View in EDS HTML Full Text PDF Full Text

Choosing 'in' Schools: Locating the Benefits of Specialisation

Saved in:
Bibliographic Details
Title: Choosing 'in' Schools: Locating the Benefits of Specialisation
Language: English
Authors: Davies, Peter, Davies, Neil, Hutton, David, Adnett, Nick, Coe, Robert
Source: Oxford Review of Education. Apr 2009 35(2):147-167.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 325 Chestnut Street Suite 800, Philadelphia, PA 19106. Tel: 800-354-1420; Fax: 215-625-2940; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 21
Publication Date: 2009
Document Type: Journal Articles
Reports - Research
Education Level: High Schools
Higher Education
Descriptors: Educational Objectives, Outcomes of Education, Academic Achievement, Foreign Countries, Course Selection (Students), Student Educational Objectives, Educational Policy, Policy Analysis, Specialization, Majors (Students), Pretests Posttests, Alignment (Education)
Geographic Terms: United Kingdom (England)
DOI: 10.1080/03054980802643298
ISSN: 0305-4985
Abstract: Recent policy in England has suggested that educational outcomes will be raised if schools specialise in particular subjects. In contrast, calls for the reform of 16-19 education have suggested that these outcomes will be improved if students become less specialised in their studies. At present, there is a limited evidence base from which to judge these arguments. In particular, we do not know the extent to which students' achievements in 16-19 education are higher when they choose subjects which play to their perceived strengths. We also do not know whether students are more likely to choose to study subjects taught by more effective departments. That is, outcomes may be affected by the relative strengths of students or departments in circumstances where there is freedom to choose. In this paper we provide evidence of the existence and strength of these relationships. This evidence suggests that reducing the scope within schools for specialisation or competition will reduce average student attainment and these effects ought to be taken into account when evaluating alternative curriculum policies. (Contains 4 tables.)
Abstractor: As Provided
Number of References: 35
Entry Date: 2009
Accession Number: EJ834864
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFkwxz6NtD5E4w8MuqjAIDtAAAA4TCB3gYJKoZIhvcNAQcGoIHQMIHNAgEAMIHHBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDAzDDcFfshTHdmS2UwIBEICBmRWUqE0ljuDOt342R3bvc2HL2GJb3L5UKgAaKI1PhZzKiIllaAakv7391BYWjH5SOApV98WygIU29nvKt3-A7eYMSa9h_99wQVK_Ww6EVRBHowVIga4SsDxGE5JNTJnus9sNoExSsntelTbGCV3_OIwHoAI0zap0b4o1v6z65XYR1lVv0Bw-PQhuQgB8rcSQ7yYvOb7E1ivNlA==
Text:
  Availability: 1
  Value: <anid>AN0037265477;oxr01apr.09;2019Mar14.13:48;v2.2.500</anid> <title id="AN0037265477-1">Choosing in schools: locating the benefits of specialisation. </title> <sbt id="AN0037265477-2">Introduction</sbt> <p>Recent policy in England has suggested that educational outcomes will be raised if schools specialise in particular subjects. In contrast, calls for the reform of 16–19 education have suggested that these outcomes will be improved if students become less specialised in their studies. At present, there is a limited evidence base from which to judge these arguments. In particular, we do not know the extent to which students' achievements in 16–19 education are higher when they choose subjects which play to their perceived strengths. We also do not know whether students are more likely to choose to study subjects taught by more effective departments. That is, outcomes may be affected by the relative strengths of students or departments in circumstances where there is freedom to choose. In this paper we provide evidence of the existence and strength of these relationships. This evidence suggests that reducing the scope within schools for specialisation or competition will reduce average student attainment and these effects ought to be taken into account when evaluating alternative curriculum policies.</p> <p>Choice and specialisation in schooling can be configured in different ways. Choice may be offered without specialisation if schools offer a common compulsory curriculum. Alternatively, choice may be offered between schools that are differentiated by the curriculum they offer, creating specialisation at the school level. In addition, choice may be offered within schools by making subjects for study options rather than requirements, generating specialisation at the pupil level. There is considerable variation in configuration of choice and specialisation across OECD countries and for some countries it is an aspect of policy that is under review. A programme of secondary school specialisation in England began in 1994 with 50 schools. By 2007 most secondary schools had adopted a curriculum specialisation. A baccalaureate qualification (IWA,[<reflink idref="bib21" id="ref1">21</reflink>]) has recently been introduced in Wales although a similar reform in England has, for the moment, been dismissed (DfES, [<reflink idref="bib9" id="ref2">9</reflink>]).</p> <p>At present, there is a very limited evidence base from which to evaluate the above policy options. This may be reflected in a contribution to recent debate by the chief inspector of schools in England (OfSTED, [<reflink idref="bib30" id="ref3">30</reflink>]): 'My own longstanding beliefs about the value of a rounded education are based <emph>on the superb grounding that I received as a pupil in a comprehensive school</emph>' (italics added). With regard to specialist schools, evidence that they can be a major instrument for transforming attainment levels is currently weak (Levačić & Jenkins, [<reflink idref="bib23" id="ref4">23</reflink>]). With regard to specialisation by pupils within schools, the curriculum for 16–19 year‐olds in England was broadened in 2000. However, this was too limited in scope to have much effect on the extent of specialisation (Hodgson <emph>et al</emph>., [<reflink idref="bib18" id="ref5">18</reflink>]). In contrast, the earlier, more radical, reform in Scotland had, according to Gamoran ([<reflink idref="bib14" id="ref6">14</reflink>]), an effect of reducing sorting by social background and may therefore be seen as a 'good thing'. Nevertheless, evidence (Dolton & Vignoles, [<reflink idref="bib11" id="ref7">11</reflink>]; Johnes, [<reflink idref="bib22" id="ref8">22</reflink>]) suggests that there is no relationship between a broader curriculum at Advanced Level (A Level) and subsequent earnings.</p> <p>We add to the evidence base through an investigation of the extent to which students' choices of optional subjects to study in 16–19 education reflect their own relative advantage and the relative effectiveness of departments responsible for teaching the optional subjects. The case for broadening the curriculum is affected by the strength of these effects. If students' performance in A Level subjects is strongly affected by their aptitude towards different subjects then reducing the scope for them to specialise in preferred subjects would reduce their average attainment. Similarly, if students are more likely to choose to be taught by departments that achieve higher levels of value added, then their average attainment would fall if this opportunity were taken away from them. Moreover, a possible source of incentives for improvements in schooling would disappear. There are grounds for believing that these sources of variation within schools could be more important to students' achievement than the variation between schools which has been the focus of choice and specialisation in government policy in England (e.g. House of Commons, [<reflink idref="bib19" id="ref9">19</reflink>]; DfES, [<reflink idref="bib8" id="ref10">8</reflink>]). The majority of variation in school effects on pupils' attainment appears to be attributable to differences within rather than between schools (Tymms, [<reflink idref="bib34" id="ref11">34</reflink>]; Nye <emph>et al</emph>., [<reflink idref="bib28" id="ref12">28</reflink>], Rivkin <emph>et al</emph>., [<reflink idref="bib31" id="ref13">31</reflink>]).</p> <p>We begin by outlining a case for anticipating benefits from increased subject choice. This is followed by sections that describe the model and the data used in this study. Our results section is followed by conclusions in which we discuss the findings in the context of previous research and suggest some implications for current policy and future research.</p> <hd id="AN0037265477-3">The rationale for expecting benefits from subject choice</hd> <p>Three arguments suggest that students might achieve higher grades if they are able to choose subjects to study. First, a student might have a relative advantage in a particular subject arising from variation in their processing abilities. The relative advantage argument depends on the existence of variation in abilities of different kinds of processing and students' personal learning history. Differences in orientation towards spheres of learning have been termed 'frames of mind' by Hudson ([<reflink idref="bib20" id="ref14">20</reflink>]) and Gardner ([<reflink idref="bib15" id="ref15">15</reflink>]).</p> <p>Second, students might develop a self‐concept through comparing their achievements with that of peers, leading to an inference that they are 'good at' one subject because they seem to perform relatively better in that subject when compared with others (Marsh, [<reflink idref="bib26" id="ref16">26</reflink>]; Eccles & Wigfield, [<reflink idref="bib12" id="ref17">12</reflink>]). Uerz <emph>et al</emph>. ([<reflink idref="bib35" id="ref18">35</reflink>]) found that students' past achievements in mathematics and languages relative to other subjects influenced their choices in their final year of secondary schooling in Holland. We might anticipate that students' awareness of any such relative advantage will improve as they gain more evidence through test and examination performance over time (Manski, [<reflink idref="bib25" id="ref19">25</reflink>]). This awareness contributes to expectations of success or failure in particular subjects. Students are more likely to achieve higher grades if they are studying subjects for which their capabilities are more suited and if they believe they are likely to be successful (Trautwein <emph>et al</emph>., [<reflink idref="bib33" id="ref20">33</reflink>]).</p> <p>Third, students might find a subject more congruent with their interests and ambitions, leading to higher motivation. Studies in different countries report the effect of students' interests and aspirations on subject choice (e.g. Elsworth <emph>et al</emph>., [<reflink idref="bib13" id="ref21">13</reflink>]; Stokking, [<reflink idref="bib32" id="ref22">32</reflink>]; Cleaves, [<reflink idref="bib6" id="ref23">6</reflink>]). When students are more highly motivated to study a subject they are more likely to persevere in the face of difficulties and be more likely to aim for deep rather than surface learning.</p> <p>Combining all three of these factors in a model of student choice, Nagy <emph>et al</emph>. ([<reflink idref="bib27" id="ref24">27</reflink>]) find significant correlations between students' subject achievement, self‐concept, interest and enrolment in specialist courses amongst Year 12 <emph>Gymnasium</emph> students in Germany. To the extent that students' specialisation improves their average grades it will also improve their future earnings (Dolton & Vignoles, [<reflink idref="bib11" id="ref25">11</reflink>]).</p> <p>Students' capacity to base their choice of subjects to study on their relative advantage depends on the adequacy of the information they have at the time of making their choices. In the case of A‐Level subjects in England students make provisional subject choices before they know the results from the public examinations they take when aged 16. However, the widespread use of 'target setting systems' in English secondary schools means that students receive explicit information about their relative performance, aside from the judgements they form on the basis of routine assessments. On this basis we would expect their relative advantage to be more important in choosing subjects such as French or mathematics for which they are able to use their previous performance in the subject as a guide than in choosing subjects like psychology or media studies which they are unlikely to have studied before.</p> <p>The reasons for anticipating a positive effect of subject choice on departmental effectiveness are discussed by Adnett and Davies ([<reflink idref="bib1" id="ref26">1</reflink>]). Students face incentives to choose subjects which are taught more effectively in so far as this will extend the range of choice they face when applying to Higher Education Institutions and might also improve their employment prospects. The publication of league tables of school performance also provides an incentive for schools to guide students towards subjects that are taught more effectively. The organisation of schooling in England provides an opportunity to distinguish between these two (student and teacher) effects. Whilst a majority of students continue their 16–19 education in the same school in which they have studied for the previous three years, a substantial minority change to study in a tertiary college. Students who switch institution at age 16 are less likely to have reliable information about the relative effectiveness of teaching of different subjects. We might therefore expect the relative performance of subject departments in sixth‐form colleges to exert less influence on rates of enrolment than the relative performance of departments within schools that cater for the full 11–18 age range.</p> <p>Previous studies of students' choice of A‐level subjects in England provide evidence of the scale of specialisation in students' choices (Bell <emph>et al</emph>., [<reflink idref="bib3" id="ref27">3</reflink>]), the effect of previous performance (e.g. Garratt, [<reflink idref="bib16" id="ref28">16</reflink>]), and career aspirations and interests (e.g. Cleaves, [<reflink idref="bib6" id="ref29">6</reflink>]). Lumby and Wilson ([<reflink idref="bib24" id="ref30">24</reflink>]) also show that type of institution attended (school or college) is associated with variation in 16–19 year‐olds' curricular choices. We add to this literature by providing a large scale multivariate analysis of the factors affecting the probability of a student choosing to study a range of A‐level subjects. We also add to the international evidence on the effect of students' relative advantage in subjects in two ways. First, our estimation of this advantage is more comprehensive than that provided by Uerz <emph>et al</emph>. ([<reflink idref="bib35" id="ref31">35</reflink>]). Second, we provide an estimate for England that may be compared with countries in which students are expected to choose a larger (e.g. the Netherlands) or smaller (e.g. Germany) number of subjects. We also include an estimate of the comparative effectiveness of subject departments. Davies <emph>et al</emph>. (forthcoming) found a small effect of the difference in past value added difference between history and geography departments on students' choice of GCSE subjects. We are not aware of any similar study for 16–19 year‐old students.</p> <hd id="AN0037265477-4">The model</hd> <p>Our model explains the probability of a student entering for examination in each of 13 subjects: biology, business studies, chemistry, economics, English, French, geography, German, history, media studies, maths, physics and psychology. Three criteria dictated this selection: subjects that attract a large number of students at A Level; subjects that fall within the groupings science, humanities and social sciences; and a mix of subjects that students will have studied before the age of 16 and subjects that they are unlikely to have studied before.</p> <p>Ideally, the model would take account of the effects of simultaneous choice between these subjects at three levels: the school, the individual and subject chosen. This can be estimated using a multivariate response model. However, these models are highly computationally intensive. Our models that simultaneously estimated subsets of choices were unstable, encountering difficulties with the convergence of the maximum likelihood iteration. This suggested that some of the assumptions of the simultaneous model were being violated. One such problem was that not all subjects were offered at all schools. This meant there was not enough variation for some of the subjects at certain schools. This problem can be overcome by dropping the school level effects. However, given that the school level effects were one of our variables of interest this solution was unsatisfactory and we explored alternative specifications.</p> <p>We estimated several models. Our first model is estimated independently for each of the 13 subjects. This model has two levels, schools and individuals, although the inclusion of an estimate of students' comparative advantage provides an aspect of a third level since students' abilities in each subject are modelled relative to their ability in the other subjects they are offered. We subsequently estimated more complex, simultaneous models using a randomly selected sub‐sample of the data. Details of these models are provided in an appendix. However, since they produced results very similar to a more simple model we concentrate on that simple model in our presentation. This random effects logit model was specified as:</p> <p>Graph</p> <p>Where <emph>i</emph> = 1,..., n students</p> <p>  <emph>s</emph> = 1,...,S institutions</p> <p>Graph</p> <p>This model is estimated independently for each of the 13 subjects. Subject is a dummy variable taking 1 if a student is awarded an examination grade for a particular subject, 0 otherwise. Given that few students who study for A Level are not entered for an examination then this is likely to provide a good indication of initial enrolment levels, although this relationship has been weakened by the Curriculum 2000 reforms. The vectors StChar and Family capture variation in the probability of choosing each subject due to student characteristics. These characteristics include an estimate of students' relative advantage for studying different A‐Level subjects. We distinguish between schools with sixth forms and sixth‐form colleges since, as discussed above, we expect that departmental effectiveness would be more important in the former, where students have easier access to information about departments' performance. Teachers in schools with sixth forms might also have better information about students' relative advantage strengthening the effect of this variable in these schools. This distinction is enabled through separate regressions for the two types of institution. The vector Dept estimates the impact of effectiveness of subject departments. This variable is lagged by three years to take account of the lapse of time between the year in which students made their subject choices and the year in which they took the final examination. The vector Year which consists of dummies for the years 1999–2001 is included to control for exogenous time trends.</p> <p>One potential source of variation over time was introduced by the Curriculum 2000 reform which saw the introduction of AS examinations one year into a student's two‐year course of study for A Levels. Whilst this reform left the typical number of subjects studied for A Level unchanged (at three), it did mean that students were able to study an additional subject for the first year of their A‐Level course and to choose which three out of their four subjects they would continue with in the second year. As a robustness check we compared results for the pre‐ and post‐reform periods in our data.</p> <p>Our model treats subject choice as a simple binary decision (you either choose the subject or not). This imposes a strong simplification on the data. In total, there are more than 20,000 different combinations of A‐Level subjects chosen by students in the years covered by this study (Bell <emph>et al</emph>., [<reflink idref="bib3" id="ref32">3</reflink>]), and even with our restricted set of 13 subjects there are 8192 different possible combinations that students could theoretically choose. This could be significant for our measure of departmental effectiveness. In principle we would expect students to be interested primarily in the importance of one department relative to another. If they are considering two subjects as alternatives, then better relative performance in department <emph>x</emph> makes it less likely that they will choose subject <emph>y</emph>. If students are considering taking two subjects as complements then better relative performance by department <emph>x</emph> may make it more likely that students will choose subject <emph>y</emph>. Evidence from studies of the impact of recent curricular reform in England (e.g. Hodgson <emph>et al</emph>., [<reflink idref="bib18" id="ref33">18</reflink>]) indicates that the majority of students are still choosing to specialise in a group of related subjects, so this complementary effect may be strong (see Appendix for a test of this hypothesis). We tested for this by including the previous value‐added performance of each of the other departments in the model. This produced very weak and inconsistent results. When we split our data for schools with sixth forms into the periods before and after the Curriculum 2000 reforms only one relationship was consistently significant across the two periods: a negative relationship between the relative performance of geography departments and the likelihood of students entering for examination in psychology.</p> <p>Given these results, we present a more straightforward estimation in this paper. We use value added rather than absolute departmental performance because the average GCSE achievement of students varies substantially from one subject to another (as shown in our results). To take account of the lag between students' choice of subject and the year in which they complete their studies our model estimates the effect of A Level value‐added departmental performance in year <emph>t‐3</emph>. Value‐added departmental performance is estimated as the departmental average residual from individual level regression of A‐level results on average GCSE results; this gives a measure of departmental performance for each year.</p> <p>To estimate students' relative advantage in their choice of A‐Level subjects we regressed grades achieved by A‐level students on a basket of their GCSE results (equation 1). The GCSE subject grades used were: mathematics, English language, average grade for science, average grade for modern foreign languages and average grade for history and geography. At the time when the data were collected all students were required to study GCSE English, mathematics and science, one modern foreign language and either history or geography. By using this range of subjects we allow for variations in students' relative advantage between subjects without substantially reducing our sample size. This equation is estimated independently for each cohort using OLS pooled across all schools. This gives estimates of average performance of all students at A Level in each year for given combinations of performance at GCSE.</p> <p>(<reflink idref="bib1" id="ref34">1</reflink>)</p> <p>Graph</p> <p>where <subs>I</subs> = 1.....n students and</p> <p>  <subs>k</subs> = 1...13 subjects</p> <p> t = 1996–1999</p> <p>Using lagged coefficients from equation (<reflink idref="bib1" id="ref35">1</reflink>) we then calculated the A‐Level grade that a student might expect in each of the following subjects: mathematics, English language, physics, chemistry, biology, geography, history, French, German, business studies, economics, psychology and media studies (equation 2).</p> <p>(<reflink idref="bib2" id="ref36">2</reflink>)</p> <p>Graph</p> <p>The grades were expressed in numerical terms currently used to express grade values where E= 40, D=60, C=80, B=100, A=120. From the results of equation (<reflink idref="bib2" id="ref37">2</reflink>) we were able to produce a predicted grade for each student in each subject. We then subtracted the average (of subjects offered at each institution) of these from the predicted grade in each subject to yield a measure of the student's relative advantage in that subject (equation 3).</p> <p>(<reflink idref="bib3" id="ref38">3</reflink>)</p> <p>Graph</p> <p>The prediction of the model is that as relative advantage increases so will the probability of entering for examination in the subject. Since this measure of relative advantage is estimated by comparing a student's achievements with our whole sample, it ignores comparison effects within the institution at which the student is studying. That is, it excludes peer comparisons suggested by Marsh ([<reflink idref="bib12" id="ref39">12</reflink>]) and Eccles and Wigfield ([<reflink idref="bib26" id="ref40">26</reflink>]).</p> <hd id="AN0037265477-5">The data</hd> <p>The dataset is a sample of schools and sixth‐form colleges from the Advanced Level Information System (ALIS) project of the Curriculum Evaluation and Management (CEM) Centre at Durham University. The institutions included in our data set all participated in the ALIS project for each of the years between 1999 and 2003, provided that at least 50 students were entered by that institution for at least one A level in each year. According to the Audit Commission ([<reflink idref="bib2" id="ref41">2</reflink>]) the minimum viable number of students in a sixth form is 150. This figure is for two cohorts of students including those studying vocational qualifications. According to OfSTED ([<reflink idref="bib29" id="ref42">29</reflink>]) sixth forms with fewer than 100 students (in two cohorts) offer a restricted curriculum. So our minimum size criterion is pitched at a level that is consistent with official judgements on the size of sixth form required to provide a satisfactory range of choice for students.</p> <p>Our total sample size is 622 institutions, although the numbers of institutions offering each subject is sometimes much lower than this (see Table 1). However, the overall sample is a large proportion of the total number of institutions of this size which were offering opportunities for A‐level study. For example, in 2000, the middle year of our sample, there were 834 schools with sixth forms in which at least 50 students were entered for A‐level examinations. Moreover, in a study of the factors affecting the likelihood of students entering for examination in seven subjects at GCSE, Davies <emph>et al</emph>. (forthcoming) found that differences between schools had little effect. Only one school level variable, the proportion of students eligible for free school meals, was significant for each subject. It is likely that this is less important in 16–19 education as students are much more likely to stay on into the sixth form if they come from a higher socio‐economic background: 17‐year‐old students whose parents are 'higher professional' and students with at least one parent with a degree have an 81% probability of being in full‐time education in England. Meanwhile, 17‐year‐old students whose parents are in 'routine' jobs have only a 51% probability of full‐time study, whilst students with neither parent with an advanced level qualification have only a 57% probability of continuing in full‐time study (DfES, [<reflink idref="bib10" id="ref43">10</reflink>]). For these reasons we consider the risk of sample bias to be small and unlikely to affect the generalisability of the results. Tables 1 and 2 present descriptive data for variables included in our models.</p> <p>Table 1. Variation by A‐level subject in percentage of students and institutions</p> <p> <ephtml> <table><thead valign="bottom"><tr><td>Row</td><td>Variables</td><td><bold>No of colleges offering</bold></td><td><bold>No of schools offering</bold></td><td><bold>Percentage of students in the sample entering for examination in this subject in those institutions offering the subject</bold></td></tr><tr><td /><td /><td /><td /><td><bold>In whole sample (%)</bold></td><td><bold>In colleges (%)</bold></td><td><bold>In schools (%)</bold></td></tr></thead><tbody><tr><td>1</td><td>Biology</td><td char=".">82</td><td char=".">527</td><td char=".">22</td><td char=".">16</td><td char=".">23</td></tr><tr><td>2</td><td>Business Studies</td><td char=".">80</td><td char=".">297</td><td char=".">18</td><td char=".">15</td><td char=".">19</td></tr><tr><td>3</td><td>Chemistry</td><td char=".">71</td><td char=".">496</td><td char=".">17</td><td char=".">10</td><td char=".">18</td></tr><tr><td>4</td><td>Economics</td><td char=".">53</td><td char=".">77</td><td char=".">14</td><td char=".">14</td><td char=".">15</td></tr><tr><td>5</td><td>English</td><td char=".">87</td><td char=".">535</td><td char=".">34</td><td char=".">34</td><td char=".">33</td></tr><tr><td>6</td><td>French</td><td char=".">63</td><td char=".">443</td><td char=".">9</td><td char=".">5</td><td char=".">10</td></tr><tr><td>7</td><td>Geography</td><td char=".">76</td><td char=".">515</td><td char=".">20</td><td char=".">12</td><td char=".">21</td></tr><tr><td>8</td><td>German</td><td char=".">44</td><td char=".">283</td><td char=".">5</td><td char=".">3</td><td char=".">6</td></tr><tr><td>9</td><td>History</td><td char=".">80</td><td char=".">516</td><td char=".">20</td><td char=".">14</td><td char=".">21</td></tr><tr><td>10</td><td>Mathematics</td><td char=".">82</td><td char=".">534</td><td char=".">23</td><td char=".">16</td><td char=".">25</td></tr><tr><td>11</td><td>Media Studies</td><td char=".">47</td><td char=".">267</td><td char=".">13</td><td char=".">6</td><td char=".">14</td></tr><tr><td>12</td><td>Physics</td><td char=".">72</td><td char=".">524</td><td char=".">14</td><td char=".">9</td><td char=".">15</td></tr><tr><td>13</td><td>Psychology</td><td char=".">84</td><td char=".">198</td><td char=".">19</td><td char=".">20</td><td char=".">19</td></tr></tbody></table> </ephtml> </p> <p>Table 2. Student background variables included in the model</p> <p> <ephtml> <table><thead valign="bottom"><tr><td /><td><bold>Students and Family Characteristics</bold></td><td><bold>For all students</bold></td><td><bold>For students in colleges</bold></td><td><bold>For students in schools</bold></td></tr></thead><tbody><tr><td>1</td><td>Average GCSE Points</td><td char=".">6.019</td><td char=".">5.765</td><td char=".">6.118</td></tr><tr><td /><td /><td char="."><bold>% of all students</bold></td><td char="."><bold>% of students in colleges</bold></td><td char="."><bold>% of students in schools</bold></td></tr><tr><td>2</td><td>Gender (Male)</td><td char=".">46.61</td><td char=".">45.33</td><td char=".">47.11</td></tr><tr><td>3</td><td><italic>Student's Job Aim</italic></td><td /><td /><td /></tr><tr><td>4</td><td>5‐Managerial‡</td><td char=".">29</td><td char=".">31</td><td char=".">29</td></tr><tr><td>5</td><td>6‐Professional‡</td><td char=".">46</td><td char=".">43</td><td char=".">47</td></tr><tr><td>6</td><td><italic>Mother's Job</italic></td><td /><td /><td /></tr><tr><td>7</td><td>5‐Managerial</td><td char=".">25</td><td char=".">23</td><td char=".">26</td></tr><tr><td>8</td><td>6‐Professional</td><td char=".">10</td><td char=".">8</td><td char=".">11</td></tr><tr><td>9</td><td><italic>Father's Job</italic></td><td /><td /><td /></tr><tr><td>10</td><td>5‐Managerial</td><td char=".">38</td><td char=".">40</td><td char=".">38</td></tr><tr><td>11</td><td>6‐Professional</td><td char=".">27</td><td char=".">20</td><td char=".">30</td></tr><tr><td>12</td><td><italic>Parental Education</italic></td><td /><td /><td /></tr><tr><td>13</td><td>Mother Educated to Degree level</td><td char=".">23</td><td char=".">19</td><td char=".">25</td></tr><tr><td>14</td><td>Father Educated to Degree Level</td><td char=".">31</td><td char=".">25</td><td char=".">34</td></tr><tr><td>‡The base reference for these categories combines all other occupational categories of skilled manual and clerical (IIIA, IIIB), semi‐skilled and unskilled and not in employment.</td></tr></tbody></table> </ephtml> </p> <hd id="AN0037265477-6">Results</hd> <p>A comparison of regressions for the pre‐ and post Curriculum 2000 reforms found only minor differences, and these were present only in a minority of subjects. These findings are consistent with previous evidence (Hodgson <emph>et al</emph>., [<reflink idref="bib18" id="ref44">18</reflink>]) that Curriculum 2000 has had a very limited effect on the pattern of subject choice. For these reasons we present the results for the models for a panel including the whole 1999–2002 period. Table 3 presents results for schools with sixth forms whilst Table 4 shows results for sixth‐form colleges. The estimates are presented as raw coefficients.</p> <p>Table 3. Random effects logit regression for schools with sixth forms</p> <p> <ephtml> <table><thead valign="bottom"><tr><td /><td>Biol.</td><td>B.Std</td><td>Chem.</td><td>Econ.</td><td>Eng.</td><td>Fren.</td><td>Geog.</td><td>Germ.</td><td>Hist.</td><td>Math.</td><td>Media Studs</td><td>Phys.</td><td>Psych.</td></tr></thead><tbody><tr><td>1 Relative Advantage (RA)</td><td char=".">0.143</td><td char=".">−0.02</td><td char=".">0.109</td><td char=".">0.025</td><td char=".">0.236</td><td char=".">0.056</td><td char=".">0.030</td><td char=".">0.020</td><td char=".">0.085</td><td char=".">0.194</td><td char=".">0.027</td><td char=".">0.088</td><td char=".">−0.01</td></tr><tr><td /><td char=".">23.48</td><td char=".">−2.63</td><td char=".">20.95</td><td char=".">4.83</td><td char=".">33.03</td><td char=".">18.01</td><td char=".">6.02</td><td char=".">11.78</td><td char=".">16.27</td><td char=".">28.67</td><td char=".">3.23</td><td char=".">20.09</td><td char=".">−0.78</td></tr><tr><td>2 Department Value Added</td><td char=".">0.011</td><td char=".">0.003</td><td char=".">0.004</td><td char=".">0.002</td><td char=".">0.009</td><td char=".">−0.00</td><td char=".">0.006</td><td char=".">0.000</td><td char=".">0.003</td><td char=".">0.003</td><td char=".">0.007</td><td char=".">0.00</td><td char=".">0.009</td></tr><tr><td /><td char=".">3.52</td><td char=".">0.94</td><td char=".">1.99</td><td char=".">0.54</td><td char=".">2.49</td><td char=".">−0.10</td><td char=".">2.01</td><td char=".">0.20</td><td char=".">1.04</td><td char=".">1.29</td><td char=".">1.68</td><td char=".">−0.28</td><td char=".">1.69</td></tr><tr><td>3 Av. GCSE</td><td char=".">0.014</td><td char=".">−0.07</td><td char=".">0.068</td><td char=".">0.016</td><td char=".">0.198</td><td char=".">0.011</td><td char=".">0.043</td><td char=".">0.023</td><td char=".">0.102</td><td char=".">0.091</td><td char=".">−0.02</td><td char=".">0.026</td><td char=".">−0.02</td></tr><tr><td /><td char=".">4.37</td><td char=".">−17.1</td><td char=".">26.57</td><td char=".">6.52</td><td char=".">41.75</td><td char=".">12.16</td><td char=".">14.17</td><td char=".">18.74</td><td char=".">31.35</td><td char=".">30.48</td><td char=".">−3.46</td><td char=".">14.87</td><td char=".">−4.05</td></tr><tr><td>4 Male</td><td char=".">−0.08</td><td char=".">0.053</td><td char=".">0.013</td><td char=".">0.081</td><td char=".">−0.12</td><td char=".">−0.01</td><td char=".">0.055</td><td char=".">0.001</td><td char=".">−0.04</td><td char=".">0.083</td><td char=".">0.023</td><td char=".">0.063</td><td char=".">−0.18</td></tr><tr><td /><td char=".">−16.3</td><td char=".">9.88</td><td char=".">3.21</td><td char=".">15.55</td><td char=".">−21.2</td><td char=".">−6.84</td><td char=".">11.47</td><td char=".">0.70</td><td char=".">−9.35</td><td char=".">17.28</td><td char=".">2.84</td><td char=".">16.10</td><td char=".">−17.1</td></tr><tr><td>5 Mother Professional</td><td char=".">0.003</td><td char=".">−0.02</td><td char=".">0.001</td><td char=".">0.000</td><td char=".">0.004</td><td char=".">0.002</td><td char=".">0.012</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.001</td><td char=".">0.011</td><td char=".">−0.01</td></tr><tr><td /><td char=".">0.36</td><td char=".">−3.28</td><td char=".">0.13</td><td char=".">0.07</td><td char=".">0.54</td><td char=".">0.59</td><td char=".">1.85</td><td char=".">−0.01</td><td char=".">0.06</td><td char=".">−0.11</td><td char=".">0.04</td><td char=".">2.17</td><td char=".">−0.89</td></tr><tr><td>6 Father Professional</td><td char=".">−0.02</td><td char=".">0.030</td><td char=".">−0.03</td><td char=".">0.016</td><td char=".">0.017</td><td char=".">0.004</td><td char=".">0.024</td><td char=".">−0.00</td><td char=".">0.014</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">0.013</td></tr><tr><td /><td char=".">−3.56</td><td char=".">4.36</td><td char=".">−5.97</td><td char=".">3.16</td><td char=".">2.40</td><td char=".">1.33</td><td char=".">4.13</td><td char=".">−0.74</td><td char=".">2.49</td><td char=".">−2.56</td><td char=".">−0.56</td><td char=".">−1.44</td><td char=".">1.47</td></tr><tr><td>7 Aspiration Professional</td><td char=".">0.036</td><td char=".">0.039</td><td char=".">0.047</td><td char=".">0.045</td><td char=".">−0.02</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.016</td><td char=".">0.037</td><td char=".">−0.03</td><td char=".">0.007</td><td char=".">−0.03</td></tr><tr><td /><td char=".">6.97</td><td char=".">6.46</td><td char=".">10.39</td><td char=".">9.43</td><td char=".">−2.64</td><td char=".">−0.49</td><td char=".">−1.33</td><td char=".">−1.16</td><td char=".">3.22</td><td char=".">7.34</td><td char=".">−3.65</td><td char=".">1.67</td><td char=".">−3.66</td></tr><tr><td>8 Mother managerial</td><td char=".">−0.01</td><td char=".">−0.02</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">0.006</td><td char=".">0.005</td><td char=".">0.008</td><td char=".">−0.00</td><td char=".">0.012</td><td char=".">−0.01</td><td char=".">0.010</td><td char=".">0.00</td><td char=".">0.015</td></tr><tr><td /><td char=".">−2.34</td><td char=".">−3.49</td><td char=".">−2.16</td><td char=".">−1.80</td><td char=".">1.03</td><td char=".">1.97</td><td char=".">1.56</td><td char=".">−0.51</td><td char=".">2.31</td><td char=".">−3.29</td><td char=".">1.10</td><td char=".">−0.17</td><td char=".">1.91</td></tr><tr><td>9 Father Managerial</td><td char=".">−0.01</td><td char=".">0.005</td><td char=".">−0.01</td><td char=".">0.005</td><td char=".">0.005</td><td char=".">0.000</td><td char=".">0.015</td><td char=".">0.002</td><td char=".">−0.01</td><td char=".">0.007</td><td char=".">0.001</td><td char=".">0.005</td><td char=".">0.00</td></tr><tr><td /><td char=".">−1.12</td><td char=".">0.87</td><td char=".">−1.33</td><td char=".">1.01</td><td char=".">0.81</td><td char=".">0.05</td><td char=".">2.92</td><td char=".">1.23</td><td char=".">−0.85</td><td char=".">1.35</td><td char=".">0.08</td><td char=".">1.35</td><td char=".">−0.33</td></tr><tr><td>10 Aspiration Managerial</td><td char=".">−0.05</td><td char=".">0.009</td><td char=".">−0.03</td><td char=".">−0.01</td><td char=".">−0.03</td><td char=".">0.001</td><td char=".">0.048</td><td char=".">0.003</td><td char=".">−0.02</td><td char=".">0.047</td><td char=".">−0.04</td><td char=".">0.050</td><td char=".">−0.02</td></tr><tr><td /><td char=".">−8.05</td><td char=".">1.30</td><td char=".">−4.75</td><td char=".">−2.40</td><td char=".">−3.45</td><td char=".">0.43</td><td char=".">7.76</td><td char=".">1.16</td><td char=".">−2.94</td><td char=".">6.80</td><td char=".">−4.29</td><td char=".">8.52</td><td char=".">−2.10</td></tr><tr><td>11 Mother with Degree</td><td char=".">0.005</td><td char=".">−0.03</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.001</td><td char=".">−0.00</td><td char=".">0.002</td><td char=".">0.006</td><td char=".">0.00</td><td char=".">−0.00</td><td char=".">0.00</td><td char=".">0.002</td></tr><tr><td /><td char=".">0.87</td><td char=".">−4.03</td><td char=".">−0.82</td><td char=".">−1.04</td><td char=".">−0.13</td><td char=".">0.41</td><td char=".">−0.74</td><td char=".">1.03</td><td char=".">1.05</td><td char=".">−0.85</td><td char=".">−0.23</td><td char=".">−0.39</td><td char=".">0.19</td></tr><tr><td>12 Father with Degree</td><td char=".">0.009</td><td char=".">−0.03</td><td char=".">0.021</td><td char=".">−0.00</td><td char=".">−0.02</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.008</td><td char=".">0.007</td><td char=".">−0.01</td><td char=".">0.002</td><td char=".">−0.01</td></tr><tr><td /><td char=".">1.63</td><td char=".">−4.87</td><td char=".">4.67</td><td char=".">−0.57</td><td char=".">−2.71</td><td char=".">−1.33</td><td char=".">−2.43</td><td char=".">−0.24</td><td char=".">1.56</td><td char=".">1.49</td><td char=".">−1.30</td><td char=".">0.50</td><td char=".">−1.18</td></tr><tr><td>13 RA X Mother Professional</td><td char=".">0.004</td><td char=".">0.005</td><td char=".">0.004</td><td char=".">−0.00</td><td char=".">−0.00</td><td char=".">−0.00</td><td char=".">0.001</td><td char=".">−0.00</td><td char=".">0.010</td><td char=".">0.005</td><td char=".">0.020</td><td char=".">−0.01</td><td char=".">0.019</td></tr><tr><td /><td char=".">0.53</td><td char=".">0.60</td><td char=".">0.67</td><td char=".">−0.14</td><td char=".">−0.16</td><td char=".">−0.31</td><td char=".">0.13</td><td char=".">−0.97</td><td char=".">1.58</td><td char=".">0.68</td><td char=".">1.69</td><td char=".">−1.14</td><td char=".">1.64</td></tr><tr><td>14 RA X Father Professional</td><td char=".">0.00</td><td char=".">0.002</td><td char=".">0.006</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.002</td><td char=".">−0.00</td><td char=".">0.003</td><td char=".">0.00</td><td char=".">0.013</td><td char=".">0.004</td><td char=".">−0.01</td></tr><tr><td /><td char=".">0.18</td><td char=".">0.28</td><td char=".">1.24</td><td char=".">−1.39</td><td char=".">−0.65</td><td char=".">−0.35</td><td char=".">0.41</td><td char=".">−0.98</td><td char=".">0.58</td><td char=".">−0.06</td><td char=".">1.46</td><td char=".">1.06</td><td char=".">−1.09</td></tr><tr><td>15 RA X Aspiration Professional</td><td char=".">0.022</td><td char=".">0.019</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.048</td><td char=".">−0.00</td><td char=".">0.005</td><td char=".">0.001</td><td char=".">0.013</td><td char=".">−0.01</td><td char=".">0.028</td><td char=".">−0.02</td><td char=".">0.037</td></tr><tr><td /><td char=".">3.99</td><td char=".">3.30</td><td char=".">−1.93</td><td char=".">−0.79</td><td char=".">7.36</td><td char=".">−0.21</td><td char=".">1.11</td><td char=".">0.94</td><td char=".">2.78</td><td char=".">−2.11</td><td char=".">3.64</td><td char=".">−4.42</td><td char=".">4.77</td></tr><tr><td>16 RA X Aspiration Managerial</td><td char=".">−0.01</td><td char=".">0.004</td><td char=".">0.005</td><td char=".">−0.01</td><td char=".">0.032</td><td char=".">−0.00</td><td char=".">0.004</td><td char=".">0.001</td><td char=".">0.007</td><td char=".">−0.01</td><td char=".">0.020</td><td char=".">0.004</td><td char=".">0.016</td></tr><tr><td /><td char=".">−0.83</td><td char=".">0.60</td><td char=".">0.78</td><td char=".">−1.39</td><td char=".">4.32</td><td char=".">−0.51</td><td char=".">0.71</td><td char=".">0.63</td><td char=".">1.14</td><td char=".">−1.09</td><td char=".">2.24</td><td char=".">0.94</td><td char=".">1.91</td></tr><tr><td>17 RA X Mother with Degree</td><td char=".">0.00</td><td char=".">0.005</td><td char=".">0.009</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.002</td><td char=".">−0.00</td><td char=".">0.001</td><td char=".">0.002</td><td char=".">0.00</td><td char=".">−0.01</td><td char=".">0.001</td><td char=".">0.011</td></tr><tr><td /><td char=".">−0.66</td><td char=".">0.74</td><td char=".">1.86</td><td char=".">−2.05</td><td char=".">−0.06</td><td char=".">0.70</td><td char=".">−0.01</td><td char=".">0.31</td><td char=".">0.29</td><td char=".">−0.74</td><td char=".">−0.57</td><td char=".">0.18</td><td char=".">1.25</td></tr><tr><td>18 RA X Father with Degree</td><td char=".">−0.01</td><td char=".">0.017</td><td char=".">−0.01</td><td char=".">0.005</td><td char=".">0.033</td><td char=".">0.002</td><td char=".">0.020</td><td char=".">−0.00</td><td char=".">0.010</td><td char=".">0.006</td><td char=".">0.012</td><td char=".">0.004</td><td char=".">0.012</td></tr><tr><td /><td char=".">−1.80</td><td char=".">2.95</td><td char=".">−1.09</td><td char=".">1.13</td><td char=".">4.84</td><td char=".">0.95</td><td char=".">4.15</td><td char=".">−0.39</td><td char=".">2.17</td><td char=".">1.14</td><td char=".">1.41</td><td char=".">1.09</td><td char=".">1.51</td></tr><tr><td>19 RA X Gender</td><td char=".">−0.03</td><td char=".">0.021</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">0.015</td><td char=".">0.006</td><td char=".">0.035</td><td char=".">0.003</td><td char=".">0.033</td><td char=".">−0.02</td><td char=".">0.019</td><td char=".">0.003</td><td char=".">0.00</td></tr><tr><td /><td char=".">−6.80</td><td char=".">4.52</td><td char=".">−2.97</td><td char=".">−1.50</td><td char=".">2.62</td><td char=".">3.17</td><td char=".">8.98</td><td char=".">2.39</td><td char=".">7.75</td><td char=".">−4.18</td><td char=".">2.75</td><td char=".">1.00</td><td char=".">−0.21</td></tr><tr><td>20 Students (n)</td><td char=".">42553</td><td char=".">28576</td><td char=".">40878</td><td char=".">29750</td><td char=".">42389</td><td char=".">40899</td><td char=".">41933</td><td char=".">32172</td><td char=".">41540</td><td char=".">42474</td><td char=".">10407</td><td char=".">42183</td><td char=".">19730</td></tr><tr><td>21 rho</td><td char=".">0.050</td><td char=".">0.079</td><td char=".">0.076</td><td char=".">0.092</td><td char=".">0.057</td><td char=".">0.069</td><td char=".">0.064</td><td char=".">0.076</td><td char=".">0.048</td><td char=".">0.092</td><td char=".">0.046</td><td char=".">0.071</td><td char=".">0.079</td></tr><tr><td>22 Log Likelihood (−)</td><td char=".">20279</td><td char=".">12384</td><td char=".">15199</td><td char=".">21847</td><td char=".">9885</td><td char=".">19672</td><td char=".">4898</td><td char=".">20074</td><td char=".">15027</td><td char=".">4164</td><td char=".">9968</td><td char=".">12551</td><td char=".">9764</td></tr></tbody></table> </ephtml> </p> <p>Table 4. Random effects logit regression for sixth‐form colleges</p> <p> <ephtml> <table><thead valign="bottom"><tr><td /><td>Biol.</td><td>B.Std</td><td>Chem.</td><td>Econ.</td><td>Eng.</td><td>Fren.</td><td>Geog.</td><td>Germ.</td><td>Hist.</td><td>Math.</td><td>Media Studs</td><td>Phys.</td><td>Psych.</td></tr></thead><tbody><tr><td>1 Relative Advantage (RA)</td><td char=".">0.119</td><td char=".">−0.01</td><td char=".">0.065</td><td char=".">0.010</td><td char=".">0.214</td><td char=".">0.023</td><td char=".">0.025</td><td char=".">0.010</td><td char=".">0.070</td><td char=".">0.153</td><td char=".">0.030</td><td char=".">0.043</td><td char=".">−0.01</td></tr><tr><td /><td char=".">15.30</td><td char=".">−1.67</td><td char=".">12.49</td><td char=".">2.16</td><td char=".">20.49</td><td char=".">10.24</td><td char=".">4.30</td><td char=".">6.15</td><td char=".">10.36</td><td char=".">19.05</td><td char=".">4.57</td><td char=".">12.33</td><td char=".">−1.11</td></tr><tr><td>2 Department Value Added</td><td char=".">0.004</td><td char=".">0.007</td><td char=".">0.004</td><td char=".">0.002</td><td char=".">0.008</td><td char=".">−0.00</td><td char=".">0.007</td><td char=".">−0.00</td><td char=".">0.007</td><td char=".">0.005</td><td char=".">−0.08</td><td char=".">0.001</td><td char=".">0.007</td></tr><tr><td /><td char=".">0.79</td><td char=".">1.54</td><td char=".">1.57</td><td char=".">0.78</td><td char=".">1.44</td><td char=".">−1.67</td><td char=".">1.91</td><td char=".">−0.60</td><td char=".">1.57</td><td char=".">1.35</td><td char=".">−1.77</td><td char=".">0.46</td><td char=".">1.61</td></tr><tr><td>3 Average GCSE</td><td char=".">0.017</td><td char=".">−0.03</td><td char=".">0.034</td><td char=".">0.017</td><td char=".">0.181</td><td char=".">0.005</td><td char=".">0.044</td><td char=".">0.008</td><td char=".">0.070</td><td char=".">0.073</td><td char=".">0.001</td><td char=".">0.017</td><td char=".">0.000</td></tr><tr><td /><td char=".">4.02</td><td char=".">−5.14</td><td char=".">12.98</td><td char=".">7.91</td><td char=".">27.53</td><td char=".">7.12</td><td char=".">12.47</td><td char=".">7.41</td><td char=".">16.16</td><td char=".">18.59</td><td char=".">0.26</td><td char=".">9.88</td><td char=".">0.04</td></tr><tr><td>4 Male</td><td char=".">−0.08</td><td char=".">0.064</td><td char=".">0.001</td><td char=".">0.050</td><td char=".">−0.16</td><td char=".">−0.01</td><td char=".">0.032</td><td char=".">0.00</td><td char=".">−0.04</td><td char=".">0.069</td><td char=".">0.027</td><td char=".">0.038</td><td char=".">−0.16</td></tr><tr><td /><td char=".">−12.5</td><td char=".">10.20</td><td char=".">0.22</td><td char=".">10.54</td><td char=".">−19.7</td><td char=".">−5.29</td><td char=".">6.29</td><td char=".">−2.03</td><td char=".">−6.07</td><td char=".">11.37</td><td char=".">4.39</td><td char=".">10.01</td><td char=".">−15.1</td></tr><tr><td>5 Mother Professional</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">0.005</td><td char=".">0.015</td><td char=".">−0.00</td><td char=".">0.00</td><td char=".">0.002</td><td char=".">0.010</td><td char=".">−0.02</td><td char=".">0.009</td><td char=".">−0.00</td><td char=".">0.003</td></tr><tr><td /><td char=".">−0.27</td><td char=".">−1.12</td><td char=".">−0.80</td><td char=".">0.77</td><td char=".">1.01</td><td char=".">−0.36</td><td char=".">0.15</td><td char=".">0.98</td><td char=".">0.91</td><td char=".">−1.63</td><td char=".">0.87</td><td char=".">−0.72</td><td char=".">0.23</td></tr><tr><td>6 Father Professional</td><td char=".">−0.01</td><td char=".">0.022</td><td char=".">−0.02</td><td char=".">−0.00</td><td char=".">0.012</td><td char=".">0.007</td><td char=".">−0.00</td><td char=".">0.00</td><td char=".">0.013</td><td char=".">−0.01</td><td char=".">0.002</td><td char=".">−0.01</td><td char=".">0.006</td></tr><tr><td /><td char=".">−1.25</td><td char=".">2.48</td><td char=".">−4.03</td><td char=".">−0.32</td><td char=".">1.01</td><td char=".">1.96</td><td char=".">−0.30</td><td char=".">0.16</td><td char=".">1.56</td><td char=".">−1.20</td><td char=".">0.30</td><td char=".">−1.31</td><td char=".">0.62</td></tr><tr><td>7 Aspiration Professional</td><td char=".">0.010</td><td char=".">0.056</td><td char=".">0.032</td><td char=".">0.024</td><td char=".">−0.02</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">0.00</td><td char=".">0.016</td><td char=".">0.039</td><td char=".">−0.02</td><td char=".">0.007</td><td char=".">−0.00</td></tr><tr><td /><td char=".">1.45</td><td char=".">7.30</td><td char=".">5.76)</td><td char=".">5.28</td><td char=".">−1.55</td><td char=".">−0.66</td><td char=".">−2.04</td><td char=".">−0.36</td><td char=".">2.25</td><td char=".">5.29</td><td char=".">−3.62</td><td char=".">1.74</td><td char=".">−0.45</td></tr><tr><td>8 Mother managerial</td><td char=".">−0.01</td><td char=".">−0.04</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.019</td><td char=".">0.001</td><td char=".">0.004</td><td char=".">0.001</td><td char=".">0.011</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">0.016</td></tr><tr><td /><td char=".">−1.23</td><td char=".">−4.83</td><td char=".">−2.22</td><td char=".">−0.61</td><td char=".">1.78</td><td char=".">0.40</td><td char=".">0.63</td><td char=".">0.74</td><td char=".">1.48</td><td char=".">−1.40</td><td char=".">−0.03</td><td char=".">−2.04</td><td char=".">1.83</td></tr><tr><td>9 Father Managerial</td><td char=".">0.010</td><td char=".">0.008</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.018</td><td char=".">0.003</td><td char=".">0.004</td><td char=".">0.002</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.001</td><td char=".">0.002</td><td char=".">0.005</td></tr><tr><td /><td char=".">1.40</td><td char=".">1.19</td><td char=".">−1.90</td><td char=".">−0.99</td><td char=".">1.84</td><td char=".">1.46</td><td char=".">0.72</td><td char=".">1.27</td><td char=".">−0.64</td><td char=".">−0.31</td><td char=".">0.21</td><td char=".">0.54</td><td char=".">0.63</td></tr><tr><td>10 Aspiration Managerial</td><td char=".">−0.03</td><td char=".">0.014</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">0.028</td><td char=".">0.002</td><td char=".">0.045</td><td char=".">0.001</td><td char=".">0.002</td><td char=".">0.031</td><td char=".">−0.04</td><td char=".">0.028</td><td char=".">−0.01</td></tr><tr><td /><td char=".">−4.20</td><td char=".">1.61</td><td char=".">−1.15</td><td char=".">−1.43</td><td char=".">2.41</td><td char=".">0.69</td><td char=".">6.06</td><td char=".">0.43</td><td char=".">0.28</td><td char=".">3.37</td><td char=".">−6.14</td><td char=".">4.61</td><td char=".">−1.55</td></tr><tr><td>11 Mother with Degree</td><td char=".">0.012</td><td char=".">−0.03</td><td char=".">0.009</td><td char=".">−0.01</td><td char=".">0.002</td><td char=".">0.001</td><td char=".">0.012</td><td char=".">0.0</td><td char=".">0.005</td><td char=".">−0.01</td><td char=".">0.011</td><td char=".">−0.01</td><td char=".">0.005</td></tr><tr><td /><td char=".">1.32</td><td char=".">−4.3</td><td char=".">1.46</td><td char=".">−2.4</td><td char=".">(0.14)</td><td char=".">0.32</td><td char=".">1.67</td><td char=".">0.29</td><td char=".">0.61</td><td char=".">−0.7</td><td char=".">1.34</td><td char=".">−3.1</td><td char=".">0.49</td></tr><tr><td>12 Father with Degree</td><td char=".">0.000</td><td char=".">0.002</td><td char=".">0.005</td><td char=".">0.014</td><td char=".">0.010</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">0.001</td><td char=".">0.011</td><td char=".">−0.01</td><td char=".">0.001</td><td char=".">0.014</td><td char=".">−0.01</td></tr><tr><td /><td char=".">0.01</td><td char=".">0.19</td><td char=".">0.91</td><td char=".">2.81</td><td char=".">0.95</td><td char=".">−1.33</td><td char=".">−1.16</td><td char=".">1.06</td><td char=".">1.47</td><td char=".">−1.46</td><td char=".">0.20</td><td char=".">3.01</td><td char=".">−1.19</td></tr><tr><td>13 RA X Mother Professional</td><td char=".">−0.00</td><td char=".">0.014</td><td char=".">−0.00</td><td char=".">0.007</td><td char=".">−0.02</td><td char=".">0.00</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">−0.01</td><td char=".">0.003</td><td char=".">0.009</td><td char=".">0.001</td><td char=".">−0.01</td></tr><tr><td /><td char=".">−0.20</td><td char=".">1.35</td><td char=".">−0.58</td><td char=".">1.22</td><td char=".">−1.01</td><td char=".">−0.50</td><td char=".">−0.64</td><td char=".">−1.28</td><td char=".">−0.98</td><td char=".">0.24</td><td char=".">1.06</td><td char=".">0.22</td><td char=".">−0.41</td></tr><tr><td>14 RA X Father Professional</td><td char=".">−0.01</td><td char=".">0.005</td><td char=".">0.009</td><td char=".">0.012</td><td char=".">−0.01</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.002</td><td char=".">0.002</td><td char=".">0.001</td></tr><tr><td /><td char=".">−0.80</td><td char=".">0.54</td><td char=".">1.73</td><td char=".">2.61</td><td char=".">−0.35</td><td char=".">−0.86</td><td char=".">−0.21</td><td char=".">−0.60</td><td char=".">−0.03</td><td char=".">−0.45</td><td char=".">0.26</td><td char=".">0.69</td><td char=".">0.11</td></tr><tr><td>15 RA X Aspiration Professional</td><td char=".">0.005</td><td char=".">0.011</td><td char=".">−0.01</td><td char=".">0.00</td><td char=".">0.034</td><td char=".">0.001</td><td char=".">0.011</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">−0.02</td><td char=".">0.016</td><td char=".">−0.01</td><td char=".">0.022</td></tr><tr><td /><td char=".">0.75</td><td char=".">1.45</td><td char=".">−2.73</td><td char=".">−1.12</td><td char=".">3.27</td><td char=".">0.66</td><td char=".">1.81</td><td char=".">0.41</td><td char=".">0.01</td><td char=".">−2.26</td><td char=".">2.58</td><td char=".">−2.27</td><td char=".">2.60</td></tr><tr><td>16 RA X Aspiration Managerial</td><td char=".">−0.02</td><td char=".">0.00</td><td char=".">−0.01</td><td char=".">0.00</td><td char=".">0.043</td><td char=".">0.00</td><td char=".">0.007</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.00</td><td char=".">0.02</td><td char=".">0.001</td><td char=".">0.031</td></tr><tr><td /><td char=".">−2.34</td><td char=".">−0.20</td><td char=".">−1.63</td><td char=".">−0.37</td><td char=".">(3.58)</td><td char=".">−0.03</td><td char=".">1.19</td><td char=".">0.20</td><td char=".">−0.15</td><td char=".">−0.43</td><td char=".">2.68</td><td char=".">0.37</td><td char=".">3.31</td></tr><tr><td>17 RA X Mother with Degree</td><td char=".">0.002</td><td char=".">0.004</td><td char=".">−0.00</td><td char=".">0.002</td><td char=".">−0.01</td><td char=".">0.000</td><td char=".">0.004</td><td char=".">0.000</td><td char=".">−0.01</td><td char=".">−0.00</td><td char=".">0.005</td><td char=".">0.009</td><td char=".">0.00</td></tr><tr><td /><td char=".">0.20</td><td char=".">0.48</td><td char=".">−0.83</td><td char=".">0.50</td><td char=".">−0.70</td><td char=".">0.01</td><td char=".">0.56</td><td char=".">0.07</td><td char=".">−0.92</td><td char=".">−0.29</td><td char=".">0.59</td><td char=".">2.34</td><td char=".">−0.29</td></tr><tr><td>18 RA X Father with Degree</td><td char=".">−0.01</td><td char=".">0.022</td><td char=".">0.00</td><td char=".">−0.01</td><td char=".">0.030</td><td char=".">0.003</td><td char=".">0.009</td><td char=".">0.00</td><td char=".">0.014</td><td char=".">0.018</td><td char=".">0.005</td><td char=".">0.00</td><td char=".">0.021</td></tr><tr><td /><td char=".">−0.95</td><td char=".">2.88</td><td char=".">−0.21</td><td char=".">−1.76</td><td char=".">2.49</td><td char=".">1.35</td><td char=".">1.35</td><td char=".">−1.16</td><td char=".">1.97</td><td char=".">2.25</td><td char=".">0.65</td><td char=".">−0.89</td><td char=".">2.21</td></tr><tr><td>19 RA X Gender</td><td char=".">−0.02</td><td char=".">0.017</td><td char=".">−0.01</td><td char=".">−0.01</td><td char=".">0.005</td><td char=".">0.003</td><td char=".">0.020</td><td char=".">0.002</td><td char=".">0.018</td><td char=".">−0.03</td><td char=".">0.020</td><td char=".">0.00</td><td char=".">0.015</td></tr><tr><td /><td char=".">−3.62</td><td char=".">2.80</td><td char=".">−2.95</td><td char=".">−1.77</td><td char=".">0.59</td><td char=".">1.90</td><td char=".">4.18</td><td char=".">1.84)</td><td char=".">3.14</td><td char=".">−4.69</td><td char=".">3.59</td><td char=".">−0.87</td><td char=".">1.64</td></tr><tr><td>20 Students (n)</td><td char=".">16235</td><td char=".">15951</td><td char=".">15259</td><td char=".">16501</td><td char=".">15948</td><td char=".">16190</td><td char=".">14790</td><td char=".">16169</td><td char=".">16286</td><td char=".">13876</td><td char=".">13580</td><td char=".">16007</td><td char=".">16083</td></tr><tr><td>21 Rho</td><td char=".">0.043</td><td char=".">0.027</td><td char=".">0.033</td><td char=".">0.106</td><td char=".">0.050</td><td char=".">0.000</td><td char=".">0.027</td><td char=".">0.068</td><td char=".">0.025</td><td char=".">0.065</td><td char=".">0.047</td><td char=".">0.018</td><td char=".">0.033</td></tr><tr><td>22 Log Likelihood (−)</td><td char=".">−20252</td><td char=".">−12388</td><td char=".">−15186</td><td char=".">−9969</td><td char=".">−21890</td><td char=".">−9862</td><td char=".">−19676</td><td char=".">−4886</td><td char=".">−20066</td><td char=".">−15065</td><td char=".">−4157</td><td char=".">−12557</td><td char=".">−9755</td></tr></tbody></table> </ephtml> </p> <p>The probability that a particular student will enter for examination in a subject can be estimated by entering the linear prediction into the logitistic cdf:</p> <p>Graph</p> <p>To provide some indication of the effect sizes we now present estimates for some selected cases. We do this by showing the effect of variation in one variable whilst holding the others constant: continuous variables are held constant at the mean unless indicated otherwise and dummy variables are given the value 0. The social background effects are either insignificant or very small, so changing the parental dummy variables from 0 to 1 has little effect on the probability estimates. The effect of changing the gender dummy is substantial, as expected. A female student in a school with an average relative advantage in biology has a probability of 0.20 of entering for examination in biology. This rises to 0.38 if their relative advantage in biology is one standard deviation above the mean. For a male the estimated probability of taking biology drops dramatically to 0.11 at the mean and 0.21 when the individual has a comparative advantage of one standard deviation above the mean. The figure in second row for each variable in Tables 3 and 4 shows the statistical significance in terms of the t statistic, which is a test of whether the coefficient is significantly different from zero.</p> <hd id="AN0037265477-7">Students' gender, background, prior achievements and aspirations</hd> <p>A student in a school (Table 3) is more likely to study any of the ten more established subjects (biology, chemistry, economics, English, French, geography, German, history, mathematics and physics) if they have a higher average GCSE grade. This effect (Tables 3 and 4, Row 3) is particularly strong for mathematics, English and history. More recent additions to the sixth form curriculum in schools (business studies, media studies and psychology) are more likely to attract students with lower performance at GCSE. This pattern is less strong in sixth‐form colleges. For media studies and psychology the effect is insignificant, but the negative relationship remains significant for business studies.</p> <p>The gender differences are as expected on the basis of national figures: a female is more likely than a male to study biology, English, French, history and psychology, and less likely to study business studies, economics, mathematics and physics. Rows 5 to 18 show the effects of home background and aspirations. These effects are best viewed collectively for each subject, although the effects of these variables frequently operate in opposite directions, particularly when the interactions are considered. Most of these marginal effects are very small, especially when compared with the coefficients for students' relative advantage. The only subject that does not conform to this pattern is business studies, particularly in schools. Overall the data offer little support for the proposition that freedom of choice of subject at A level within an institution reinforces socio‐economic differences.</p> <hd id="AN0037265477-8">The effect of students' relative advantage in a subject</hd> <p>Students are more likely to study a subject if it is associated with their academic strengths. This is visible in the 'Student's relative advantage' (Row 1 in Tables 3 and 4). In most of the subjects which students are most likely to have studied before (biology, chemistry, English, French, German, history, mathematics and physics) the effect of a one standard deviation increase in relative advantage is greater than that associated with being male. The only exception is geography. In all but two of these subjects (geography and history), the effect of a one standard deviation rise in relative advantage is larger than a one grade increase in average GCSE grade. If the model is run without the inclusion of the relative advantage variable much of its effect is picked up by average GCSE and gender, generating a misleading impression of the effect of these two variables. The smallest effects are found in those subjects (business studies, media studies, economics and psychology) which students are less likely to have studied before. The relationship is significantly negative for business studies. It is possible that some of this difference between 'old' and 'newer' subjects could be accounted for by differences in the relative difficulty of these subjects for all students. There is evidence (CEM Centre, [<reflink idref="bib5" id="ref45">5</reflink>]), for example, that, on average, students with the same GCSE scores attain lower grades in mathematics, French and physics than in psychology, business studies and media studies. However, English is rated easier than average in these measures, yet secures the highest effect in terms of relative advantage. Economics is rated more difficult than average yet it has a somewhat low relative advantage score.</p> <p>For the traditional subjects these relative advantage effects are stronger in schools than in sixth‐form colleges, whilst the effects are stronger for business studies, media studies and psychology in sixth‐form colleges than in schools. This could indicate the effect of school teachers' knowledge and guidance or the effect of some students attending sixth‐form colleges wanting a change from their previous curriculum experience even when they have been relatively successful at it. This evidence of the effect of relative advantage suggests that students, on average, do develop a fairly accurate assessment of their relative strengths and that they choose the subjects at A Level in which they are most likely to do well. This effect is stronger in schools than in colleges. The hypothesis that the coefficients for sixth forms and schools are not significantly different can be tested using a likelihood ratio test of the restricted (pooled) likelihood against the two unrestricted likelihoods. This hypothesis cannot be rejected for German, but it can be rejected at the 5% level for French, media studies and history and at the 0.1% level for all other subjects.</p> <hd id="AN0037265477-9">The effect of relative departmental performance</hd> <p>The rho statistics in rows 21 of Tables 3 and 4 show the proportion of the unexplained variance at the institution level: the likelihood of studying a particular subject varies with the institution which a student attends and the results are highly significant. These effects are similar in scale to school effects on attainment (Nye <emph>et al</emph>., 2004). Typically, these subject choice effects are larger for schools than for sixth‐form colleges. These effects might be generated by school policies and the socio‐economic composition of the institution's intake.</p> <p>In contrast, observed departmental effectiveness is only significant for three out of the 13 subjects in schools. There are no significant effects in sixth‐form colleges. Even in those instances where departmental effectiveness is significant the marginal effect is negligible when compared with the effect of a one standard deviation change in students' relative advantage. However, there are a number of reasons for being cautious in interpreting this result. First, suppose that there is a correlation between departmental performance in schools at GCSE and at A Level. In this case our measure of students' relative advantage will incorporate an effect of department effectiveness. This provides another possible explanation for the relative advantage effect being lower in sixth‐form colleges than in schools with sixth forms. Second, if value added to the student reduces as class size increases then there would be a tendency towards equalising value added across subjects. If a department proved itself relatively more effective it would attract more students, but this would subsequently reduce its relative effectiveness. Third, there are difficulties in finding an appropriate estimate for departmental effectiveness in the multi‐dimensional choice framework that exists for A Level. For each of these reasons it is possible that there are departmental effects that we are not picking up in our analysis.</p> <hd id="AN0037265477-10">Conclusions</hd> <p>For most subjects that students have studied before entering the sixth form, our results suggest that a student's relative advantage in a subject is a more powerful predictor of the likelihood that they will enter for examination in an A‐Level subject than either gender or the student's average prior achievement across subjects. This contrasts with previous studies (such as Uerz <emph>et al</emph>., [<reflink idref="bib35" id="ref46">35</reflink>]) which conclude that gender is the key factor. The inclusion of students' average GCSE grade in our model reduces the coefficient on our gender variable, since on average females get higher average GCSE grades than males in England. However, Uerz <emph>et al</emph>. include a similar variable in their study. Moreover our calculation of students' relative advantage incorporates gender differences in relative achievement at GCSE. The gender effect is less than the relative advantage effect in most of the established school subjects (biology, chemistry, English, French, German, history, mathematics and physics). However, the gender effect is greater than relative advantage for geography and for subjects that students are less likely to have studied before.</p> <p>Measures of students' relative advantage, such as the one used in this study, are imprecise. However, the scale of the coefficients and the robustness of the results to different configurations of the model, suggest several implications. First, a move from a type of open choice system, as presently exists in England, to a baccalaureate system which significantly reduced the scope for specialisation would be likely to reduce students' average attainment. This follows because our calculations of students' relative advantage indicate that students' achievements at GCSE are significant predictors of difference between the grades they are likely to obtain in different subjects. Our data also show that where students and schools have evidence of prior achievement in a subject this exerts a strong effect on choices for further study. It appears that they do make well informed, perhaps, 'rational' decisions. Moreover, the lack of evidence (Dolton & Vignoles, [<reflink idref="bib11" id="ref47">11</reflink>]; Johnes, [<reflink idref="bib22" id="ref48">22</reflink>]) that students following a broad curriculum receive a wage premium in the labour market does not support an argument that a broad curriculum has employment benefits which should be taken into account beyond effects on average examination performance.</p> <p>Second, our data present consistent evidence that better decisions are made when students have previous experience of studying a subject as anticipated by Manski ([<reflink idref="bib25" id="ref49">25</reflink>]). In the case of business studies there is actually a small negative relationship between relative advantage and the likelihood of entering for examination in the subject. A curriculum that offers increasing choice as students mature will always face this kind of problem, but it does suggest that it is valuable to give students some experience of a new subject area before they are fully committed to it. The extent to which this opportunity has been incorporated within the 16–19 curriculum in England since 2000 does not appear to have made a significant impact on the problem.</p> <p>Third, there is a small tendency for students to be less inclined to be influenced by their relative advantage when switching to study in a sixth‐form college than when they continue to study at a school with a sixth form. Some students choose to study at a sixth‐form college as an alternative to continuing in the same school. These students may wish to abandon subjects that they associate with an educational pathway they are rejecting. However, the data do provide some grounds for thinking that guidance in sixth‐form colleges might be strengthened to encourage students to reflect more carefully on their relative advantage when making their choice of subjects.</p> <p>Our evidence provides little support for an expectation that more effective departments will attract more students. We tried a variety of approaches to identifying an effect of competition between departments but none of these produced consistent and significant results. The only result that was consistent in all our investigations was that relatively more effective biology departments in schools tend to attract more A‐Level students. This does not conclusively demonstrate that there are no effects in other subjects, for reasons we have discussed earlier. However, it does suggest that subject choice, at least at A Level, impacts on school effectiveness through the relative advantage of students more powerfully than the relative effectiveness of departments.</p> <p>Nevertheless, schools do make a difference. Our estimates of the proportion of the variance in the likelihood of a student entering for examination in a subject that is attributable to the institution are comparable in size to estimates of school effects on student attainment. One possible explanation for this is that our measure of departmental effectiveness does not adequately capture the ways in which students and school managers weigh up the qualities of different departments. Another possibility is that criteria used to grant access to courses vary across institutions. These are matters for further research. Perhaps the most striking school effect is seen in the subjects that are relatively new to the curriculum. Psychology and media studies are offered by less than half of all schools, whilst business studies is offered by just over half of all schools. Economics is offered by only one school in seven. However, in those institutions that offer students an opportunity to study these subjects their popularity is similar to most of the more traditional subjects. It seems therefore that specialisation by schools can restrict, rather than enhance, the opportunities open to students.</p> <hd id="AN0037265477-11">Notes on contributors</hd> <p>Nick Adnett is Professor of Economics at Staffordshire University. His work on the economics of education concentrates on competition and collaboration in schooling markets and student finance in Higher Education.</p> <p>Robert Coe is Director of Secondary Projects at the Curriculum Evaluation and Management Centre, Durham University. He is editor with A.J. Visscher of <emph>School Improvement through Performance Feedback</emph>, published by Swets and Zeitlinger.</p> <p>Neil Davies is Research Assistant at the Centre for Market and Public Organisation, University of Bristol.</p> <p>Peter Davies is Professor of Education Policy and Director of the Institute for Education Policy Research at Staffordshire University. He is author, with Nick Adnett, of <emph>Markets for Schooling: an economic analysis</emph>, published by Routledge, and editor of the <emph>International Review of Economics Education</emph>.</p> <p>David Hutton is Research Associate at the Curriculum Evaluation and Management Centre, Durham University.</p> <hd id="AN0037265477-12">Acknowledgements</hd> <p>The paper is an outcome of the ESRC Project RES‐000‐22‐0090 'Within School Competition and Pupil Achievement'. We are grateful for helpful comments on earlier drafts received from Ray Bachan, Karen Hancock, George Leckie, Hans Luyten, Jean Mangan, Ulrich Trautwein, Helen Watt, Frank Windmeijer and two anonymous referees.</p> <hd id="AN0037265477-13">Appendix</hd> <p>A simultaneous version of the model was estimated without school level effects on a reduced sample of 994 individuals. This model is based on the Geweke‐Hajivassiliou‐Keane estimator for a multi‐variate normal (Hajivassiliou <emph>et al</emph>., [<reflink idref="bib17" id="ref50">17</reflink>]). In STATA this is the mvprobit command coded by Cappellari and Jenkins ([<reflink idref="bib4" id="ref51">4</reflink>]).</p> <p>The model can be expressed as a M=13 equation multivariate probit model, the school level variances are removed. The 13 equations are jointly estimated using a multivariate normal cdf. The simultaneity occurs through the error term; it allows for the error terms to be correlated across the different choices:</p> <p>Graph</p> <p>Where, <emph>u<subs>kit</subs></emph>, <emph>k</emph> = 1,...,13 are error terms distributed as a multivariate normal each with a mean of zero and a variance‐covariance matrix V, where V has values of 1 on the leading diagonal and correlations <emph>ρ<subs>jk</subs></emph> = <emph>ρ<subs>kj</subs></emph> on the off diagonal elements.</p> <p>The resulting variance covariance matrix:</p> <p>Graph</p> <p>estimating each equation independently is a restriction on this model of <emph>ρ<subs>jk</subs></emph> = 0∀<emph>ij</emph>, <emph>i</emph> ≠ <emph>j</emph>.</p> <p>This model can be estimated using a log‐likelihood function based on the multivariate normal cdf;</p> <p>Graph</p> <p>Where Φ<subs>13</subs> (<emph>µ</emph>i, Ω) is a multivariate normal cdf.</p> <p>This model was estimated using STATA on a random 1% sample of the original dataset. This model cannot be directly compared to the original regressions as it is a restricted sample and it does not contain school effects. However, all of the covariance terms were as expected. For example, biology is positively correlated with chemistry and physics is positive correlated with maths. Examining the independence restriction <emph>ρ<subs>jk</subs></emph> = 0∀<emph>ij</emph>, <emph>i</emph> ≠ <emph>j</emph> using a likelihood ratio test showed that hypothesis of independence is strongly rejected: chi(<reflink idref="bib156" id="ref52">156</reflink>)=469, and p<0.01.</p> <p>However, the coefficients are very similar to the simple model and the previous significant effects remain significant. This finding is particularly secure in the case of our measure of comparative advantage. None of our conclusions based on the independent equations would differ. This model does not allow us to control for school level heterogeneity.</p> <ref id="AN0037265477-14"> <title> References </title> <blist> <bibl id="bib1" idref="ref26" type="bt">1</bibl> <bibtext> Adnett, N. and Davies, P.2005. Competition between or within schools? Re‐assessing school choice. Education Economics, 13: 109–121.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref36" type="bt">2</bibl> <bibtext> Audit Commission. 1996. Trading places: the supply and allocation of school places, London: The Audit Commission.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref27" type="bt">3</bibl> <bibtext> Bell, J.F., Malacova, E. and Shannon, M.2005. The changing pattern of A Level/AS uptake in England. The Curriculum Journal, 16: 391–400.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref51" type="bt">4</bibl> <bibtext> Cappellari, L. and Jenkins, S.P.2003. MVPROBIT: Stata module to calculate multivariate probit regression using simulated maximum likelihoodStatistical Software Components S432601, Boston College Department of Economics, revised 25 January 2006. Available online at: <ulink href="http://ideas.repec.org/c/boc/bocode/s432601.html">http://ideas.repec.org/c/boc/bocode/s432601.html</ulink> (accessed 18 November 2008)</bibtext> </blist> <blist> <bibl id="bib5" idref="ref45" type="bt">5</bibl> <bibtext> CEM (Curriculum, Evaluation and Management Centre). 2007. A Level subject difficulties, Durham: Curriculum, Evaluation and Management Centre, University of Durham. Available online at <ulink href="http://www.alisproject.org/Documents/Alis/Research/ALevel%20Subject%20Difficulties.pdf">http://www.alisproject.org/Documents/Alis/Research/ALevel%20Subject%20Difficulties.pdf</ulink> (accessed 12 September 2007)</bibtext> </blist> <blist> <bibl id="bib6" idref="ref23" type="bt">6</bibl> <bibtext> Cleaves, A.2005. The formation of science choices in secondary school. International Journal of Science Education, 27: 471–486.</bibtext> </blist> <blist> <bibl id="bib7" type="bt">7</bibl> <bibtext> Davies, P., Telhaj, S., Hutton, D., Adnett, N. and Coe, R.forthcoming. Competition, cream skimming and department performance within secondary schools. British Educational Research Journal,</bibtext> </blist> <blist> <bibl id="bib8" idref="ref10" type="bt">8</bibl> <bibtext> Department for Education and Skills. 2001. Schools achieving success, London: HMSO.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref2" type="bt">9</bibl> <bibtext> Department for Education and Skills. 2005a. 14–19 Education and skills, London: HMSO. Cm. 6746</bibtext> </blist> <blist> <bibtext> Department for Education and Skills. 2005b. Youth cohort study: the activities and experiences of 17‐year‐olds: England and Wales, London: HMSO.</bibtext> </blist> <blist> <bibtext> Dolton, P. J. and Vignoles, A.2002. Is a broader curriculum better?. Economics of Education Review, 21: 415–429.</bibtext> </blist> <blist> <bibtext> Eccles, J. S. and Wigfield, A.2002. Motivational beliefs, values, and goals. Annual Review of Psychology, 53: 109–132.</bibtext> </blist> <blist> <bibtext> Elsworth, G.R., Harvey‐Beavis, A., Ainley, J. and Fabris, S.1999. Generic interests and school subject choice. Educational Research and Evaluation, 5: 290–318.</bibtext> </blist> <blist> <bibtext> Gamoran, A.1996. Curriculum standardization and equality of opportunity in Scottish secondary education: 1984–90. Sociology of Education, 69: 1–21.</bibtext> </blist> <blist> <bibtext> Gardner, H.1983. Frames of mind, New York: Basic Books.</bibtext> </blist> <blist> <bibtext> Garratt, L.1985. Factors affecting subject choice at A‐Level. Educational Studies, 11(2): 127–132.</bibtext> </blist> <blist> <bibtext> Hajivassiliou, V., McFadden, D. and Ruud, P.1996. Simulation of multivariate normal rectangle probabilities and their derivatives theoretical and computational results. Journal of Econometrics, 72(1–2): 85–134.</bibtext> </blist> <blist> <bibtext> Hodgson, A., Howieson, C., Raffe, D., Spours, K. and Tinkin, T.2004. Post‐16 curriculum and qualifications reform in England and Scotland: lessons from home‐international comparisons. Journal of Education and Work, 17: 441–465.</bibtext> </blist> <blist> <bibtext> House of Commons. 1992. Choice and diversity: a new framework for schools, London: HMSO. White Paper, minutes of evidence from the Department of Education, Cmnd 236i</bibtext> </blist> <blist> <bibtext> Hudson, L.1968. Frames of mind: ability, perception and self‐perception in the arts and sciences, London: Methuen.</bibtext> </blist> <blist> <bibtext> Institute for Welsh Affairs (IWA). 1999. The WelshBac: from Wales to the World, Cardiff: IWA.</bibtext> </blist> <blist> <bibtext> Johnes, G.2003. Curriculum, Lancaster: University of Lancaster. Lancaster University Management School Working Paper 2003/086</bibtext> </blist> <blist> <bibtext> Levačić, R. and Jenkins, A.2006. Evaluating the effectiveness of specialist schools in England. School Effectiveness and School Improvement, 17: 229–254.</bibtext> </blist> <blist> <bibtext> Lumby, J. and Wilson, M.2003. Developing 14–19 education: meeting needs and improving choice. Journal of Education Policy, 18: 533–550.</bibtext> </blist> <blist> <bibtext> Manski, C.1993. "Adolescent econometricians: how do youth infer the returns to schooling?". In Studies of supply and demand in higher education, Edited by: Clotfelter, G. and Rothschild, M.43–57. Chicago: University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Marsh, H. W.1986. Verbal and math self‐concepts: an internal/external frame of reference model. American Educational Research Journal, 23: 129–149.</bibtext> </blist> <blist> <bibtext> Nagy, G., Trautwein, U., Köller, O., Baumert, J. and Garrett, J.2004. Gender and course selection in upper secondary education: effects of academic self‐concept and intrinsic value. Educational Research and Evaluation, 12: 323–345.</bibtext> </blist> <blist> <bibtext> Nye, B., Konstantopoulos, S. and Hedges, L.V.2004. How large are teacher effects?. Educational Evaluation and Policy Analysis, 26: 237–257.</bibtext> </blist> <blist> <bibtext> OfSTED. 2001. Standards and quality in education 1999–2000, London: HMSO.</bibtext> </blist> <blist> <bibtext> OfSTED. 2005. Ofsted praises improving education system but warns of class gap holding back some young people from achieving full potential, London: OfSTED. Available at: <ulink href="http://www.ofsted.gov.uk/news/index.cfm?fuseaction=news.details&id=1660">http://www.ofsted.gov.uk/news/index.cfm?fuseaction=news.details&id=1660</ulink> (accessed 14 October 2005)</bibtext> </blist> <blist> <bibtext> Rivkin, S. G.Hanushek, E. A.2005. Teachers, schools, and academic achievement. Econometrica, 73(2): 417–458.</bibtext> </blist> <blist> <bibtext> Stokking, K.M.2000. Predicting the choice of physics in secondary education. International Journal of Science Education, 22: 1261–1283.</bibtext> </blist> <blist> <bibtext> Trautwein, U., Köller, O., Lüdtke, O. and Baumert, J.2005. "Student tracking and the powerful effects of opt‐in courses on self‐concept". In International advances in self research: volume 2, Edited by: Marsh, H. W., Craven, R. G. and McInerney, D. M.Greenwich, CT: Information Age Press.</bibtext> </blist> <blist> <bibtext> Tymms, P.1992. The relative effectiveness of post‐16 institutions in England (including Assisted Places Scheme schools). British Educational Research Journal, 18: 175–192.</bibtext> </blist> <blist> <bibtext> Uerz, D., Dekkers, H. and Béguin, A.A.2004. Mathematics and language skills and the choice of science subjects in secondary education. Educational Research and Evaluation, 10: 163–182.</bibtext> </blist> </ref> <aug> <p>By Peter Davies; Neil Davies; David Hutton; Nick Adnett and Robert Coe</p> <p>Reported by Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib21" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib30" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib23" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib18" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib14" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib11" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib22" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib19" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib34" firstref="ref11"></nolink> <nolink nlid="nl10" bibid="bib28" firstref="ref12"></nolink> <nolink nlid="nl11" bibid="bib31" firstref="ref13"></nolink> <nolink nlid="nl12" bibid="bib20" firstref="ref14"></nolink> <nolink nlid="nl13" bibid="bib15" firstref="ref15"></nolink> <nolink nlid="nl14" bibid="bib26" firstref="ref16"></nolink> <nolink nlid="nl15" bibid="bib12" firstref="ref17"></nolink> <nolink nlid="nl16" bibid="bib35" firstref="ref18"></nolink> <nolink nlid="nl17" bibid="bib25" firstref="ref19"></nolink> <nolink nlid="nl18" bibid="bib33" firstref="ref20"></nolink> <nolink nlid="nl19" bibid="bib13" firstref="ref21"></nolink> <nolink nlid="nl20" bibid="bib32" firstref="ref22"></nolink> <nolink nlid="nl21" bibid="bib27" firstref="ref24"></nolink> <nolink nlid="nl22" bibid="bib16" firstref="ref28"></nolink> <nolink nlid="nl23" bibid="bib24" firstref="ref30"></nolink> <nolink nlid="nl24" bibid="bib29" firstref="ref42"></nolink> <nolink nlid="nl25" bibid="bib10" firstref="ref43"></nolink> <nolink nlid="nl26" bibid="bib17" firstref="ref50"></nolink> <nolink nlid="nl27" bibid="bib156" firstref="ref52"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ834864
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Choosing 'in' Schools: Locating the Benefits of Specialisation
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Davies%2C+Peter%22">Davies, Peter</searchLink><br /><searchLink fieldCode="AR" term="%22Davies%2C+Neil%22">Davies, Neil</searchLink><br /><searchLink fieldCode="AR" term="%22Hutton%2C+David%22">Hutton, David</searchLink><br /><searchLink fieldCode="AR" term="%22Adnett%2C+Nick%22">Adnett, Nick</searchLink><br /><searchLink fieldCode="AR" term="%22Coe%2C+Robert%22">Coe, Robert</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Oxford+Review+of+Education%22"><i>Oxford Review of Education</i></searchLink>. Apr 2009 35(2):147-167.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 325 Chestnut Street Suite 800, Philadelphia, PA 19106. Tel: 800-354-1420; Fax: 215-625-2940; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 21
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2009
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22High+Schools%22">High Schools</searchLink><br /><searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Educational+Objectives%22">Educational Objectives</searchLink><br /><searchLink fieldCode="DE" term="%22Outcomes+of+Education%22">Outcomes of Education</searchLink><br /><searchLink fieldCode="DE" term="%22Academic+Achievement%22">Academic Achievement</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Course+Selection+%28Students%29%22">Course Selection (Students)</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Educational+Objectives%22">Student Educational Objectives</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Policy%22">Educational Policy</searchLink><br /><searchLink fieldCode="DE" term="%22Policy+Analysis%22">Policy Analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Specialization%22">Specialization</searchLink><br /><searchLink fieldCode="DE" term="%22Majors+%28Students%29%22">Majors (Students)</searchLink><br /><searchLink fieldCode="DE" term="%22Pretests+Posttests%22">Pretests Posttests</searchLink><br /><searchLink fieldCode="DE" term="%22Alignment+%28Education%29%22">Alignment (Education)</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22United+Kingdom+%28England%29%22">United Kingdom (England)</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/03054980802643298
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0305-4985
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Recent policy in England has suggested that educational outcomes will be raised if schools specialise in particular subjects. In contrast, calls for the reform of 16-19 education have suggested that these outcomes will be improved if students become less specialised in their studies. At present, there is a limited evidence base from which to judge these arguments. In particular, we do not know the extent to which students' achievements in 16-19 education are higher when they choose subjects which play to their perceived strengths. We also do not know whether students are more likely to choose to study subjects taught by more effective departments. That is, outcomes may be affected by the relative strengths of students or departments in circumstances where there is freedom to choose. In this paper we provide evidence of the existence and strength of these relationships. This evidence suggests that reducing the scope within schools for specialisation or competition will reduce average student attainment and these effects ought to be taken into account when evaluating alternative curriculum policies. (Contains 4 tables.)
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: Ref
  Label: Number of References
  Group: RefInfo
  Data: 35
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2009
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ834864
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ834864
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/03054980802643298
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 21
        StartPage: 147
    Subjects:
      – SubjectFull: Educational Objectives
        Type: general
      – SubjectFull: Outcomes of Education
        Type: general
      – SubjectFull: Academic Achievement
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Course Selection (Students)
        Type: general
      – SubjectFull: Student Educational Objectives
        Type: general
      – SubjectFull: Educational Policy
        Type: general
      – SubjectFull: Policy Analysis
        Type: general
      – SubjectFull: Specialization
        Type: general
      – SubjectFull: Majors (Students)
        Type: general
      – SubjectFull: Pretests Posttests
        Type: general
      – SubjectFull: Alignment (Education)
        Type: general
      – SubjectFull: United Kingdom (England)
        Type: general
    Titles:
      – TitleFull: Choosing 'in' Schools: Locating the Benefits of Specialisation
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Davies, Peter
      – PersonEntity:
          Name:
            NameFull: Davies, Neil
      – PersonEntity:
          Name:
            NameFull: Hutton, David
      – PersonEntity:
          Name:
            NameFull: Adnett, Nick
      – PersonEntity:
          Name:
            NameFull: Coe, Robert
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 04
              Type: published
              Y: 2009
          Identifiers:
            – Type: issn-print
              Value: 0305-4985
          Numbering:
            – Type: volume
              Value: 35
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Oxford Review of Education
              Type: main
ResultId 1