ethnic and gender difference in educational achievements

Upload: aino-vehmasto

Post on 07-Apr-2018

223 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    1/16

    Short report

    Ethnic and gender differences in educational achievement andimplications for school improvement strategies

    Feyisa Demie

    Head of Research and Statistics, London Borough of Lambeth Education

    Directorate, International House, Canterbury Crescent, London SW9 7QR

    Summary

    A vital element in school improvement is raising the levels of achievement of

    under-performing groups of pupils in schools. This short report examines the

    extent and reasons for underachievement throughout the key stage 1 and 2 and

    GCSE school years in inner city local education authorities. Statistical trends and

    patterns of performance are analysed, by gender and ethnic factors, to illustrate

    differences in attainment. This is followed by a discussion of the implications of

    the empirical evidence for school improvement. Overall, the message from our

    study is that there are no simple explanations for ethnic differences in educationalachievement. Ethnic heritage does not presuppose underachievement. Some

    ethnic minority groups, such as Indian and Chinese, have levels of attainment

    above the average of the white UK groups. Others such as Caribbean, Portuguese

    and African are under-performing. Additionally, ethnic minority achievements

    are differentiated by both gender and level of uency in English. The ndings of

    this study suggest that pupils in the early stages of English uency perform at

    very low levels, while bilingual pupils who are reasonably procient in English

    perform better, on average, than English-only speakers at KS1, KS2 and GCSE.Whatever the pupils ethnic origin, girls tend to perform at higher levels than boys

    at all key stages. Several other factors are likely to have an inuence on perform-

    ance at the end of each stage of the National Curriculum. This study also raises

    wider questions about how contextual analyses can be used for school self-evalu-

    ation.

    Keywords: ethnicity, gender, school improvement, educational achievement, key

    stage performance

    Educational Research Vol. 43 No. 1 Spring 2001 91106

    Acknowledgements: The views expressed in this short report are those of the author

    and do not necessarily represent those of the London Borough of Lambeth. The author

    would like to acknowledge helpful comments and assistance on the draft copy from

    Rebecca Butler and Anne Taplin.This report is based on a longer paper that was presentedat the Annual Conference of the British Educational Research Association, University of

    Sussex, 4 September 1999.

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    2/16

    Introduction

    Ethnic-related differences in achievement have generated much attention and

    debate since the 1960s, and from the 1970s on, a modest but steady stream of

    reports have been published. The focus of policy-makers concern, although

    earlier concentrating exclusively on black Caribbean pupils, has lately includedthe performance of all ethnic groups in the UK. The most comprehensive inu-

    ential policy studies and inquiry were undertaken by the Rampton Committee

    (1981); Swann Committee (1985); and the ILEA Longitudinal Literacy Survey

    (Mabey, 1981); and the ILEA Junior School Study: School Matters (Mortimore

    et al., 1988). Each of these appeared to show considerable under-achievement of

    Caribbean and other black pupils, when compared with the average levels of

    achievement of white and Asian children.

    Some of the ndings in these reports are supported by recent studies. Severalstudies have found that ethnic background factors affect educational outcomes

    and penalize schools with a high number of pupils from ethnic minorities. Pupils

    from the main ethnic minority groups tend to have a level of attainment below

    the average for that of their white peers (Nuttall et al., 1989; Jesson, Gray and

    Tranmer, 1992; Drew and Gray, 1990; Demie, Reid and Butler, 1997).

    Mortimore et al. (1988) and Drew and Gray (1990) further reveal African and

    Caribbean under-achievement, even after controlling for social class. These

    results are strikingly similar to the ndings of a longitudinal study of pupils in

    multi-ethnic comprehensive schools, which also found that the white pupils

    attained the best average examination achievements when compared to ethnic

    minority pupils (Smith and Tomlinson, 1989).

    Although the above ndings show that ethnic minority pupils tend to have

    lower levels of attainment than white pupils, recent inner London studies suggest

    that some ethnic minorities were performing better than the white pupils

    (Thomas, Pan and Goldstein, 1994; Kendall, 1995, 1998). For example, the

    analysis of the ILEA data by Nuttall et al. (1989) conrms that all ethnic groups

    performed signicantly better than English, Scottish, Welsh and Irish groups.Only the Caribbean pupils achieved less well than the white ethnic group. In con-

    trast, other ethnic minorities such as black African, Indian,Pakistani,Bangladeshi

    and Chinese obtained signicantly higher scores (Thomas et al., 1994; Kendall,

    1995, 1998). One of the possible reasons for improvement by some minority

    groups is that bilingual learners may start school as low attainers in verbal reason-

    ing, but make substantial progress in language skills while attending secondary

    schools (Demie et al., 1997;Thomas et al., 1994). A signicant limitation of both

    the NFER and Institute of Education studies was that the sample size of someof the ethnic groups used in the analysis was too small for a comparative analy-

    sis with larger ethnic groups such as English, Scottish, Welsh, African and

    Caribbean (Demie et al., 1997). Furthermore, the white population of inner

    London is not representative socially or economically with the total white popu-

    lation of the UK (Nuttall et al., 1989).

    More recently, an OFSTED-commissioned research report also reviewed the

    state of recent changes in the educational achievements of ethnic minority pupils.

    The results conrm previous research ndings which suggest considerable under-achievement of Caribbean and other black pupils, on average, compared with

    white and Asian children. In addition, the report concludes that if ethnic diver-

    92 Educational Research Volume 43 Number 1 Spring 2001

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    3/16

    wasted (Gillborn and Gipps, 1996, p. 7). This concern has increased in the wake

    of recent key stages 1 and 2 and GCSE results,which show the underachievement

    of African, Caribbean and white working-class boys in both primary and secon-

    dary schools.

    Various possible explanations were considered for the differences in perform-

    ance between different ethnic groups and underachievement of Caribbean andother black children.A number of researchers have argued that ethnic differences

    in attainment can only properly be understood in relation to social class, gender

    and levels of English uency. For example, Arnot et al. (1998) reveal that, overall,

    white girls are outperforming white boys from the professional and intermediate

    classes, and Asian and African Caribbean boys from these classes are outper-

    forming their female counterparts.

    Previous research has also demonstrated a link between level of uency in

    English and under-achievement (Demie et al., 1997; Kendall et al., 1995, 1998).Children who have English as a second language and are non-uent in English

    have restricted access to the National Curriculum and are severely disadvantaged.

    The research ndings from inner London show that children who are not uent

    in English tend to do less well in KS1, KS2 and GCSE than those pupils who are

    fully uent in English (ILEA, 1987, 1990; Demie et al., 1997;Kendall et al., 1995,

    1998). The London Borough of Tower Hamlets provides an interesting example.

    In comparison with other LEAs, the borough has the largest number of

    Bangladeshi pupils speaking languages other than or in addition to English. Sta-

    tistical evidence from the LEA indicates that Bangladeshi pupils who were not

    uent in English are achieving considerably below the level of other ethnic

    groups. This pattern changes signicantly, however, when the different levels of

    uency in English are taken into account. The borough has made a priority of

    identifying the language needs of pupils and has tackled Bangladeshi under-

    achievement through Section 11 funded Language Achievements Projects. As a

    result of such targeted initiatives, the performance of Bangladeshi pupils has con-

    tinued to improve, with stage 3 and stage 4 bilingual pupils achieving better, on

    average, than English-only speakers (London Borough of Tower Hamlets, 1994,1998). These ndings from inner London are signicant for education policies

    which aim to target resources to raise the achievements of ethnic minority pupils.

    They illustrate the importance of additional English language support to help

    pupils to be able to fully access the National Curriculum.

    Other studies also provide an alternative explanation for the underachievement

    of ethnic minority children, including unintentional racism (Rampton, 1981;

    Swann, 1985); differences in socio-economic conditions (Swann, 1985;

    OFSTED, 1996); prejudice on the part of some teachers; inappropriate curric-ula and teaching materials; the discouraging effect of relatively poor employment

    prospects after leaving schools, resulting from discrimination in the labour

    market; lack of adequate support to schools and teachers from some Caribbean

    and other black parents; and inadequacy of the understanding of Caribbean and

    other black children by both schools and teachers (Rampton, 1981).

    This research on performance by ethnic group and gender has been valuable

    in advancing our knowledge on the nature and extent of difference in educational

    outcomes. However, much of the work in this area mainly relates to secondaryschools, where statistical records are more extensive. There have been compar-

    able differences at primary schools too, but unfortunately, there is insufcient evi-

    Short report 93

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    4/16

    gender factors to explain differences in achievement between ethnic groups are

    rarely reported. Thus it is not possible to tell from most studies whether white

    pupils do better than the major ethnic minority groups through KS1 and KS2 in

    primary schools, or whether boys of different ethnic groups perform better than

    girls. Furthermore, previous studies lacked data on differences in performance

    between the different ethnic groups by level of uency in English.This report considers evidence from an inner London borough and examines

    performance differences among the main ethnic groups, during the early stages

    of primary education and at the end of their secondary schooling by gender and

    level of English acquisition.

    Method

    The sample

    This paper is an LEA case study using 1998 key stages 1 and 2 and GCSE data.

    The sample consisted of 2,340 pupils who had completed KS1, 2,267 pupils at

    KS2 and 1,225 pupils at GCSE in inner city schools. The analysis is based on

    the results of these pupils.

    Measures of pupil background Free school meals this measure is frequently used as proxy for level of depri-

    vation in the area served by schools.

    Ethnic group each pupils ethnic origin was recorded in one of 14 ethnic

    groups. Details of the ethnic groups and other descriptive statistics of the KS1,

    KS2 and GCSE cohorts used in the study are given in the Appendix.

    Fluency in English a pupils uency in English is measured by one of the four

    stages developed by the ILEA, ranging from stage 1 (beginner) to stage 4

    (uent). The uency codes used and their denitions are summarized below:

    Stage 1: Bilingual English learners who might be able to engage in classroom

    learning activities using their mother tongue, but need support to operate in

    English. (Bilingual refers to all pupils who have access to or need to use two

    languages at home and at school. It does not imply uency in either language

    and includes pupils just beginning to learn English.)

    Stage 2: Bilingual English learners who can engage in all learning activities

    but whose spoken and/or written English clearly shows that English is not theirrst language. Their oral English is well developed but their literacy develop-

    ment is such that they need considerable support to operate successfully in

    written activities in the classroom.

    Stage 3: Bilingual pupils whose oral and written English is developing well and

    who can engage successfully in both oral and written activities, but where

    further support is considered necessary for a variety of possible reasons, e.g.

    pupils who can read aloud accurately but experience disproportionate dif-

    culty in reading comprehension.Stage 4: Bilingual pupils whose use of English and engagement with the cur-

    riculum are considered successful and who do not require additional support.

    94 Educational Research Volume 43 Number 1 Spring 2001

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    5/16

    All primary and secondary schools were asked to provide the name, date of birth,

    sex, meals status, ethnic group and uency in English for each pupil taking KS1

    and KS2 tests and GCSE examinations.

    Measures of performance

    Pupils aged ve to 16 years in schools are taught the National Curriculum sub-

    jects, including English, mathematics and science.This is divided into four stages:

    KS1, KS2, KS3 and KS4 and is age dependent.

    The National Curriculum sets the standard of achievement in each subject

    and, for most subjects, these standards range from levels 1 to 8 between KS1 and

    KS3. Pupils climb the levels as they get older and learn more. Thus a typical

    seven-year-old is expected to achieve level 2 and an 11-year-old level 4. There-

    fore the measure used in the analysis is level 2 or above for KS1 performance,level 4 or above for KS2 and A*C for GCSE.An overall indicator of pupil attain-

    ment in KS1 and KS2 was also derived by taking the average of the three tests

    i.e. English, mathematics and science for each school and the LEA.

    Attainment at the end of KS1, KS2 and GCSE

    Table 1 presents a summary of the results for KS1 and KS2. Overall, this infor-

    mation indicates that 77 per cent of seven-year-olds achieved level 2 or above and

    54 per cent of 11-year-olds achieved level 4 or above.

    Prior to the introduction of the National Curriculum, the statistic routinely

    available for monitoring performance at GCSE was the proportion of pupils

    obtaining ve or more grade A*C. There is no data for KS1 and KS2 which

    would permit trend analysis, but the results for GCSE indicate another dimen-

    sion to the overall picture of performance between 1990 and 1998.The propor-

    tion getting 5+ A*C passes has been rising in recent years. Figure 2 shows the

    improvement in performance. In general, the LEA schools have been achievinghigher 5+ A*C grades than in previous years. There was an increase from 16.9

    per cent in 1990 to 28.8 per cent in 1998. This represents an improvement of

    11.9 per cent. The national gure improved between 1992 and 1998 by 8 per

    cent (see Figure 1).

    Recent data allow us to analyse the results by ethnic background. The LEA

    schools contain a high proportion of English/Scottish/Welsh and Caribbean and

    African pupils. Of about 28,000 pupils on roll, 27.4 per cent were English/Scot-

    tish/Welsh, followed by 22.7 per cent Caribbean, 17.2 per cent African, 10.7per cent other black, 5.9 per cent other white and 3.8 per cent Portuguese.

    There are much lower proportions of other ethnic minority pupils. The find-

    ings confirm that there were substantial differences in performance between

    different ethnic groups at each key stage. Of all the three main ethnic groups,

    English/ Scottish/Welsh did better in their performance compared to the LEA

    average.

    African pupils outperform Caribbean pupils at each key stage. They also have

    higher levels of achievement than Pakistani,Vietnamese, Irish and English/Scot-tish/Welsh pupils at key stage 1, although this is not true of key stages 2 and 4.

    One reason is that KS2 and KS4 pupils are not as long-established in the borough

    Short report 95

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    6/16

    96 Educational Research Volume 43 Number 1 Spring 2001

    LE1

    Performanceatkeystages1and2andGCSE

    ,byethnicbackground

    c

    KS1cohorts

    KS1Results

    KS2co

    horts

    KS2Results

    GCSEco

    horts

    GCSE

    roun

    d

    No.of

    %ageac

    hieving

    No.of

    %ageac

    hiev

    ing

    No.of

    %a

    geac

    hiev

    ing

    pupils

    leve

    l2+

    pupi

    ls

    leve

    l4+

    pupi

    ls

    A*Cgrades

    an

    435

    81.1

    378

    48.3

    196

    37.2

    ladeshi

    40

    72.5

    35

    46.0

    22

    37.0

    bbean

    516

    79.6

    575

    47.0

    251

    25.6

    ese

    21

    85.7

    20

    77.3

    22

    50.5

    n

    53

    83.0

    47

    66.0

    27

    58.6

    tani

    34

    76.5

    45

    66.7

    10

    39.0

    amese

    20

    76.7

    27

    59.3

    17

    48.6

    rblack

    266

    75.1

    242

    54.7

    76

    33.4

    sh/Scot

    tish/Welsh

    636

    79.7

    546

    60.0

    301

    40.3

    k

    11

    78.8

    9

    61.3

    3

    78.2

    17

    62.7

    36

    59.3

    26

    57.8

    sh

    17

    64.7

    21

    55.0

    7

    15.3

    rwhite

    131

    67.4

    107

    46.0

    71

    42.7

    uguese

    89

    44.6

    83

    0.0

    45

    30.2

    assied

    54

    56.8

    63.0

    35

    age

    76.6

    53.6

    37.2

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    7/16

    less uent in English and others are recent arrivals or refugees. The combined

    inuence of these factors can be seen in the relatively poor performance of

    African children.

    The achievement of Caribbean pupils is a particular cause for concern. People

    of Caribbean origin make up the largest ethnic minority in the LEA and have

    been a focus of attention among policy-makers. This is partly a reection of nd-

    ings from London Reading Testand GCSE results which have consistently shown

    that Caribbean pupils have performed below both the average for the borough

    and that for the other main ethnic groups. This is further supported by an analy-

    sis of National Curriculum results that show at the end of KS2 and GCSE,

    Caribbean pupils achieved below the borough average (see Table 1). It is inter-

    esting to note here that Caribbean pupils performed better than Bangladeshi

    pupils at KS1 and KS2, but at the end of KS4, this performance was reversed in

    favour of the Bangladeshi pupils. This nding mirrors the ndings of previous

    research into the performance of Bangladeshi pupils, based on the LondonReading Test and GCSE results. The research shows that once bilingual pupils

    become uent in English, they perform up to and, in many cases, beyond the

    LEA average. Bangladeshi pupils at KS1 and KS2 have lower uency levels, and

    as they progress to higher levels of English uency, they perform better than the

    Caribbean group. Overall, at KS2 and KS4 the Caribbean pupils appear to

    achieve lower scores than all major ethnic groups.

    The results of the National Curriculum tests at ages seven, 11 and 16 years in

    the LEA also show that Bangladeshi pupils are the main underachieving groupat the infant and junior levels. However, by GCSE, Bangladeshi pupils are per-

    forming close to the borough average, and have overtaken Caribbean pupils. The

    main reason for the Bangladeshi underachievement from this study is clear; at

    age seven about 90 per cent of Bangladeshi children in the LEA were in the early

    stages (1, 2 or 3) of English uency. By age 11, just under 67 per cent of

    Bangladeshi children were in the early stages of uency, while by the time they

    took GCSE, the majority were reasonably procient in English. These gures

    offer much encouragement. They demonstrate that once the disadvantage of lan-guage is overcome, it is possible for an ethnic group to catch up with other groups

    who have outperformed them at the early stages of education.

    Short report 97

    FIGU RE 1 Proportion of pupils gaining 5+ A*C passes, 19908

    Source: DfEE Statistics for Education: Public Examinations GCSE and GCE in England,

    19901998.

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    8/16

    Vietnamese pupils are relatively small ethnic groups in the LEA, and so it is more

    difcult to draw rm conclusions from the data. However,Table 1 replicates nd-

    ings from other studies, that Chinese and Indian pupils tend to be the highest-

    performing ethnic group (Thomas et al., 1994; Kendall et al., 1995, 1998).

    Vietnamese pupils in this study also perform well, especially at KS2 and KS4,

    although they under-perform at KS1.Again, this could, in part, be due to improv-ing performance as prociency in English improves. Information about new

    groups, such as Portuguese pupils, is more scarce and they are not recognized

    nationally as a separate category. The extra category was added due to anecdotal

    evidence that Portuguese pupils were under-performing. This is borne out by

    their performance at KS1,which is the lowest of any ethnic group in the LEA.

    Factors inuencing school performance

    A number of factors may be responsible for the under-achievement of ethnic

    minority pupils, including levels of English language acquisition, poverty and

    social class. A major factor in the under-performance of ethnic groups such as

    African and Portuguese at KS2 is the level of uency in English. For example,

    the results of the KS2 analysis show that the percentage of pupils attaining level

    4 and above in each subject increased as the stage of prociency in English

    increased. Bilingual pupils who were fully uent in English were much more likely

    to get level 4 or above in KS2 when compared with English-only speakers (see

    Figure 1 and the Appendix).

    There was little difference in the performance of stage 1 and 2 pupils in math-

    ematics at KS2, but there was a marked difference for English, as might be

    expected. Generally, the ndings of this study indicate that the higher the level

    of uency in English, the better the pupil performance.

    A similar pattern emerges when the data are further analysed by ethnic back-

    ground. Table 2 shows the association between stages of English uency, ethnic-

    ity and achievement for KS2. The results are especially important because theysupport the nding of previous studies that performance increases with the

    increased level of uency in English for the major under-performing groups.

    Analysis of the 1998 GCSE examination results also shows that background

    factors, such as uency in English and economic disadvantage, continue to have

    a strong inuence on the examination performance of different groups of pupils.

    For example, the results of the GCSE analysis by uency in English indicates that

    the percentage of pupils attaining A*C grades increases as stage of English pro-

    ciency increases. Free school meals also demonstrate a similar pattern,with about28.9 per cent of those entitled to free meals achieving A*C grades compared with

    98 Educational Research Volume 43 Number 1 Spring 2001

    TABLE 2 Key stage 2 performance and stages of English language uency, by

    ethnic group (level 4+)

    Stages of English KS2 results (%) No. of pupils

    uency

    African Bangladeshi Portuguese African Bangladeshi Portuguese

    Bilingual stage 1 18 0 0 11 1 25

    Bilingual stage 2 26 22 30 65 12 27

    Bili l 3 58 51 73 115 13 15

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    9/16

    43.6 per cent who were not (see the Appendix).This result therefore further con-

    rms previous ILEA ndings, which show that the higher the proportion eligible

    for free meals, the lower is the performance in examination results.

    Apart from the factors discussed above, earlier research has frequently revealed

    the discrepancy between the performance of boys and girls (Arnot et al., 1998;

    Gillborn and Gipps, 1996). Evidence from a number of studies suggests that girlsoutperformed boys in virtually all subjects (Warrington and Younger, 1997;

    Gallagher, 1997; OFSTED, 1996) at each key stage.

    In this study, performance of boys and girls, by ethnic background, was exam-

    ined for all key stages. Table 3 repeats the pattern established earlier, whereby

    girls tend to outperform boys at each key stage. Overall, the ndings of the results

    between key stages indicate that girls achieve higher averages than boys. This is

    true for African, Caribbean, English/Scottish/Welsh, and Vietnamese pupils at all

    key stages. It also conrms that for Bangladeshi pupils boys achieve better thangirls at KS2, but at KS4 girls are catching up and in fact doing better than boys.

    However, sex and ethnic differences in achievement are not uniformly in favour

    Short report 99

    FIGU RE 2 Key stage 2 performance, by stages of English uency

    FIGU RE 3 GCSE A*C grade performance and stages of English uency

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    10/16

    KS2. In English, Caribbean, English/Scottish/Welsh, African, and Vietnamese

    girls do signicantly better than boys at all key stages. A similar pattern is

    observed for maths and science for these major ethnic groups. In addition, the

    results conrmed that Chinese and Vietnamese boys achieve better than girls in

    mathematics at KS2 and Indian boys in maths and English and Pakistani boys in

    science. By the end of KS4, however, girls do better in all subjects than boys.

    Given the relatively small number involved in some ethnic groups, this picture

    may well vary from one year to another.

    As always such ndings are incomplete without further analysis by social class.

    In this case, there are no data within the LEA which would permit an analysis of

    the performance by ethnic background and social class. It would be of con-

    siderable interest to know how the interaction of these variables inuences per-

    formance at KS1, KS2 and GCSE. However, the ndings of research studiespublished a decade ago show that social class is strongly associated with achieve-

    ment regardless of ethnic background (Drew and Gray, 1990).

    Differences in performance between schools and key issuesfor school improvement

    Another important aspect of this study has been the feedback to schools. Eachschool is provided with their own performance data for each of the three key

    stages. Figure 4 (and the Appendix) provide a sample of benchmarking infor-

    mation that is used by the LEA to support school improvement and target setting

    in schools. It also gives school-by-school differences in performance by ethnic

    background. The LEA provides this information and other detailed statistical

    data as part of its support for school self-evaluation (see Demie, Taplin and

    Butler, 1999).

    The issues which arise from the above ndings, and critical analysis of thesendings, have important messages to convey regarding educational opportunity

    of under-performing groups in schools. Many of them are familiar equal oppor-

    100 Educational Research Volume 43 Number 1 Spring 2001

    TAB LE 3 Performance, by key stage, gender and ethnic background (percen-

    tages)

    Ethnic KS1 level 2+ KS2 level 4+ GCSE A*C

    background

    Girls Boys Average Girls Boys Average Girls Boys Average

    African 84.0 78.5 81.1 50.0 46.0 48.3 41.8 27.9 37.2

    Bangladeshi 75.0 70.8 72.5 43.0 49.0 46.0 41.7 32.5 37.0

    Caribbean 80.7 78.5 79.6 48.0 46.0 47.0 28.6 18.3 25.6

    Chinese 80.5 92.6 85.7 78.0 77.0 77.0 58.3 35.7 50.5

    English/Scottish/

    Welsh 79.6 75.9 79.7 63.0 59.0 60.0 40.2 40.5 40.3

    Indian 84.5 80.0 83.0 63.0 67.0 66.0 58.5 58.7 58.6

    Irish 13.3 83.3 62.7 66.0 56.0 59.3 40.2 47.5 57.8

    Vietnamese 88.9 58.3 76.7 74.0 43.0 59.0 67.3 10.0 48.6Portuguese 51.5 40.8 44.6 N/A N/A N/A 33.3 26.1 30.2

    All pupils 79.4 74.4 76.6 56.0 53.0 53.7 41.2 34.6 37.2

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    11/16

    schools to use individual pupils performance data for school self-evaluation and

    target setting.

    Overall, the results in Figure 4 show how well schools can do, whatever their

    circumstances. They also conrm that there is a wide range of performance

    between schools within the LEA. However, while overall there is a relationship

    between ethnic background and achievement at KS2, some schools with highlevels of disadvantage also have excellent results with English/Scottish/Welsh,

    Caribbean and African children. These schools might be considered to be doing

    better than expected and may be seen as benchmarks for success. In contrast,

    there are schools with low scores based on the background indicators above which

    attained a lower percentage at level 4 or above (see also ibid.).These schools might

    be considered to be doing less well than expected. Our benchmarking information

    is based on a crude analysis, but conrms that there is a wide range of perform-

    ance between schools within the LEA.It is important to note that the relative position of a school on each of the indi-

    cators may be due to a number of factors. Figure 4 shows the comparative pos-

    ition for schools on a variety of indicators, including KS2 performance by three

    main ethnic groups within the LEA. Information of this nature,while illustrating

    the position of schools relative to the LEA average, does not of course explain

    why they occupy that position: it is a matter for headteachers, governors and the

    LEA to make sound judgements about why schools with similar characteristics

    perform differently. Comparative information such as this is not a substitute for

    detailed monitoring or review in schools themselves. However, it triggers a series

    of questions concerning school improvement, and suggests areas of discussion

    once headteachers, governors and teachers have compared their own schools

    performance with those in the bar graphs, the LEA and national averages:

    How well are we currently doing?

    How does my school compare with other local authority schools and the

    national average?

    Why is it that some schools can achieve results that are so much better? What is it that these schools are doing in raising the achievement of under-

    performing groups that we can learn from?

    What can be done to address and improve the underachievement of different

    ethnic groups?

    These are questions which headteachers, governors and school staff should ask

    themselves as a basis for improving their work and raising achievement of under-

    performing groups in schools.

    Conclusion

    In this study we have examined the educational achievement of ethnic minority

    pupils at the end of each of three key stages. The ndings of this study suggest

    that children from different ethnic groups show differences in educational attain-

    ment at the end of each key stage. Indian, Vietnamese and Chinese children

    achieve higher results, on average, than African, Caribbean, Irish and English/Scottish/Welsh pupils. Caribbean, African and Portuguese pupils are the main

    underachieving ethnic groups. Underachievement becomes pronounced for

    Short report 101

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    12/16

    102 Educational Research Volume 43 Number 1 Spring 2001

    GURE

    4

    Keystage2performance,bymainethnic

    groups(level4orabove)

    A

    0%

    0%

    0%

    0%

    00%

    37

    47

    5

    66

    8

    09

    03

    54

    63

    71

    75

    23

    26

    06

    33

    14

    70

    04

    42

    52

    39

    76

    07

    41

    44

    40

    31

    74

    13

    05

    77

    46

    10

    32

    85

    57

    02

    49

    65

    01

    59

    53

    55

    62

    36

    35

    88

    58

    38

    34

    72

    19

    27

    29

    30

    oo

    al

    B

    A

    0%

    0%

    0%

    0%

    00%

    3

    0

    47

    04

    37

    5

    54

    286

    6

    23

    36

    63

    33

    01

    15

    71

    08

    20

    10

    55

    73

    77

    85

    46

    70

    5

    88

    09

    32

    44

    06

    05

    13

    74

    03

    38

    40

    78

    35

    14

    26

    57

    02

    62

    45

    41

    53

    34

    19

    29

    7569

    52

    39

    72

    27

    49

    oo

    al

    S

    TISH

    ,

    ELSH

    0%

    0%

    0%

    0%

    00%

    2

    2

    4

    4

    63

    07

    70

    03

    1

    08

    36

    74

    77

    73

    75

    20

    02

    2

    58

    530

    5

    40

    31

    54

    32

    66

    23

    78

    10

    59

    06

    45

    46

    57

    72

    37

    88

    33

    38

    85

    554

    1

    62

    35

    09

    65

    39

    30

    27

    52

    34

    01

    44

    14

    69

    19

    49

    oo

    al

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    13/16

    However, this picture needs to be qualied in a number of ways. First, the

    different ethnic groups often show different gender composition. Gender is

    strongly associated with achievement regardless of ethnic background. The nd-

    ings of this research indicate that girls outperform boys at KS1, KS2 and GCSE.

    Second, none of these ethnic categories is homogeneous.The African group, for

    example, is itself made up of groups whose language uency levels differ from oneparticular ethnic constituent to another. Pupils from Somalia and Zaire, and some

    ethnic groups from the former French colonies and North Africa, are considered

    as having a lower level of uency in English when compared to African ethnic sub-

    groups, for example, from Kenya, Ghana, Zimbabwe, South Africa or Nigeria.

    Third, it must be taken into consideration that the numbers in some of these

    ethnic groups are relatively small, and any conclusions or interpretations should

    be made with care, since the performance of a few pupils can signicantly weight

    the overall performance of a group.

    Implications for school improvement

    The current local and national government school improvement initiatives put

    much emphasis on the monitoring of performance and the need to identify the

    factors behind underachievement. The focus has been on disseminating per-

    formance data to schools to help support them in identifying areas for improve-

    ment. The ndings in this study show that Caribbean, African, Bangladeshi and

    Portuguese pupils are under-performing compared to English/Scottish/Welsh

    and other ethnic groups. This is hardly surprising, given that some of the African,

    Bangladeshi and Portuguese pupils are the most recently arrived pupils or

    refugees, and therefore represent the most economically deprived groups with

    additional needs for English language support. The under-performance of these

    groups remains a cause for concern and is obviously an issue that policy-makers

    and schools need to address. There is a need for strategies to be developed to

    raise levels of achievement among these groups.

    Implications for English language support

    The ndings of this research conrm that there is a strong relationship between

    the stages of uency in English and the achievement of bilingual pupils. In

    general, empirical evidence from the LEA shows that the performance levels of

    ethnic minority pupils increase as uency in English increases. Pupils in the early

    stages of English uency perform at very low levels, while bilingual pupils who

    are reasonably procient in English perform better, on average, than English-onlyspeakers at KS1, KS2 and at GCSE. Bilingual pupils assessed as fully uent in

    English perform much higher than the national average at all ages. These nd-

    ings offer much encouragement for policy-makers and school improvement prac-

    titioners. They demonstrate that once the disadvantage of language is overcome,

    it is possible to attain high levels of achievement for all key stages.

    Research agenda for addressing ethnic differences in educational

    achievement

    Our study is not complete, and it is useful to identify some of the gaps revealed

    Short report 103

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    14/16

    ethnic differences in achievement reveals that there has so far been little research

    into:

    what strategies schools are adopting to address underachievement of different

    ethnic groups

    the extent to which different strategies are addressing ethnic differences inachievement

    how the experience of successful schools may be disseminated within the LEA

    to address underachievement in other similar schools.

    The rst step in answering these questions is to identify successful schools and

    the strategies used in raising the achievements of under-performing groups,

    including ethnic minority pupils. Some schools are already using a wide range of

    different initiatives and good practices to address the issue. The strategies mostfrequently raised by schools during training sessions on school self-evaluation

    were: target setting, new teaching methods, parental involvement, role model-

    ling/mentoring, staff training, language support for bilingual pupils through the

    Ethnic Minority Achievement Grant and literacy and numeracy initiatives.

    However, the extent to which these strategies have successfully addressed this

    issue has not been documented. There is a need to explore further these strat-

    egies and draw good practices that may be shared between similar schools.

    The data used in the research do not reveal whether the variations in attainment

    are the result of differences between pupils that existed before they arrive in

    school, or whether they have developed during the course of schooling. To answer

    these questions requires,at the very least, that each pupils progress is charted from

    the point of arrival in school to the point at which the National Curriculum tests

    and public examinations are taken. Further, it would be interesting to examine

    whether certain groups fare better at one school than another. Any extension to

    this work could involve using the data for value-added analysis which takes into

    account such factors and weights them statistically. Another factor which has not

    been examined here is that of social class. These ethnic groups may show a totallydifferent class composition, which may further explain the difference in achieve-

    ment between the groups. Further research over time will help to determine the

    trends of achievement.

    Despite these limitations, the paper contains a wealth of empirical data relat-

    ing to the performance of different ethnic groups that may be used as baseline

    information for subsequent studies.

    References

    ARNOT, M., GRAY, J., JAMES, M., RUDDUCK, J. and DUVEEN, G. (1998). Recent

    Research on Gender and Educational Performance. OFSTED Reviews of Research.

    London: HMSO.

    DEMIE, F., REID, A. and BUTLER, R. (1997). Pupil Achievement and Ethnic Background:

    Results of the Analysis of the 1997 National Curriculum Results. Research and Statistics

    Unit, Southwark Education, London.

    DEMIE, F., TAPLIN, A. and BUTLER, R. (1999). School Prole: Making Figures Speak

    for Themselves. London: Research and Statistics Unit, Lambeth Education.

    DREW D and GRAY J (1990) The fth year examination achievements of black young

    104 Educational Research Volume 43 Number 1 Spring 2001

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    15/16

    Evidence on the Apparent Underachievement of Boys. Belfast: Queens University, School

    of Education.

    GILLBORN, D. and GIPPS, C. (1996). Recent Research on the Achievement of Ethnic

    Minority Pupils: OFSTED Reviews of Research. London: HMSO.

    INNER LONDON EDUCATION AUTHORITY (1987). Actual and Predicted Exami-

    nation Scores in Schools. ILEA Research and Statistics Branch.

    INNER LONDON EDUCATION AUTHORITY (1990). Differences in Examination

    Performance. ILEA Research and Statistics Branch.

    JESSON, D., GRAY, J. and TRANMER, M. (1992). GCSE Performance in Notting-

    hamshire, 1991: Pupil and School Factors. Shefeld: University of Shefeld/Notting-

    hamshire County Council.

    KENDALL, L. (1995). Report on the Analysis of 1994 Examination Results: NFER/AMA

    Project on Examination Results in Context. London: Association of Metropolitan

    Authorities.

    KENDALL, L. (1998). Report on the Analysis of 1997 Examination Results: NFER/AMA

    Project on Examination Results in Context. London:Association of Metropolitan Author-

    ities.

    LONDON BOROUGH OF TOWER HAMLETS (1994). Analysis of the 1994 GCSE

    Results. London Borough of Tower Hamlets.

    LONDON BOROUGH OF TOWER HAMLETS (1998).Achieving against the Odds: A

    Summary of Performance in Tower Hamlets. London Borough of Tower Hamlets.

    MABEY, C. (1981). Black British literacy, Educational Research, 23, 2, 8395.

    MORTIMORE, P., SAMMONS, P., STOLL, L., LEWIS, D. and ECOB, R. (1988).

    School Matters:The Junior Years. Yeovil: Open Books.

    NUTTALL, D., GOLDSTEIN, H., PROSSER, R. and RASBASH, J. (1989). Differen-

    tial school effectiveness, International Journal of Educational Research, 13, 76976.

    OFFICE FOR STANDARDS IN EDUCATION (1996). The Gender Divide: Performance

    Differences between Boys and Girls at School. London: HMSO.

    RAMPTON REPORT. GREAT BRITAIN. DEPARTMENT OF EDUCATION AND

    SCIENCE (1981). West Indian Children in Our Schools. London: HMSO.

    SMITH, D. and TOMLINSON, S. (1989). The School Effect: A Study of the Multi-social

    Comprehensive. Exeter: Policy Studies Institute.

    SWAN REPORT. GREAT BRITAIN. DEPARTMENT FOR EDUCATION AND

    SCIENCE (1985). Education for All. London: HMSO.

    THOMAS, S., PAN, H. and GOLDSTEIN, H. (1994). Report on the Analysis of 1992

    Examination Results: AMA Project on Putting Examination Results in Context. London:

    Association of Metropolitan Authorities.

    WARRINGTON, M. and YOUNGER, M. (1997). Gender and achievement: the debate

    at GCSE, Education Review, 10, 1, 217.

    Short report 105

  • 8/6/2019 Ethnic and Gender Difference in Educational Achievements

    16/16

    Appendix:Key

    stages1and2testperformance,byethni

    cgroup,genderandsubjects

    Test/Tas

    kKey

    Stage

    1(percentageac

    hiev

    ing

    leve

    l2+)

    Test

    /Tas

    kKey

    Stage

    2(percentageac

    hiev

    ing

    leve

    l4+)

    CohortNum

    ber

    Re

    ading

    Writing

    Mathemat

    ics

    Aggregate

    Eng

    lish

    Mat

    hemat

    ics

    S

    cience

    Aggregate

    KS2

    KS1

    Boys

    GirlsBoys

    GirlsBoys

    GirlsBoys

    Girls

    Boys

    GirlsBoys

    GirlsBoy

    sGirlsBoys

    GirlsBoy

    sGirlsBoys

    Girls

    Unclassied

    63.2

    55.0

    68.4

    60.0

    68.4

    55.0

    66.7

    56.7

    66%

    61%

    67%

    51%

    70%

    71%

    68%

    61%

    11

    4

    126

    28

    26

    African

    76.7

    84.0

    76.2

    84.0

    82.5

    84.0

    78.5

    84.0

    46%

    52%

    45%

    48%

    47%

    50%

    46%

    50%

    14

    1

    141

    223

    212

    Bangladeshi

    62.5

    75.0

    70.8

    62.5

    79.2

    87.5

    70.8

    75.0

    47%

    43%

    47%

    48%

    53%

    38%

    49%

    43%

    1

    9

    141

    24

    16

    Caribbean

    76.4

    80.2

    75.7

    81.4

    83.3

    80.6

    78.5

    80.7

    43%

    54%

    45%

    41%

    50%

    48%

    46%

    48%

    26

    2

    256

    263

    253

    Chinese

    88.9

    83.3

    88.9

    83.3100.0

    75.0

    92.6

    80.5

    69%

    73%

    85%

    80%

    77%

    80%

    77%

    78%

    1

    3

    15

    9

    12

    Indian

    68.2

    90.3

    77.3

    90.3

    72.7

    90.3

    72.7

    90.3

    67%

    63%

    72%

    58%

    61%

    68%

    67%

    63%

    1

    8

    19

    22

    31

    Pakistani

    84.6

    57.1

    92.3

    71.4

    84.6

    81.0

    87.2

    69.8

    53%

    77%

    67%

    73%

    67%

    59%

    62%

    70%

    1

    5

    22

    13

    21

    Vietnamese

    50.0

    83.3

    37.5

    91.7

    87.5

    91.7

    58.3

    88.9

    38%

    77%

    54%

    69%

    38%

    77%

    44%

    74%

    1

    3

    13

    8

    12

    Otherblack

    65.5

    76.0

    70.3

    78.5

    77.9

    84.3

    71.2

    79.6

    46%

    61%

    53%

    55%

    49%

    63%

    49%

    60%

    9

    5

    96

    145

    121

    ESW

    73.4

    83.7

    72.0

    83.3

    82.2

    86.5

    75.9

    84.5

    50%

    62%

    60%

    59%

    62%

    67%

    57%

    63%

    28

    3

    245

    354

    282

    Greek

    60.0

    66.7

    80.0

    83.3100.0

    83.3

    80.0

    77.8

    100%

    50%

    100%

    50%

    50%

    50%

    83%

    50%

    2

    4

    5

    6

    Irish

    83.3

    0.0

    75.0

    20.0

    91.7

    20.0

    83.3

    13.3

    47%

    71%

    53%

    57%

    67%

    71%

    56%

    67%

    1

    5

    7

    12

    5

    Turkish

    60.0

    42.9

    60.0

    85.7

    80.0

    85.7

    66.7

    71.4

    43%

    78%

    57%

    44%

    57%

    67%

    52%

    63%

    7

    9

    10

    7

    Otherwhite

    65.4

    64.2

    67.9

    73.6

    69.2

    73.6

    67.5

    70.5

    39%

    44%

    53%

    42%

    51%

    45%

    48%

    44%

    7

    9

    78

    78

    53

    Portuguese

    35.6

    36.4

    37.8

    59.8

    48.9

    59.1

    40.8

    51.5

    n/a

    n/a

    n/a

    n/a

    n/a

    n/a

    n/a

    n/a

    n/a

    n/a

    45

    44

    Unclassied

    77.8

    76.2

    80.0

    81.0

    84.4

    73.8

    80.7

    77.0

    66%

    60%

    67%

    51%

    70%

    70%

    68%

    61%

    12

    2

    134

    54

    48

    Freemeal

    62.8

    71.5

    62.4

    72.5

    73.5

    77.1

    66.2

    73.7

    38%

    51%

    47%

    44%

    47%

    50%

    44%

    48%

    48

    0

    482

    559

    516

    Paidmeal

    78.6

    83.4

    79.9

    85.3

    85.1

    86.8

    81.2

    85.2

    55%

    64%

    59%

    58%

    61%

    64%

    58%

    62%

    47

    4

    436

    626

    537

    Unclassied

    68.2

    65.4

    72.7

    65.4

    72.7

    65.4

    71.2

    65.4

    74%

    65%

    74%

    52%

    78%

    75%

    76%

    64%

    9

    4

    113

    31

    32

    Englishonly

    72.7

    80.7

    72.7

    83.3

    81.2

    83.3

    75.5

    82.4

    46%

    58%

    52%

    50%

    54%

    58%

    51%

    55%

    70

    7

    640

    120

    92

    Stage1

    33.3

    30.4

    35.8

    51.1

    48.3

    51.1

    39.1

    44.2

    14%

    0%

    21%

    35%

    21%

    25%

    19%

    20%

    1

    4

    20

    194

    167

    Stage2

    78.4

    78.4

    81.4

    83.2

    86.1

    83.2

    82.0

    81.6

    22%

    19%

    28%

    23%

    37%

    28%

    29%

    23%

    6

    5

    53

    59

    81

    Stage3

    98.3

    100.0

    94.9

    100.0

    98.3100.0

    97.2

    100.0

    47%

    51%

    57%

    44%

    55%

    45%

    53%

    47%

    9

    3

    98

    33

    36

    Stage4

    90.9

    91.7

    87.9

    94.4

    97.0

    94.4

    91.9

    93.5

    66%

    80%

    69%

    73%

    67%

    75%

    67%

    76%

    10

    3

    128

    802

    693

    Allpupils

    71.4

    77.5

    72.0

    73.9

    79.8

    81.7

    74.4

    79.4

    49%

    58%

    54%

    51%

    56%

    58%

    53%

    56%

    107

    6

    1052

    1239

    1101

    nic

    k-

    und

    e als el n

    cy

    glish

    kground

    icators