Research in Science Education, 1997, 27(2), 195-214

Factors Affecting Student Career Choice in Science: An Australian Study of Rural and Urban Schools Deidra J Young and Barry J Fraser

Curtin University of Technology Brian E Woolnough

University of Oxford

Abstract In 1990/91, a research study was undertaken in England on the Factors Affecting Schools' Success in Producing Engineers and Scientists (FASSIPES). This study was conducted by Brian Woolnough at the Oxford University Department of Educational Studies, in conjunction with the Institute of Physics and the Institution of Electrical Engineers and National Power (Woolnough, 1991, 1994; Woolnough et al., 1997). Principally, Woolnough attempted to ascertain why young people chose to pursue a career in the physical sciences and engineering. In addition, characteristics of schools which appeared to influence students to pursue a study of science were investigated. A number of countries have since replicated this study as an international research cooperative and the National Key Centre for School Science and Mathematics, Curtin University of Technology, Perth, Western Australia participated on behalf of Australia. Currently, the following countries have contributed to FASSIPES International in addition to England and Australia: Canada, China, Japan and Portugal. In this study, we compared the career aspirations of 729 Year 11 and 12 science and physics students, from 20 rural and urban high schools in Western Australia. While we were particularly interested in students who chose science and engineering careers, we also found that career aspirations were different for rural and urban students. Summarily, our study found significant cultural differences between rural and urban students and these are discussed here, although in many ways these students had very similar reasons for choosing to become a scientist or engineer. This article describes some of the similarities and differences found between rural and urban students. Two distinct types of schools (urban versus rural) were found to contribute to a school culture. In this article, the term "urban" refers to those schools located in the metropolitan urban and suburban areas of Perth, Western Australia. The term "rural" refers to those schools located in what is generally referred to as rural and remote areas of Western Australia based on statistical local areas. This classification was first developed by Arundell using the 1986 census data (Department of Primary Industries and Energy, 1991) and is currently being updated by Shore using the 1991 census data (Australian Bureau of Statistics). People living in rural Western Australia have many different social, cultural, economic and educational pursuits. This study was not meant to be all-inclusive, but rather attempted to look into a few schools and their local communities in both rural and urban locations. These schools ranged markedly in demographic terms and we have only begun to understand how the community influences these schools. A major research project investigating rural education in Western Australia over a five year period (1995-99), the Western Australian School Effectiveness Study, seeks to elucidate those features which make rural schools effective despite considerable obstacles (see Young, 1996, 1997).

196

YOUNG, FRASER AND WOOLNOUGH Background to Western Australian Research in Rural Schools

That rural students participate in education to a lesser degree than students from metropolitan areas has long been the subject of concern to both policy makers and the government (Department of Primary Industries and Energy, 1988; Dawkins & Kerin, 1989). Why this low level of participation of rural students in secondary and tertiary education exists is important for the future of young people in rural Australia. Major hurdles to participation in and completion of postcompulsory schooling and higher education were identified in rural and remote areas of Western Australia by McGregor and Latchem 1(1991, p. 3): 1. 2. 3. 4. 5. 6.

lack of a culture conducive to the pursuit of learning and higher education study; lackof courses at the local level (many students cannot or do not wish to leave their home areas and, in any case, leaving is disadvantageous to regional development); lack of comprehensive Year 11 and 12 courses facilitating access to higher education; lack of information about courses and careers, entry requirements, entry procedures, and detailed up-to-date information on units, personnel and practices; lack of tutorial and peer support in study through distance education; lack of local study centres; studying in isolation proves very daunting for remote/disadvantaged students.

A small study conducted by Stevens and Mason (1992, p. 62) in one rural school in Western Australia found that the "main influence on the career choices have come from members of the students' nuclear families, particularly from mothers." These researchers found a lack of interpersonal support for young people in their study of a small Western Australian community. Most of the support provided appeared to come from within the students' nuclear families. Another important factor was the socioeconomic background of the local community and of the student's family. There is some indication that what is being measured in rural-urban difference studies is socioeconomic status and/or ethnicity. For example, Easton and Ellerbruch (1985) found that the poorer rural US students scored considerably lower on citizenship and social studies tests than did students from upper socioeconomic urban communities. In another US study which held socioeconomic level and ethnicity constant, no urban-rural achievement gap was found (Edin~on & Martellaro, 1984). Occupational and educational aspirations of rural young people is of considerable importance to rural Australians. It is not enough to have the fight attitude and the top tertiary entrance examination score, if the student faces insurmountable barriers to accessing further education and employment. However, in conducting research with a number of rural Australian teachers previously, many teachers felt that students' aspirations were too high! That is, these students were trying to sit tertiary examinations in Year 12 when the teachers did not feel that these students had any hope of gaining a high enough score to get into a university. Of course, this is anecdotal evidence and little research has been conducted in rural schools to follow student aspirations. It is also arguable that these teachers had generally low expectations of their students to go on to higher education. Stringfield and Teddlie's exemplary research into 16 paired rural and urban schools in the United States suggested that teachers in rural schools had higher expectations for their students, when compared with teachers in urban schools (Stringfield & Teddlie, 1991). The problems faced by Australian rural students are confounded. Fundamentally, when these students grow from adolescence to mature adulthood, they must also face the reality that there is little for them in their town/farm/rural area. In order for these students to attain their potential in life choices, they must make a choice. Either they can stay with their families in their rural location and enjoy the rural lifestyle they are accustomed to, or they must move to the city to

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

197

either look for work or further their education in vocational colleges or university. It is obvious to these students that education will expand and fulfill their lives; often parents send their children to boarding schools in the city in order to prepare them for the new changes which lie ahead. Unfortunately, some of these students, who are accepted into higher education courses, become extremely lonely and disheartened and return to their rural home. Of course many others are keen to leave home and become independent. It appears that this is sometimes related to the social network which rural students develop when they arrive in the city. Hektner attempted to disentangle the rural young person's aspiration for social mobility and preferences for residing in rural locations (Hektner, 1995). Irk his study of midwestern US schools, Hektner found a substantial amount of conflict experienced by rural students in choosing to leave or stay at home. Rural students were more likely to have conflicting aspirations about wanting to live at or near home and wanting to move out in order to move up. Stevens' investigation of influences on vocational choices of senior high school students in a rural community demonstrated that rural students have to make career decisions at an earlier age than urban students (Stevens, 1995). This study also found a significant difference between the rural working class and the rural middle class. That is, parents who are able to send their children to boarding schools in order to complete the final two years of high school did so from a superior financial base. In the rural school which Stevens studied, there was negligible provision for students to complete their high school education, with the result that the working class families were disadvantaged and unlikely/unable to send their children to boarding schools. Further, Stevens noted a difference in the students' perceptions of the world and their ability to cope in an urban school environment. Many rural students were supplied with inadequate information and counselling in order to choose their school subjects for their chosen occupations and also experienced conflict regarding the superiority of the urban lifestyle which lay before them. These findings are similar to McCracken and Barcinas (1991), whose study of rural schools in Ohio revealed that rural students tended to be more homogeneous, come from larger families and have lower socioeconomic status. Rural parents tended to have a lower educational attainment and were less likely to expect their children to attain an education beyond high school. McCracken maintained that these parental and home influences helped to explain why rural students chose lower educational courses. In the United States, rural youth were more likely to select vocations which they had been able to observe or experience, such as a~icultural college or technical colleges. Students in rural areas had lower income expectations, did not observe many highincome workers and those students who were bright and capable tended to be sent away to complete their education. The discrepancy in educational aspirations between rural and non-rural students seems clear, yet the reasons for it are not. Initiatives to raise students' aspirations in rural settings have had limited research foundations. However, it is the hope of a number of researchers that research can be developed which can make a difference both to the research field and to the student (Breen, 1989; Cobb, Mclntire, & Pratt, 1989; Hansen & McIntire, 1989; McCaul, 1989; Pratt & Skaggs, 1989; Preble, Phillips, & McGinley, 1989; Quaglia, 1989; Reid, 1989; Sherwood, 1989; Walberg, 1989). There appears to be a distinct relationship between socioeconomic status, occupational aspirations and educational aspirations and this theme has been the subject of research by Hailer and Virkler (1993). These important relationships framed this research study of the psychological, socioeconomic and classroom influences on occupational aspirations and educational aspirations.

198

YOUNG, FRASER AND WOOLNOUGH The Western Australian Context

In Western Australia, schools are scattered throughout metropolitan and rural locations with wide ranging community values and beliefs. There are farming, mining, tourism, fishing communities, with small and large schools. Many of these schools are in remote locations, while others are located in established, conservative towns. While.students in rural and remote locations participate in similar courses to those students residing in metropolitan Perth, they often have limited access to higher education course information. These students also tend to be motivated by different social factors than students from suburban Perth. The purpose of this paper is to describe some of these out-of-school and in-school factors.

Research Framework This research was based upon the hypothesis that the students' choice of career is influenced by their ability and personality, by the experiences they had in school and out of school, and by the values placed on careers in science and engineering by society (Woolnough, 1994). If we knew what certain schools did which made them successful in producing large numbers of engineers and scientists, other schools might benefit from this knowledge (Woolnough, 199 I). In the UK study, it was found that out-of-school factors were clearly important; these included the students' own ability and personality, the students' home back~ound and the recognition and status that society gives to science and engineering. However, this study also showed that schools themselves and the experience that students receive within them do make a difference. The purpose of this research was to establish whether these trends were similar in Western Australia. That is, to what extent do out-of-school and in-school factors influence a student's choice of career in science or engineering.

Methodology During 1993 and 1994, a total of 729 Year 11 and 12 science students were surveyed in 20 Western Australian high schools located in rural and urban locations. This research project also involved an interview procedure. The questionnaires were derived from the Woolnough study in England (Woolnough, 1991, 1994; Woolnough et al., 1997). In the questionnaire, Year I 1 and 12 students were asked about the subjects they studied, the courses they hoped to study in higher education, influences of family and other relatives, and what kind of career they hoped to achieve. They were also asked about their classroom activities and to describe their personality. Students were asked how they reacted to various types of activities in their science lessons, the factors which influenced their choice of a career in higher education in one of the physical sciences or engineering. Finally, students were asked to complete a personality profile using a seven-point semantic differential. These students were also asked to provide an open-ended response about how they came to decide to, or decide not to, pursue a career in science or engineering. Further, twelve students who indicated that they planned to continue their studies in the field of engineering and science were interviewed from four of the rural schools using focus groups. To preserve anonymity, these students and schools were given pseudonyms. We used an interview schedule for the students who were interviewed in focus ~oups. While an interview schedule was used, we tried to remain flexible and respond to the students' description of family and other

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

199

influences in their decisions. The purpose of these interviews was to provide us with in-depth information about why students decided to choose a career in science or engineering.

Preparation of Variables and Scales For the purposes of this study, we were interested in those factors which would influence a student to aspire to a career in science or engineering. The outcome measure for these analyses was a d u m m y variable (HiEdPhs): 1 if the student indicated an intention to study and pursue a career in science or engineering and 0 if not. The students were asked three types of questions in the questionnaire in order to address the research questions. We investigated the student items in Student Activity in School Science Classroom Scales (5-point Liken Scale), Encouraging and Discouraging Influences Scales (5-point L i k e n Scale) and the Personality Trait Items (7-point Semantic Differential Scale), using the following scales as described in Woolnough (1994) as a guide (see Appendices A, B and C for the individual items). Although adhering to Woolnough's scales in principle, there were some minor changes due to the results of the statistical analyses suggesting that some items should not be included. Following are the specific research questions guiding the study and the variables and scales associated with each research question. Research Question 1: Do schools make a difference to the students' career choice? Do students' career choices vary with the type and location of the school: rural/urban; private/public; size of school; singlesex/coeducational? This research question was investigated by including four school level variables in the statistical analyses which were three dummy variables and an ordinal variable: Rural School (1 for rural and 0 for urban schools); Private School (1 for private and 0 for public schools); Size of School (an ordinal categorical measure ranging from 1 to 8; 1 reflecting a small school to 8 reflecting a large school); Single-sex School (1 for single-sex schools and 0 for coeducational schools). Research Question 2: Does the student activity in the school science classroom influence the students' career choice of science or engineering? Two scales were used as described below in order to answer this research question. Student Activity in School Science Classroom Scales (see Appendix A): StudCent TeaCent

Student centred teaching (Items 1, 6, 10, 15) Teacher centred teaching (Items 2 and 9)

Research Question 3: Are there out-of-school and in-school influences on the students' career choice? Six scales were used as described below in order to answer this research question. Encouraging and Discouraging Influences Scales (see Appendix B): ExCurAct

Extracurricular activities (Items 12, 13, 15, 16)

200

YOUNG, FRASER AND WOOLNOUGH InClassAct CareerAsp ExtFacts DiffofSub HEIncent

In-class activities (Items 1,2, 7, 8, 10) Career aspirations (Items 21-23) External factors (Items 26 and 27) Difficulty of subject (Items 17 and 18) Higher education incentive (Item 19)

Research Question 4: Does the personality of students make a difference to their career choice? One scale and three items were used as described below in order to answer this research question. Personality Traits (see Appendix C): Extrovert Thing Slogger Sharp

Extroverted personality (Items 4, 11, 13, 15, 16) Preference for material things, not people (Item 5) Hardworking person(Item I) Clever person (Item 2)

Each scale was prepared using the confirmatory factor analysis procedure in LISREL, except for the scales HEIncent, Thing, Slogger and Sharp, which had only one item for each scale. LISREL requires a minimum of two items in order for analyses to proceed. Because maximum likelihood estimation assumes that the items are continuous responses, the weighted least squares method of estimation was used. This type of estimation is desirable for items which have ordinal response categories such as the Likert scale in making adjustments for the items not being normally distributed and categorical in nature (see Chapter 7, LISREL 8 User's Reference Guide, J6reskog & S6rbom, 1996). While confirmatory factor analysis estimates factor scores using the Weighted Least Squares procedure, a congeneric model for each scale was estimated, the composite variables were constructed using a proportionally weighted scale score as described by Rowe (1996, p. 47). This technique involves dividing the factor recession score obtained in the confirmatory factor analysis by the total factbr're~ession scores for that scale. The resulting scale is now comparable to the other similar scales. The importance of preparing composite scales from categorical response items using this type of methodology is that: LISREL's One Factor Congeneric Measurement Model accounts for the individual and joint measurement error of the items, minimising this error and increasing the reliability of the composite scale; and, LISREL's Weighted Least Squares procedure minimises problems with the ordinal, non-normally distributed items by calculating an appropriate weight matrix from the estimated asymptotic covariance matrix W of the polychoric and polyserial correlations (J6reskog & S6rbom, 1993, p. 45). The procedure used here consisted of estimating the ksi (~) and the regression factor scores and then converting this score to a proportion of 1.0. Where there was only one item used for a scale, no confirmatory factor analyses could be estimated and these was simply 1.0 for one item. There were eight scales which were suitable for analysis using the one-factor congeneric model and these results are presented in Table 1; these were scales 1 to 7 and scale 9. These results show good reliabilities of the items (multiple correlations), large loadings (Lambda X's) and goodness of fit statistics which are reasonable (greater than 0.80). Scales 8 and 10 to 12 were not suitable for this type of analysis. While the scales 1 to 9 were combined into a single nine-factor model, there were far too many variables for this analysis, with not enough data available to produce an asymptotic covariance matrix. Therefore, the goodness of fit statistic is a poor estimate of reliability of a structural equation model.

F A C T O R S A F F E C T I N G S T U D E N T C A R E E R C H O I C E IN S C I E N C E

201

Table 1

One-factor and Nine-factor Confirmatory Factor Analysis (Weighted Least Squares) Scale

Item

1.

StudCent

Science Lesson I, 6, 10, 15

Lambda X (LX) and Multiple correlations (R:) 0.81 0.84 0.81 0.79 1.00 1.00

Chisquare

Goodness of fit

Coefficient of determination

325.54

0.81

0.69

1.00

1.00

2.

TeaCent

Science Lesson 2, 9

0.88

0.88

1.00

1.00

88.15

0.89

na

0.94 1.00

0.94 1.00

173.28

0.91

0.82

377.59

0.84

0.80

92.65

0.94

na

148.86

0.80

0.40

35.37

0.97

na

na

na

na

na

352.45

0.85

0.85

3.

ExCurAct

Career 12, 13, 15, 16

0.89 0.83 1.00

1.00

4.

InClassAct

Career 1, 2, 7, 8,

0.89

0.86

0.82

10

0.83

0.78

1.00

1.00

1.00

1.00 0.87 1.00

1.00

5.

CareerAsp

Career 21, 22, 23

0.91 1.00

0.91 1.00

6.

ExtFacts

Career 26, 27

0.86

0.86

1.00

1.00

0.92 1.00

0.92 1.00

na

na

na

na

0.83 0.89

0.83 0.80

0.84 1.00

1.00

1.00

1.00

na na na nn na

na na na na na

na

na

na

na

na

na

na

na

na

na

na

na

na

na 3201.18

0.76

7.

DiffofSub

Career 17, 18

8.

HEIncent t

Career 19

9.

Extrovert

Person 4, 11, 13, 15, 16

1.00

10. Thing Person*

Person 5

11. Slogger Person*

Person I

12. Sharp Person*

Person 2

9-Factor Model

Factors 1-9

t For those scales with one observed item only, it was not possible to perform confirmatory factor analysis.

In this investigation, the sample size was too small for the number o f variables to be measured simultaneously using the weighted least squares m e t h o d o l o g y (see J 6 r e s k o g & S6rbom, 1996, p. 173, A p p e n d i x B. 18, o f P R E L I S 2 manual). J6reskog and S6rbom introduced a m i n i m u m s a m p l e size restriction for estimating asymptotic c o v a r i a n c e matrices, an essential part o f the w e i g h t e d least squares m e t h o d o l o g y . The m i n i m u m sample size is (k(k - 1))/2, w h e r e k is the n u m b e r o f variables. F o r this study, a total of 50 observed variables were being analysed with a s a m p l e size o f 729. T h e m i n i m u m required sample size, after listwise deletion, was 1225 and there simply was

202

YOUNG, FRASER A N D W O O L N O U G H

too many variables and too few cases. On a final note, the purpose of this study was to examine the effect of these variables on student career choice o f science or engineering and the one-factor models provided the necessary information about the reliability and construction of the composite scales.

Comparison of Urban and Rural Schools Before proceeding to more complicated statistical analyses, the composite scales and the response variable of interest, Choice, of a Career in Science or Engineering (HiEdPhs), were investigated for urban/rural differences using a t-test procedure. Table 2

Descriptive Statistics and t-Tests Comparing Urban and Rural Schools for the Career Choice Variable and the Composite Scales Scale HiEdPhs StudCent TeaCent ExCurAct

Number of items

Mean

Cronbach Aapha reliability coefficient

Urban

Rural

t-value

1 4 2 4

.24 3.44 4.02 3.33

na 0.63 0.41 0.63

.25 3.35 4.12 3.29

.21 3.58 3.83 3.40

-.89 3.37** -3.64** 1.93"

InClassAct 5 3.64 0.66 3.59 3.71 1.82" CareerAsp 3 3.97 0.73 4.03 3.91 -1.42 ExtFacts 2 3.27 0.36 3.26 3.30 .64 DiffofSub+ 2 3.21 0.66 3.20 3.19 -.09 HEIncentt 1 3.19 na 3.19 3.21 .30 Extrovert 5 4.42 0.65 4.40 4.46 .65 Thing Person t 1 4.49 na 4.43 4.54 .78 Slogger Person t 1 5.02 na 5.10 4.91 - 1.40 Sharp Person* I 5.46 na 5.47 5.49 .23 * p < .05 t For those scales with one observed item only, it was not possible to measure reliability with Cronbach's alpha. ** p < .01

Overall, there were negligible differences between urban and rural schools for these variables, however the rural schools appeared to have less teacher centred learning and more student centred learning (see StudCent and TeaCent in Table 2). Further, the extra-curricular activities and inclassroom activities appeared to influence career choice to a greater extent among rural students (see ExCurAct and InClassAct in Table 2). However, there were no statistically significant differences in student choices of a career in science or engineering. The dependent variable HiEdPhs showed that 24% of all the Year 11 and 12 science students were planning to proceed to a higher education course in science or engineering, with 25% of the urban student cohort and 21% of the rural student cohort.

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

203

Binomial Variance Component Model The nature of the data collected is hierarchical or clustered, consisting of students grouped within classes and classes grouped within schools. In this study, we looked at the effect of students (level 1 units) grouped within schools (level 2 units). While we may hope that students are grouped within schools in a random way, there are many reasons to believe that schools may have characteristics of their own which contribute towards student outcomes. It is these school effects which we have attempted to account for in our multilevel model of analysis. Because our outcome measure of, interest was a dichotomous variable, that is HiEdPhs had two possible choices (1,0), our statistical analysis was of a binomial distribution. In order to fit a nonlinear model with a binomial distribution, the software MLn was used (Version 1.0a developed by Rasbash and Woodhouse (1995, 1996)). This software was especially useful for performing multilevel analysis of data at the school and student level and estimating multiple dependent variables of both discrete and continuous nature (including variables with binomial distributions). A more detailed theoretical description of multilevel modelling may be found in Goldstein (1995). With these two specific requirements, we found that MLn was suitable for this sort of analysis.

Results

Multilevel Model Analysis When the multilevel binomial model was estimated with the school and student effects included, the level-1 variation was 1 indicating that the assumption of binomial variation was reasonable. Between school variations were minimal at 6.38% and most of this variation was due to the private, lower socioeconomic schools in the sample. Only a few school level variables were included in this model. These included Rural, a dummy variable where rural schools were coded 1 and urban schools were coded 0; Private, a dummy variable where private schools were coded 1 and government schools were coded 0; Size of School, coded progressively from 1 to 8 from smaller to larger high schools; and Single-sex school, where single-sex schools were coded 1 and coeducational schools were coded 0. Of these school level effects, the variable Private School was statistically significant in predicting student choice of a career in science or engineering (see estimate of - 1.65 in Table 3). That is, for this small sample of private schools, the students from private schools were less likely to choose a higher education course/career in science or engineering than the students from government schools. The explanation for this anomaly is probably due to the small sample size and that the private schools in this study were Catholic schools serving the poorer communities. Of the other student level factors influencing career choice, career aspirations (,80), external factors (.94) and higher education incentive (.39) were significantly associated with the likelihood that a student would choose a career in science or engineering. For these students, job satisfaction, status, salaries, family influences, hobbies, and expected ease of gaining higher education entry and funding all influenced their decision positively. For these students, they did not consider the difficulty of the subject (.27) nor the classroom activities (-.31) and extra-curricular activities (.02). However, teacher centred learning did have a negative association with their choice (.70). When student personality scales were investigated in the same multilevel model, there was a link with the "Sharp Person" scale, that is, the sharp type of person was more likely to select a career in science and engineering. This scale was measured by the semantic differential between

204

YOUNG, FRASER AND WOOLNOUGH

clever and stupid (see item 2 in Appendix C). That is, those students who tended to categorise themselves as clever, also tended to aspire to careers in science or engineering. These findings are similar to those of Woolnough (i994), who distinguished the engineer as the Thing person and the scientist as the Sharp person. In this study, the combined engineer/scientist category was not associated with the Thing type of personality, but rather the Sharp type. Important differences did exist however in the way this study constructed personality scales, with Thing, Slogger and Sharp One being single items. This was due to the lack of cohesiveness of the items used in Woolnough's study with the Australian data. Table 3 Estimation of Student and School Coefficients for Career Choice in Science or En~ineerin8 t Explanatory model Explanatory model Fixed parameters estimate standard error School Effects Rural" school (1,0) -.37 .39 Private school (1,0) - 1.65" .56 Size of School (1-8) -. 16 .11 Single-sex school (1,0) .13 .92 Science Learning Strategies Student Centred Learning Teacher Centred Learning

-.70*

.29 .20

Influences on Career Choice Extra-curricular Activities In-class Activities Career Aspirations External Factors Difficulty of Subject Higher Education Incentive

-.02 -.31 .80* .94* .27 .39*

.34 .32 .25 .33 .19 .22

Personality Traits Extrovert Thing Person Slogger Person Sharp Person

.01 .11 .15 .31"

.20 .14 .13 .18

-.23

Random Binomial Variance Total Variance School Level 0 Student Level 1 t N = 729 Year I l and 12 physics students in 20 schools * Explanatory variable significantly influences student choice of higher education courses in science

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

205

Qualitative Research Findings Interpretive social science has traditionally avoided mixed methods because this has usually meant treating qualitative methods subservient to the positivist paradigm. However, in this study we elected to use the grounded theory methodology (Strauss & Corbin, 1990, pp. 23-25; 1994, pp. 273-285) to "clarify concepts and the relationships among them." That is: The grounded theory approach is a qualitative method that uses a systematic set of procedures to develop an inductively derived grounded theory about a phenomenon. The purpose of grounded theory method is to build theory that is faithful to and illuminates the area under study9 (Strauss & Corbin, 1990, p. 24) Twelve students were interviewed in focus groups from four rural schools. All students were asked the following research question in order to provided further information about the role of the home, the student and the school in determining a students' career choice in science or engineering: "How did you come to decide to pursue or not to pursue a career in science or englneerlne,. These interviews provided an interesting array of information regarding their career aspirations and views about science and engineering careers through individual and group interviews. Below is presented a summary of these interviews sectioned by the student's career choice. Following this summary, a descriptive account of two rural communities and their schools is presented: Mintown (a small mining community) and Bordon (a large rural centre). Pseudonyms are used here to protect the identities of the interviewees. 9

"

o'0"

Engineering A number of rural students who wanted to pursue a career in engineering described engineering summer camps at the University of Western Australia as very influential in their choice. In addition, the Australian Defence Force Academy actively recruited these students, both rural and urban. The importance of influential family members who worked in the field of engineering and work experience in the field of engineering was also highlighted. One rural student was involved in the CSIRO's (Commonwealth Scientific and Industrial Research Organisation) 2040 program and this led him to try for engineering. For urban students, there appeared to be different kinds of reasons for choosing engineering, although the National Science Summer School was an important influencing factor for at least two urban students. Some urban students appeared to have an interest in science, while others were influenced by family members. Again, work experience confirmed many students' choice of engineering. Finally, there was an awareness of the prestige and financial security provided by a career in engineering. Many rural students expressed a frustration at the lack of career guidance information available to them. These findings showed that it was not necessarily the personality type which determined a student's career choice (as indicated in the quantitative survey), but rather out of school influences such as family members and work experiences seemed to elevate or direct student aspirations towards a particular career path.

Science It was interesting to note that many students did not perceive nursing or medicine to be a scientific career and listed these careers under non-science courses. These students appeared to categorise nursing and medicine as human service type of careers. Female students who wanted

206

YOUNG, FRASER AND WOOLNOUGI-I

to study nursing were often motivated by a desire to care for people. Further, female students described their choice of a biological science career as caring for animals and the environment. In this case, the personality type was an influencing facior as shown in the quantitative survey. Non-science

courses

Of the students who did not wish to study science in the future, many female students had a negative perception of science. They felt that they would be unable to express themselves creatively if they chose a scientific career. Examples of female quotations from urban schools are: "I decided not to pursue a career in science or engineering, because I have more interest in children and helping them learn for their future." "I would not like a career in science or engineering, because the way that they have been taught in this high school is dull and boring. The majority of chemistry and physics teachers need to teach these courses in a more interesting way to stimulate the Students' minds." In the following two sections, we present a picture of two rural communities: a small mining community and a large rural centre in Western Australia. The relationship between the school and the community is described based upon interviews with students, teachers and deputy principals. In this study, this relationship was fundamental to the ambiance and environment of the school and predict that this relationship must be incorporated in further educational research. Mintown

Mintown is a small mining town in Western Australia. People in Mintown were distinctly more relaxed than those in the city (Perth). Teachers considered the relaxed pace of life in this town to be a disadvantage for student performance. They were continually frustrated by the lack of competitive drive amongst their students. We noted that these same teachers enjoyed the lifestyle themselves and seemed to have become a part of the community. While most teachers at Mintown Senior High School tended to be young and enthusiastic, due to the Education Department policy of sending inexperienced teachers out to the country to teach, there was now a growing tendency for older, family oriented teachers to apply to come to Mintown to teach. These older teachers saw Mintown as an ideal location to raise their children and get away from the city pressures. Teachers in this school reported a sense that the stress of the city was beneficial for academic achievement, but detrimental to the family and health. Most students at Mintown do not go on to university. The more able students tended to mix with lower ability students in large classes here, this is mostly due to the higher retention rate. Teachers feel that this mixing has a negative influence on those students who have the potential and ability to achieve at the tertiary level. Although able students could move to the city and attend university, they were more comfortable in the town, living with their families who earned good salaries in the mining companies. Additionally, the more able students had little incentive to study hard at home. With their peers influencing them to spend time socialising, their attitude to study was poor and difficult to improve. Three Year 12 physics students from Mintown Senior High School who planned to go on to study engineering were interviewed. Two were from Mintown and one was from Blacktown. When asked about the problems of choosing a career in the country, these students felt that the access to career information was poor and that they had no idea of their competitiveness by academic standards in the city. While they were keen to continue their studies in engineering, they were not sure whether they were good enough. The lack of competitiveness was a clear disadvantage to these students' confidence.

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

207

Two Year 12 physics students from a private high school in Mintown were also interviewed. These students appeared to be greatly influenced by family members to study science in high school. They felt that their teacher was enthusiastic and a good story teller. These students hung around other students who planned to go to university and appeared to encourage one another to continue on to higher education courses. These students were keen to get away from the town and live and study in the city. Overall, these students appeared to have more positive influences to go on to study higher education. These influences included the family, peer groups and school support systems providing an expectation that these students can aspire to tertiary studies. Bordon Bordon is a community of 29,000 people and has two government schools, Bordon Senior High School and Lane Senior High School, and one Catholic high school, St. Ruth's College. It is a regional centre serving a large area of farming and mining communities. People come to this town for exports: shopping, business, education, spare parts for machinery and it is a major port for most of the surrounding districts. It is also a government centre and wholesaling centre. Throughout Western Australia, approximately 12% of all Year 12 students don't do any Tertiary Entrance Examination (TEE) subjects. In Bordon, this figure is 30%. That is, approximately three times the number of students don't want to go to university, when compared with the state average; they do want to gain their high school graduation. The state average for students who apply to get into university is 30%, while at Bordon Senior High School this figure is 20%. There appeared to be a lower proportion of students getting into university in rural schools, of those who apply, when compared to the metropolitan settings. As the application rate is also lower, it would appear that rural students are disadvantaged in terms of their aspirations and their chances. Further research is certainly warranted here. Reasons given for the lower aspirations of students in Bordon by the Deputy Principal at Bordon Senior High School were: 1. 2. 3.

The benigness of people in Bordon makes it difficult for the children to leave. It is more comfortable for them to stay. They are not motivated to seek a university education. The family is important to them and to the overaU culture of the town. The children appear to hold old-fashioned, country West Australian values. They are very happy with the pedestrian, quiet, secure, family values, sporting club, local entertainment and local opportunities.

While there was not a lot of motivation and hard study here, the Deputy Principal felt that the school should match the community and not dance to some other tune. In Bordon, aspirations to study engineering appeared to be related to friends and family influences as in the cases of John and Robert, two Year 12 physics students from Lane Senior High School. John became interested in mining engineering because his sister extolled this type of career. She had won a scholarship from the mining industry and he saw this as an opportunity to gain financial assistance. He had also been down to Kalgoorlie recently to see how she worked and studied at the mining school. John was impressed with the community of students and the accommodation provided for them. The country life is a big factor in John's career choice. He hates the city and prefers to remain in a small, rural community. John's father was reasonably well to do with assets in local fishing areas. He was a taxi driver and very proud of his children's aspirations.

208

YOUNG, FRASER AND WOOLNOUGH

Robert does not like the cities either. He does not like being cooped up and prefers a career which would allow him to work outside. Robert has his heart set on being a geophysicist. He has a friend of the family who has written to them about being a geophysicist. Robert also has good memories of growing copper sulphate crystals at school. Born and raised on a farm, Robert's brother has a Bachelor of Agriculture from Muresk and his sister is studying dentistry. Both his brother and father are supportive of his aspirations. All of the male members of the family run the farm. The two brothers organise soil and plant tissue analysis. Additionally, there are sheep and cattle on this farm. Robert wants to try a different life for a while and perhaps return to the farm later on.

Summary of Qualitative Research In this study, we found that students tended to choose their future careers on the basis of their experiences, their family support and advice, their gender, the information presented either directly or indirectly to them at school and their own beliefs and value systems. Because most of these factors varied with the type of community in which the student lived and went to school, there were distinct differences between the rural and urban students. For example, many students chose careers which were attainable to them in the country and would provide employment for them in the country. Rural students appeared to have a wider access to alternative work experiences which were "outdoors" in nature, such as marine biology and mining engineering. Additionally, while some rural students were reportedly reluctant to leave their communities (teacher perceptions), most of the Year 12 science students whom we interviewed were optimistic about studying higher educational courses in other states or cities. This contrasted with teachers who often told of school leavers who went to Perth to attend university and returned six months later, unhappy with the city life and homesick. When students were interviewed they did not seem at all concerned about leaving their homes and families. These students appeared anxious to become independent and start new ventures in higher education. According to teacher reports, most students who succeeded in the city had good support networks and the drive to overcome their difficulties.

Summary of Quantitative and Qualitative Research By combining the emergent findings from the qualitative research (interview data) and the quantitative research (questionnaire data) we are able to demonstrate certain influences which influenced student career choice of science or engineering. Many of the factors highlighted by statistical analysis were confirmed by student interviews. For example, out-of-school and inschool experiences and activities had an important influence on the rural students' career choice. The effect of personality was not statistically significant on career choice, but it was highlighted many times in the interviews. There is perhaps doubt about the validity of the personality measures selected for the survey and these need to be further investigated. However, student voices were important to this study to clarify and add to our understanding of student aspirations and ambitions. While some influences were positive, there were negative influences as well. These are summarised in Figure 1 into three broad categories: out-of-school factors, in-school factors and personality factors. In terms of our original research questions, we found that there were school effects, classroom effects, out-of-school/in-school factors and personality effects which influenced the student choice of career. Some of our findings were confirmed by both qualitative and quantitative research, while others were indicated by one or the other methodology. Further research is necessary here.

FACTORS AFFECTING STUDENT CAREER C H O I C E IN SCIENCE

209

Discussion While this study was limited to measuring the effects of other factors on career and higher education course choices of a selected sample of Western Australian rural and urban Year 11 and 12 science students, we were aware that the low level of participation of rural students in secondary and tertiary education would form the basis of a more substantial research study. Our study involved a small number of schools and students, but the findings are pertinent to an overall understanding of why students choose science and engineering careers.

Out-of-school Factors Influencing Student Career Choice 1. 2. 3. 4. 5. 6. 8. 9. 10.

Work related experiences Friends and family support and influences Where would the career locate the person (outdoors, rural) Reluctance or willingness to leave their community and family Support networks in near the higher education institution of their career choice Prestige and financial security Financial constraints (e.g., available scholarships or support) Extracurricular activities Scientific hobbies

In-school Factors Influencing Student Career Choice 1. 2. 3. 4.

Teacher enthusiasm Access to career information and advice Teacher centred learning (negative influence) Higher education incentive - ease of entry to higher education

Personality Types Influencing Student Career Choice 1. 2. 4. 5. 6.

Caring oriented personalities preferred scientific careers involving animals and/or people Competitive drive to succeed Desire to express themselves creatively (negative influence) Career aspirations Perception of themselves as being clever (self-concept)

Figure 1. Factors influencing student career choice.

While the statistical analyses revealed effects within the classroom, there were other nonschool effects such as the student's personality and other motivational effects. Upon being interviewed, we found the importance of the home, family support, peer expectations and school support in helping a student to choose his or her career was highlighted. This study showed that the effect of the school in influencing career choice was confounded by the student's own motivation (psychological aspects) and the student's friends and family (sociological effects). While many researchers study these effects separately, especially in achievement studies, these results suggest that these variables are inter-related and confounded (Figure 2).

210

YOUNG, FRASER AND WOOLNOUGH

Figure 2. Effects on student career and higher education aspirations.

The importance of these findings lies in the variety of social, school and personal influences interacting upon the student to form his or her career choice. These findings are consistent with Woolnough (Woolnough, 1991, 1994; Woolnough et al., 1997) who described the complexity of factors which influence students' choices. In conclusion, this study suggests that further research be conducted in rural communities to identify other pertinent factors associated with student career aspirations and higher education course choices. Secondly, students and schools in rural communities should be provided with better access to a wider selection of career information and work experiences.

Acknowledgments

FASSIPES: Factors Affecting Schools' Success in Producing Engineers and Scientists The questionnaires were developed and.'tested in England by Brian E Woolnough, Oxford University Department of Educational Studies, Oxford UK. Questionnaires were modified for the Australian setting and are available from Deidra Young, Curtin University of Technology. The authors are ~ateful to Professor Murray Aitkin, previously at the University of Western Australia, and currently at the University of Newcastle Upon Tynne, for significant advice regarding the analysis of binomial outcome measures, while maintaining multilevel analytical procedures. We would also like to thank Joanne Tiros, Beverley Webster and Carolyn Montgomery for their research assistance in the collection and analysis of this data. Further, we would like to acknowledge the invaluable statistical advice provided by Kenneth Rowe, Research Fellow, Centre for Applied Educational Research, The University of Melbourne and Leigh Smith, Head of School, Psychology, Curtin University of Technology, Perth.

Correspondence: Dr Deidra J Young, Australian Research Fellow, Science and Mathematics Education Centre, Curtin University of Technology, GPO Box U1987, Perth, Western Australia, 6845, Australia. Internet email: [email protected]

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

211

References Breen, D. T. (1989). Enhancing student aspirations: A goal for comprehensive developmental guidance programs. Research in Rural Education, 6(2), 35-38. Cobb, R. A., Mclntire, W. G., & Pratt, P. A. (1989). Vocational and educational aspirations of high school students: A problem for rural America. Research in Rural Education, 6(2), 11-16. Dawkins, J. S., & Kerin, J. C. (1989). A fair go: The federal government's strategy for rural and education training. Canberra: Australian Governement Publishing Service. Department of Primary Industries angl Energy. (1988). Education in rural Australia: A discussion paper prepared for the rural and allied industries council Canberra: Australian Government Publishing Service. Department of Primary Industries and Energy. (1991). Rural remote and metropolitan regions classification. Canberra: Australian Government Publishing Service. Easton, S. E., & Ellerbruch, L. W. (1985). Update on the citizenship and social studies achievement of rural 13-year-olds. Bozeman: Montana State University. (ERIC Document Reproduction Service No. ED 262 946). Edington, E. D., & Martellaro, H. C. (1984). Variables affecting academic achievement in New Mexico schools. Las Cruces: New Mexico Center for Rural Education. (ERIC Document Reproduction Service No. ED 271 267). Goldstein, H. (1991). Nonlinear multilevel models, with an application to discrete response data. Biometrica, 78, 45-51. Goldstein, H. (1995). Multilevel statistical models. London: Edward Arnold. Hailer, E. J., & Virkler, S. ]'. (1993). Another look at rural-nonrural differences in students' educational aspirations. Journal of Research in Rural Education, 9(3), 170-178. Hansen, T. D., & McIntire, W. G. (1989). Family structure variables as predictors of educational and vocational aspirations of high school seniors. Research in Rural Education, 6(2), 39-49. Hektner, J. M. (1995). When moving up implies moving out: Rural adolescent conflict in the transition to adulthood. Journal of Research in Rural Education, 11(1), 3-14. Jrreskog, K., & Srrbom, D. (1993). New features in PRELIS 2. Chicago: Scientific Software International. Jrreskog, K., & Srrbom, D. (1996). LISREL 8: User's reference guide. Chicago: Scientific Software International. McCaul, E. (1989). Approaches to dropout prevention: The philosopher's stone revisited. Research in Rural Education, 6(2), 25-30. McCracken, J. D., & Barcinas, ]'. D. (1991). Differences between rural and urban schools, student characteristics, and student aspirations in Ohio. Journal of Research in Rural Education, 7(2), 29-40. McGregor, A. L., & Latchem, C. R. (1991). Networks for learning: A review of access and equity in post-compulsory education in rural and remote areas of the State of Western Australia. Perth: Western Australian Office of Higher Education. Pratt, P. A., & Skaggs, C. T. (1989). First generation college students: Are they at greater risk for attrition than their peers? Research in Rural Education, 6(2), 31-34. Preble, B., Phillips, P., & McGinley, H. (1989). Maine's aspirations movement: Reaching out to youth. Research in Rural Education, 6(2), 51-59. Quaglia, R. (1989). Student aspirations: A critical dimension in effective schools. Research in Rural Education, 6(2), 7-9. Rasbash, J., & Woodhouse, G. (1995). MLn command reference: Version 1.0. London: Multilevel Models Project, Institute of Education, University of London.

212

YOUNG, FRASER AND WOOLNOUGH

Rasbash, J., & Woodhouse, G. (1996). MLnfor Windows: Software for n-level analysis. London: Multilevel Models Project, Institute of Education, University of London. Reid, J. N. (1989). The rural economy and rural youth:~Challenges for the future. Research in Rural Education, 6(2), 17-23. Rowe, K. I. (1996). Multilevel modelling with MLn: An integrated course. Melbourne: Centre for Applied Educational Research, University of Melbourne. Sherwood, R. A. (1989). A conceptual framework for the study of aspirations. Research in Rural Education, 6(2), 61-66. Stevens, K. (1995). Vocational choice for senior high school students in rural Australian communities. Journal of Research in Rural Education, 11(3), 182-186. Stevens, K., & Mason, D. (1992). Making career choices in rural Western Australia. In C. Boylan (Ed.), Rural education: In pursuit of exceUence. Proceedings of the Eighth Annual Conference of the Society for the Provision of Education in Rural Australia. University of New England, Armidale, July 1992. Strauss, A., & Corbin, J. (1990). Basics of qualitative research. London: Sage. Strauss, A., & Corbin, .l. (1994). Grounded theory methodology. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research. (pp.). Thousand Oaks, CA: Sage Publications. Stringfield, S., & Teddlie, C. (1991). School, classroom and student level indicators of rural school effectiveness. Journal of Research in Rural Education, 7(3), 15-28. Walberg, H. J. (1989). Student aspirations: National and international perspectives. Research in Rural Education, 6(2), 1-6. Woolnough, B. (1991). The making of engineers and scientists: Factors affecting schools success in producing engineers and scientists. Oxford: Oxford University Department of Educational Studies. Woolnough, B. (1994). Factor's affecting students' choice of science and engineering. International Journal of Science Education, 16(6), 659-676. Woolnough, B. E., Guo, Y., Leite, M. S., de Salmeida, M. J., Ryu, T., Wang, Z., & Young, D. J. (1997). Factors affecting student choice of career in science and engineering: Parallel studies in Australia, Canada, China, England; Japan and Portugal. Research in Science and Technological Education, 15(1), 105-121. Young, D. J. (1996, November). The effect of the classroom environment on career choice: A rural perspective. Paper presented to the Joint Annual Conference of the Australian Association for Research in Education and Educational Research Association, Singapore. Young, D. J. (1997, January). Western Australian school effectiveness study: Features of effective schools. Paper presented to the International Congress for School Effectiveness and School Improvement, Memphis, Tennessee.

Appendix A How Students React to Various Types of Activities in School Science Lessons [Science Lesson Items] (Likert Response Set: Strongly A~ee, Agree, Unsure, Disagree, Strongly Disagree) 1. 2. 3. 4. 5.

I find the opportunity to plan my own experiments very satisfying I feel happiest when clear instructions are ~ven to follow during practical experiments School science is about learning scientific facts and theories School science is about learning to do science through scientific investigations Standard experiments, written up correctly, give confidence to continue with science

FACTORS AFFECTING STUDENT CAREER CHOICE IN SCIENCE

213

6. Extended practical projects show me what science is like and makes me interested in it 7. The best notes are short and concise 8. I find I need to write quite a lot in order to really express myself satisfactorily 9. I feel most confident when the science lessons are well structured and teacher directed 10. I value the opportunity when the teacher lets us plan our own activities 11. Student work is marked objectively by the teacher 12. The most effective form of assessment is self assessment by the student (me) 13. Involvement in science clubs is an unhelpful distraction from the learning of real science 14. Parents are not involved in the work of the school science department 15. Involvement in science and technology competitions is great fun 16. Local engineers can bring a stimulating dimension into science lessons. 17. Work experience in science based industry turns people off jobs in science or engineering.

Appendix B Significant Factors in Influencing Choice of Career and Higher Education Courses [Career Items] (Likert Response Set: Strongly Agree, Agree, Unsure, Disagree, Strongly Disagree) 1. The teaching in the science department 2. The personal encouragement given by science teachers 3. Supportive mathematics teaching in the school 4. Supportive science teaching in the school 5. Supportive technology teaching in the school 6. Advice from the careers staff 7. The practical nature of the science lessons 8. The intellectual satisfaction in doing science 9. The amount of involvement with human issues 10. The amount of self-expression allowed in science lesson 1 I. The tradition of good examination results in science 12. Outside speakers and visits to science firms and industrial sites 13. Local scientists and engineers coming into the school 14. Work experience in local companies 15. Involvement in science clubs (photographic, radio, etc...) 16. Involvement in science competitions (e.g., Science Talent Search; Titration Stakes; Physics Quiz) 17. The level of difficulty of the sciences at school 18. The amount of work required for school sciences 19. The ease of entry to tertiary courses in science and engineering 20. The possibility of Austudy in higher education 21. The status of jobs in science and engineering 22. The likely salaries in science and engineering jobs 23. The likely job satisfaction in science and engineering jobs 24. The way in which sophisticated technology is used in military weapons 25. The job situation in local science based industry 26. Experience of your family in science based industry 27. Scientific hobbies and fiddling with gadgets at home

214

YOUNG, F R A S E R A N D W O O L N O U G H Appendix C

Self-perception o f Personality Traits [Personality Items] (Seven point semantic differential) I. Hardworking 2. Clever 3. Introverted 4. Self-confident 5. Task-centred 6. Verbose 7. Tender-minded 8. Abstract thinker 9. Creative I0. Convergent thinker 1 I. Outgoing 12. Communicating best in words 13. Dominant 14. Conscientious 15. Adventurous 16. Self-sufficient 17. Greedy 18. Enthusiastic

1 I 1 1 1, 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

4 4 4 4 4 4 4 4 4 4 4 4 4 4 4. 4 4 4

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7

Lazy Stupid Extroverted Insecure Person-centred Concise Tough-minded Practical worker Systematic Divergent thinker A loner Best in diagrams Submissive Unreliable Timid Dependent on others Generous Disinterested