Student Academic Records in SA

1

Student academic language proficiency an
investigation into a non-first language
developmental intervention by Dr Cay van der
Merwe, Central University of Technology,
Bloemfonteinand Prof Driekie Hay, University of
the Free State, Bloemfontein SAAIR, Port
Elizabeth September 2009
2

Introduction

• English is the mother tongue of only 8,6 of
  South African students
• As from Grade 4 for the majority of learners no
  mother tongue instruction
• Academic language proficiency level of 85 below
  Grade/K 12

UNIVERSITEIT VAN DIE VRYSTAAT UNIVERSITY OF
THE FREE STATE YUNIVESITHI YA FREISTATA Tel
[phone] E-mail hayd.rd_at_ufs.ac.za
www.ufs.ac.za
3

Implications for higher education

• Students coming from a variety of educational
  backgrounds
• Rote learning
• Problems with note taking, synthesising,
  analysing, critical thinking, reasoning as
  vocabulary is lacking
• Poor reading development levels

UNIVERSITEIT VAN DIE VRYSTAAT UNIVERSITY OF
THE FREE STATE YUNIVESITHI YA FREISTATA Tel
[phone] E-mail hayd.rd_at_ufs.ac.za
www.ufs.ac.za
3
4

Theoretical points of departure

• educational viewpoint collapse of the
  secondary school system
• Moral / transformational viewpoint social
  responsibility
• Economical stance costs of failure / shortages
  in the economy

UNIVERSITEIT VAN DIE VRYSTAAT UNIVERSITY OF
THE FREE STATE YUNIVESITHI YA FREISTATA Tel
[phone] E-mail hayd.rd_at_ufs.ac.za
www.ufs.ac.za
4
5

Role of academic language proficiency testing

• Most SA universities UCT developed PTEEP
• Tests among others comprehension
• Less than 50 placed in an academic language
  proficiency course
• Purpose of our paper is to share findings of an
  investigation into a non-first language
  development intervention

UNIVERSITEIT VAN DIE VRYSTAAT UNIVERSITY OF
THE FREE STATE YUNIVESITHI YA FREISTATA Tel
[phone] E-mail hayd.rd_at_ufs.ac.za
www.ufs.ac.za
5
6
METHODOLOGY

• Analyse UFS data of students from 2006/7, on
• School-leaving outcomes
• English proficiency test outcomes
• English development course outcomes
• End of first year results over all courses

7
3 TYPES OF DATA ANALYSES

• Pre-test vs Post-test
• Means
• Students selected on basis of vulnerability by
  PTEEP (English Proficiency Test), were directed
  to English Development Course. After about 8
  months, underwent similar English Proficiency
  test outcomes compared by Student ttest for
  dependant samples - accounts for
  intracorrelation.
• Result Significant (a0.5) increase in means of
  pre-test to post test
• But Means were 37 and 47, whereas maximum
  scored by anyone in 2006/7 in PTEEP was 90
• And Comparisons of mean differences for
  particular constraints (e.g. vocabulary) werent
  available could prove interesting.

8
3 TYPES OF DATA ANALYSES (cont..)

• CORRELATIONS
• Correlation matrix of 5 variables (not age)
  with 1st and 2nd Eng. Proficiency test scores of
  this group of about 200 students yielded
• - No significant correlation between 1st test
  score and end-of-year results, but
• - Significant correlation between 2nd test
  score and end-of-year results.
• - Bigger and significant correlation between
• 2nd test score and score in English
  Development Course

9
Comparisons among 2006-2007 first-year intakes

• The variables are
• PTEEP Test in English for Educational Purposes
  (UCT)
• Mark Mark attained in an English
  development/basic Language course
• Age Age in years at enrolment (February)
• Mscore An aggregated/weighted total of
  school-leaving results
• Enrol Number of courses (subjects) for which
  a student enrolled
• Passed Number of courses passed during the
  first year.
• Subj. Pass Ratio Number of courses passed
  divided by number of courses enrolled.
• Avg Percent Average percentage attained over
  all courses at end of first year.
• Race African (Black) / White (Caucasian) /
  Coloured / Asian.

10
Comparisons among 2006-2007 first-year intakes
(cont)

• Summarised statistics for these variables

11
Race as a factor

• One-way ANOVAs on PTEEP Mark (EngDevCourse),
  SubjPassRatio and AvgPercent.
• All means differed significantly (a0.05) over
  races.
• Followed-up with post-hoc tests to see which
  differ.
• Used Scheffe (conservative test) to see which
  races differ with regard to variables above.
• .

12
Race as a factor

• Example Scheffe post-hoc table, variable PTEEP
  score (Marked differences are significant at
  plt.05000 )
• African and White means differed significantly
  from each other and from Coloured and Asian
  groups (significantly higher mean scores) in
  PTEEP tests.
• With end of year outcomes, the White students
  results differed significantly from the other 3
  groups, but the paired comparisons for the
  latter means did not differ that much.

Race 1 2 3 4
M47.658 M66.116 M58.581 M57.560
African 1   0.0000 0.0000 0.0000
White 2 0.0000   0.0000 0.0000
Coloured 3 0.0000 0.0000 0.9700
Asian 4 0.0000 0.0000 0.9700
13
PTEEP scores cut-off points

• Choice of cut-off level
• 49 - must do English Development Course
• 50 - dont do English Development Course
• Four categories 0-44 45-49 50-54 55
• ANOVAs on MARK (in English Development Course)
  SubjPassRatio and AvgPercent showed highly
  significant (a0.0001) mean differences.
  Post-hoc tests showed

14
Test group versus control group

• Compared outcomes for groups who tested below 50
  in PTEEP, between those who did an English
  academic proficiency course and those who should
  have attended but (voluntarily) did not.
• End-of schools results (composite) not applied in
  these analyses

15
Test group versus control group (contd)

• Distribution of two groups, in AvgPercent, for
  example.

16
Test group versus control group contd Means
of average percentage obtained differed
significantly () between the two groups.
Variable Mean ENProf Mean NONE t-value df p Valid N ENProf Valid N NONE
SbjPass Ratio 0.6729 0.6481 0.8216 721 0.4116 633 90
Avg Percent 50.6867 46.3404 3.0043 740 0.0028 648 94
17
Essays

• Voluntary essays on students own perceptions of
  benefits of development in academic English
  proficiency were very positive.
• Extend qualitative investigations to systemic
  analysis, with random sample of essays and
  interviewsas well.

18
CONCLUSION
HE obliged to invest resources in English
Academic Proficiency improvement. Trade-off
between burden of shouldering secondary
educations level shortcomings, taking decade at
least,withStudents success .less attrition,
efficient throughput, economic gain, less wastage
and trauma.
Thank You
19

Student academic language proficiency an
investigation into a non-first language
developmental intervention by Dr Cay van der
Merwe, Central University of Technology,
Bloemfontein Prof Driekie Hay, University of the
Free State, Bloemfontein SAAIR Port
Elizabeth September 2009