Journal of the Royal Statistical Society: Series A (Statistics in Society)

Cover image for Vol. 179 Issue 3

Edited By: H. Goldstein and L. Sharples

Impact Factor: 1.702

ISI Journal Citation Reports © Ranking: 2015: 13/49 (Social Sciences Mathematical Methods); 24/123 (Statistics & Probability)

Online ISSN: 1467-985X

Associated Title(s): Journal of the Royal Statistical Society: Series B (Statistical Methodology), Journal of the Royal Statistical Society: Series C (Applied Statistics), Significance


Multivariate multilevel analyses of examination results by M. Yang et al,
Journal of the Royal Statistical Society, Series A, Statistics in Society, Volume 165 (2002), Part 1, pages 137 - 153

The analyses in the paper were based on A/AS level results in different mathematics papers of 52,587 individuals from 2,592 institutions in England in 1997. The data file 'RSS-A633.TXT' contains 30 fields with one blank space between two fields. The name and code of each field are listed as follows:

  FieldName  Level defined   Coding
  Gender  Individual  Male=0, Female=1
  GCSE-TOT-Score   Individual  Min=0, Max=103
  GCSE-Num-taken   Individual  Min=2, Max=13
  GCSE-MATH-Score   Individual  Min=0, Max=8
  GCSE-MATH-NUM   Individual  Min=1, Max=4
Exam Board   Individual  Associated Exam Board=1
    Cambridge Board=2
    London Board=3
    Oxford Board=5
    Joint Matriculation=6
    Oxford & Cambridge=7
  Mean-GCSEtot   Institution  Min=2.19, Max=7.41
  Mean-GCSEMaths   Institution  Min=0.5, Max=7.36
  Age-Month   Individual  Min=216, max=227
  Institution-type   Institution  LEA Maintained Comprehensive=1
    LEA Maintained Selective=2
    LEA Maintained Modern=3
    Grammar Comprehensive=4
    Grammar Selective=5
    Grammar Modern=6
    Independent Selective=7
    Independent Non-selective=8
    Six-Form College=9
    FE College=10
  Mathscore  Individual  Min=0, Max=10 (outcome variable)
  A-main-maths  Within individual  Alevel Main Maths=1, others=0
  A-pure-maths  Within individual  Alevel Pure Maths=1, others=0
  A-Applied-maths  Within individual  Alevel Applied maths=1, others=0
  A-Further-maths  Within individual  Alevel Further maths=1, others=0
  AS-main  Within individual  ASlevel main maths=1, others=0
  AS-pure  Within individual  ASlevel Pure maths=1, others=0
  AS-Applied  Within individual  ASlevel Applied maths=1, others=0
  AS-Further  Within individual  ASlevel Further maths=1, others=0
  A-combination-1  Individual  A(M)+A(F)+more=1, others=0
  A-combination-3  Individual  A(M)+AS(F)+more=1, others=0
  A-combination-2  Individual  A(M)+AS(others but F)=1, others=0
  A-combination-4  Individual  A(M)+A(others but F)=1, others=0
  A-combination-5  Individual  A(P)+A(A)+more=1, others=0
  AS-combination-6  Individual  AS(M)+more than 1A=1, others=0
  AS-combination-7  Individual  AS(A)+more than 1A=1, others=0
  School ID  School  Ranged from 1 to 2592
  Individual ID  Individual  Ranged from 1 to 245 (nested within school)
  subject level  Within individual  Alevel=1, ASlevel=2
  MathCode  Within individual  Main maths=1
    Pure maths=3
    Applied maths=5
    Further maths=13

All models in the paper were fitted in using the program MLwiN. Details of the program can be found at the web site ( The full paper can also be found and down-loaded from the page (

For technical questions in fitting models appeared in the paper, contact

Min Yang

Department of Mathematical Sciences

Institute of Education

20 Bedford Way

London WC1H 0AL



tel: 44(0)207 612 6682

Dataset (556kb)

Ehrenberg law-like relationship and anthropometry by N. Forcheh, Journal of the Royal Statistical Society, Series A, Statistics in Society, Volume 165 (2002), Part 1, pages 155 - 172.

The data and programs comprise the following files:

a) Contents

1. A627.dat Nutritional data of Ngamiland under-fives: (ASCII format)

2. zhfa.sps SPSS syntax file for computing z-scores for heigth-for-age

3. zwfa.sps SPSS syntax file for computing z-scores for weigth-for-age

4. zwfh.sps SPSS syntax file for computing z-scores for weigth-for-height

1.1 Description of A627.dat

The Data file A627.DAT contains 26 variables. The data have been written in fixed format, with the data layout and formats as follows:

  Variable  Start  End  Format
  Col Col DISTRICT  1  2  F2.0
  CHILD_NO  3  4  F2.0
  CHLD_SEX  5  5  F1.0
  WT_BIRTH  6  9  F4.0
  BRE_FED  10  10  F1.0
  NMEAL  11  11  F1.0
  AGE  12  15  F4.0
  SEX_HEAD  16  16  F1.0
  AGE_HEAD  17  18  F2.0
  MSTART  19  19  F1.0
  EDU_HEAD  20  21  F2.0
  ILL2WKS  22  22  F1.0
  ZZ  23  24  F2.0
  ZHFA  25  30  F6.2
  ZWFA  31  36  F6.2
  ZWFH  37  42  F6.2
  ZSD_WFH  43  43  F1.0
  ZSD_WFA  44  44  F1.0
  ZSD_HFA  45  45  F1.0
  LOG_WT  46  50  F5.1
  HT_METRE  51  55  F5.2
  E_LOG_WT  56  61  F6.3
  EDLEVEL  62  69  F8.0
  DSTAT  70  76  F7.3
  EH_NSTAT  77  77  F1.0
  EH_WFH1  78  78  F1.0

1.2 Variables and value Labels for A627.dat


51 Ngamiland West

52 Ngamiland East


1 Male

2 Female


Yes 2



1 1 Meal

2 2 Meals

3 3 Meals

4 4 Meals

5 5 Meals

6 Exlusively-b-fed

AGE - Age in Months of Child

SEX_HEAD - Sex of Head of House hold

1 Male

2 Female

AGE_HEAD - Age of Head of Household

0 M Less than 1 year

98 Over 97 years

99 M Not Known

MSTART - Marial Status of Head of HHold

1 Married

2 Living together

3 Never married

5 Separated

6 Divorced

7 Widowed

EDU_HEAD - Level of Education of Head of HHold

1 Primary STD 1

2 Primary STD 2

3 Primary STD 3

4 Primary STD 4

5 Primary STD 5

6 Primary STD 6

7 Primary STD 7

8 Unknown Primary

11 Secondary 1

12 Secondary 2

13 Secondary 3

14 Secondary 4

15 Secondary 5

16 Unknown Secondar

21 Post Sec. Cert.

22 Post Sec. Diplom

23 Post Sec. Degree

24 Post Degree

25 Unknwn Post Sec.

99 Never Attended


1 Yes

2 No

ZZ - Cases with missing obs in one or more var.

0 1+ missing

1 No Missing

ZHFA - Standardised Z-score for Height for Age

ZWFA - Standardised Z-score for Weight for Age

ZWFH - Standardised Z-score for Weight for Height

ZSD_ WFH - Weight for Height (Wasting)

1 Severe

2 Moderate

3 Normal

ZSD_WFA - Weight for Age (Malnourished)

1 Severe

2 Moderate

3 Normal

ZSD_HFA - Height for Age (Stunting)

1 Severe

2 Moderate

3 Normal

LOG_WT - Log10(Weight in Kg)

HT_METRE - Height in Metres

E_LOG_WT - Fitted Log-Wt = 0.8h+0.39

EDLEVEL - Education Level of Head

DSTAT - Ehrenberg's Deviation Scores (Log(W) - (0.8h+0.39)

EH_NSTAT - Nutritional Status based on Ehrenbergs Adjusted Relationship

1 Severe

2 Moderate

3 Normal

EH_WFH1 - Nutritional Status based on Ehrenbergs Relationship

1 Severe

2 Moderate

3 Normal

2) A Note about the .SPS files

The files are text files and can be read and modified using any editor. These files can be used to compute the corresponding z-scores, with the names of the variables: height, weight and age changed to correspond to those used in your working data file. The means and standard deviations used are based on the NCHS/WHO recommended references discussed in the paper.

The weight-for-height program includes the variable "age" so that it can be used to compute both weight-for-length (lying height) and weight-for-stature (standing height). Here the NCHS cut-of at 24months is used, but modifications could be made if a child under 24months was measured standing, or a child over 24months was measured lying.

The programs can also be modified to compute z-scores with different means and/or standard deviations.

Consider a typical command line: IF(QF492=1 and Age = 0 ) ZHFA = (Ht_cm - 50.5 )/ 2.3.

Here, QF492 represents the variable, "Gender of child": 1=male, 2=female. The values: 50.5 and 2.3 are respectively, the expected height (cm) and standard deviation of a newborn boy.

3. Contact

Dr. N. Forcheh,

Department of Statistics,

Universtity of Botswana,

Private Bag UB00705,




Tel: 267-355 2695

Dataset (21kb)