Coefficients for Four Outcome Variables for the Model (see Figure 1)1 Outcome Dep Var2 Year Sex Abil3 Abil4 QSL (Stud) (Schl) Complete Complete 16 (06) 11 25 31 Year 12 (N=376) QSL -22 21 15 (02) - Further Study 25 16 11 22 16 Study (N=343) QSL -13 23 16 (03) - Plans for Plans 25 15 26 18 14 Work/Study (N=304) QSL -26 19 23 (06) - Type of Course 14 09 24 39 23 Course (N=262) QSL -28 15 15 (-03) - Median All Outcomes 21 12 17 24 18 Coefft (All 4) QSL -24 20 15 (03) - Notes 1. All coefficients have been multiplied by 100, and those shown in parentheses were not significant at the 0.05 confidence level. 2. Two dependent variables are shown for each form of the model: the outcome variable itself, and the appropriate intervening Quality of School Life (QSL) variable. For Completion of Year 12, the QSL variable included the General Satisfaction and the Teacher scales; for Intentions for Further Study, it included only the General Satisfaction scale; for Plans for Work/Study balance, it included the General Satisfaction, Teacher and Achievement scales; and for the Type of Course Intended, it included the General Satisfaction, Teacher and Identity scales. 3. Each student's own ability as perceived by themselves. 4. Each student's ability as perceived by the school determining a graded class placement for mathematics. Although the numbers of students included in the four analyses and the QSL variables used were different, the four sets of results were quite consistent (see Table 8). Consequently the results have been aggregated in that table by using the median value of the path coefficient from each of the independent and intervening variables to indicate the general levels of the relationships found. These overall relationships are shown in Figure 9, with only the significant paths included in the model. Year level was generally strongly negatively related to the QSL variable, indicating that student satisfaction with school declined over most of the period of secondary schooling. Year level was also strongly related, but positively so, with the intention to continue with education at school and after. More students in higher year levels intended to complete their secondary schooling, to go on to further study, particularly full time study, to go on to University rather than College, and least of all to do a TAFE course. Sex of student was strongly related to QSL, with the girls being generally more satisfied with their lives at school. Sex was slightly less-strongly related to educational intentions generally, and not related significantly to intentions to complete school. For each outcome variable, the female students had higher educational aspirations than the males. As described above, two measures of student ability were used: the results of student responses to a five-point scale asking them to compare themselves with other students in their year level, and the graded mathematics class they were placed in. The latter was used as a measure of the school's assessment of each student's ability. While only the students' perceptions of their ability were significantly related to QSL, both the student and school measures were directly related to educational intentions, with the school measure being somewhat the stronger of the two. It should be noted that the correlation between the two ability measures (r=0.46) would have affected the size and consistency of the regression coefficients linking these variables with both QSL and educational intentions. Of course the indirect paths from the background variables (year level, sex and ability) through the quality of school life variable, would add slightly to the direct paths shown in Table 8 and Figure 9. Clearly all these variables were important for students' educational intentions. The quality of school life, as perceived by students, was also important for these intentions, and it is through improving the quality of life at school that schools can make some impact on the intentions and therefore on the futures of the students for whom they are responsible. It was clear in this study that student general satisfaction was the most consistently important of the school life measures used, with the scale measuring the students' relationships with teachers being the next most important, and the students' sense of identity achieved within the school context also important for the course they intended to pursue after leaving school. On average, 23 per cent of the variance in the outcome variables was explained by the independent and the intervening QSL variables in the model used. FINAL COMMENTS The quality of their lives at school is important in itself for the students who spend at least ten years attending such institutions, and should be considered as significant for this reason alone. Additionally, however, the current social and political emphases placed on completing secondary school and gaining post-secondary job-related educational qualifications means that some students who would have left school at the end of Year 10 are staying on to complete secondary school. This in turn makes further educational options available to them. Given the importance of students' perceptions of the school for their own well-being as well as for their educational aspirations, the need is even more apparent for individual schools to consider how to improve students' perceptions of school life. The experience at Dunn High School in rural NSW is that the quality of school life questionnaire, when linked with other information about its students and the functioning of the school, provided important information which can be used as a basis for discussion among staff and for action in the school community. REFERENCES Ainley, J. & Bourke, S. (1988). "Quality of life in primary schools". Paper presented at the Annual Conference of the AARE, Armidale, NSW. Ainley, J., Reed, R. & Miller, H. (1986). School Organization and the Quality of Schooling. Hawthorn, Victoria: ACER. Bourke, S.F. (1986). The Development of a Quality of Primary School Life Questionnaire. Hawthorn, Victoria: ACER (mimeo). Manning, E.J. (1988). "The Quality of School Life in the Senior Years". Educational and Developmental Psychologist, 5(2), pp.4-6. Williams, T. & Batten, M. (1981). The Quality of School Life. Hawthorn, Victoria: ACER. NOTE Figures 1 to 9 are available on request from Dr S.F. Bourke, Faculty of Education, University of Newcastle, N.S.W. 2308.