University of Bristol 14th of March, 2008 Presentation

Download Report

Transcript University of Bristol 14th of March, 2008 Presentation

CREOLE
21 October, 2009
Using English in final examinations:
Rationales and realities
SPINE team
2
What is SPINE?
• Student Performance in National Examinations: the
dynamics of language in school achievement
(SPINE) www.bristol.ac.uk/spine (ESRC/DfID
RES-167-25-0263)
• Bristol team: Rea-Dickins, Yu, Afitska, Sutherland,
Olivero, Erduran, Ingram, Goldstein
• Zanzibar team: Z. Khamis, Mohammed, A. Khamis,
Abeid, Said, Mwevura
• LTA = high stakes; > 50% of school aged children
leave school at the end of Basic Education as
unsuccessful
3
Case 1: context
Khadija is 15 years old and in Form 2 of secondary
school. Her learning in primary school was entirely
through the medium of Kiswahili, with English taught as
a subject. She experienced an abrupt shift from
Kiswahili as medium of instruction on transition to
secondary school as in Form 1 she was expected to
learn all her subjects through English. However, in
reality both her L1 (Kiswahili) and L2 (English) are
used when she is being taught Maths, Science,
Geography and her other subjects. But, at the end of
Form 2 she is required to demonstrate her subject
learning in formal school examinations entirely through
the medium of English.
4
4
• Insert the SPINE diagram here
5
The Policy & the Politics: historical perspectives
• Continuous assessment (CA) was introduced in Tanzania
schools in 1976 after Musoma Resolution (NECTA
Guidelines 1990:1).
• In Musoma Resolution, Nyerere (1974) (see Lema et al
2006:111-112) criticised the examination system existed
since independence and addressed the need to change it.
• He maintained that the assessment procedures valued more
competitive, written type, theoretically and intellectually
focused examinations.
6
The Policy & the Politics: historical perspectives
• He saw the need to make Assessment procedures
more friendly, less competitive and practically
focused.
In that manner, he directed that students’ assessment
must;
• Include works that can enable students to function
effectively in their environments.
• Cover both students’ performance in theoretical
and practical works.
• Be carried out throughout the year.
7
The Policy & the Politics: historical perspectives
• The National Examination Council of Tanzania
(NECTA), started to include projects, character
assessment, exercises and tests in the
assessment system
Teachers were then given responsibility to;
• measure students’ ability to use their acquired
knowledge and skills in their local environments.
• keep records of their progress continuously
8
Tanzania and Zanzibar Education Policies
Both The Tanzania Education and Training Policy
(1990:79) and The Zanzibar Education Policy
(2006:28) stated that continuous assessment (based
on teacher assessment) combined with the final exam
would be the basis for awarding of certificates at
Secondary Education levels.
• The Zanzibar Education Master Plan of 19962006 emphasized consistency between teacher
assessment and National examinations for
selection purposes and quality control (p.43)
9
Language policy and Assessment
• Students’ assessment is affected by Language Policy because
students have to demonstrate their knowledge and abilities through
a language ratified by the system.
• The Zanzibar Education Policy of 1995, Kiswahili version Section
5.72 (p28), The Zanzibar Education Master Plan of 1996-2006 (p43)
and the Education Policy of 2006 (p38), all emphasized the use of
English Language as a medium of instruction for all subjects in
secondary education except for Kiswahili and Islamic studies.
10
Language policy and Assessment
• So, corollary to this, English, as a language of
instruction in teaching science at secondary level
has automatically become language of
assessment; continuous assessment and
examination.
• This implies that students’ opportunities to
demonstrate their knowledge in
examination is limited if they are not
proficient in the language
•
11
Coursework assessment as part of
national examining: implementation issues
At the level of Form II, Teachers Assessment (TA) was
implemented basing on the following assessment
procedures:
• Class works
• Oral questions
• Home works
• Weekly tests
• Terminal tests
•
12
Continuous Assessment format & Guidelines
• Term: ---------
Class: --------- subject---------
Pupils’
Month
Month
Month
Month
Total
Exam Grand
name
1
2
3
4
C/work
60%
10%
10%
10%
10%
40%
5
6
3
8
22
X
Total
100%
30
52
13
Continuous Assessment format & Guidelines
• Note: Coursework assessment will be obtained from
class works, home works and weekly tests or as it will
be directed otherwise by the Department or the
Ministry. All the marks scored by the pupil from the
mentioned activities within a particular month, will be
combined and transformed into percentage of ten (10%)
before filled in the relevant space in this form.
•
Source: Ministry of Education Working Guidelines: Guideline
No. 9 (1995)
14
Continuous Assessment format & Guidelines
Additional notes from another version
• If the term is less or more than four (4) months, four activities
chosen by the teacher him or herself should be picked at
equal intervals. The teacher is advised to give as many tests
as he or she can in a term but should choose only four for
recording.
• In getting pupil’s marks for terminal exam for each subject,
the exam itself should contribute 60% and classroom
assessments (coursework assessments) should contribute
40%
15
Coursework assessment as part of national
examining: implementation issues
• The language to be used is as indicated in the
lesson plan.
Source: Ministry of Education Working Guidelines: Guideline
No. 9 (1995)
• These notes appear to contradict with the first
version which might be leading to confusion. This
version requires the teacher to find average of
marks in a month while the second requires the
teacher to choose one activity from which the
marks can be recorded.
16
Coursework assessment as part of national
examining: implementation issues
• Homework, class work, oral questions, weekly tests
make 40% and the terminal exam makes 60% which
when combined make 100%.
• The hundred percent obtained is sent to the
Department of Curriculum and Examination for further
processing.
There was variability among teachers on:
• The specific activities that constitute teacher
assessment marks.
17
Coursework assessment as part of national
examining: implementation issues
• Number of activities they should take marks from
for each month.
• Decision on students’ absentees
• How to get 10% of marks for each month; some
picked the one that students did better, some
picked randomly, and some put average.
• Understanding of what assessment is to be done.
• Awareness of assessment guidelines
18
Coursework assessment as part of national
examining: implementation issues
• Percentage of marks that go to the final
examination
• Sharing what teachers know and practice about
assessment guidelines among teachers within
schools
• Level of monitoring of Teacher Assessment by
head teachers.
• So, students from the same school and from
different schools are assessed in different ways.
19
Overall, teachers were observed using L1 during the lessons in 49% of
cases. Specifically, they did so some of the time in 20% of cases, most of
the time in 11% of cases; and rarely in 18% of cases.
20
Teachers presented concepts clearly most of the time (28% of cases),
some of the time (62% of cases). Teachers’ presentation of concepts
was unclear (6%) or confusing (3%) in 9% of cases.
21
In 42% of cases observed, teachers rarely probed pupils’ comprehension. They did
not probe pupils’ comprehension at all (24%). Teachers frequently probed pupils’
comprehension (8%); and sometimes probed pupils’ comprehension (24%).
22
Teachers provided feedback to learners using wide range of strategies
(1%); with hardly any or no feedback in 19% of cases. In 78% of
cases, teachers were observed either using some range (24%) or a
limited range (54%) of feedback to the learners
23
In 83% of cases, learners rarely (28%), almost never or never (55%) provided
extended responses; 16% pupils provided extended responses some of the time;
only 2% learners were observed providing extended responses most of the time.
24
Findings
24
25
Findings
25
26
Student performance on exam items
English RC Qn: How whales resemble man
45 students took this item:
• 35.6% = no answer
• 26.7% = wrong answer
• 28.9% = partially correct answer
•
8.8% = correct answer
27
Interview: D1 who didn’t answer Q3 explains
D1:
Int:
D1:
Int:
D1:
“because I did not understand by this
this … resemble” (lines 115-117)
“If I tell you that resemble means ‘to
look like’ … can you do the question
now?
“Yes”
OK so what’s the answer?
“Man … is warm blooded … and whales
also … whales have lungs and man
also have lungs …” (122-133)
28
29
Maths: original & modified question
30
Linguistic accommodations make a difference
Modifications to item
– i.e. changing the word ‘below’ to ‘under’/ ‘younger’ cf.
“below 14 years” (refer to the age where the item is)
was interpreted in 3 different ways:
• includes the 14 year olds (3+2+5+4+2=16)
• cells on the left of the cell containing 14 (10,11,12,13)
• cell below the cell containing 14 (which says 2)
31
Biology:
responses to original item
• No answer
=
67.4%
• Wrong answer
=
21.7%
• Partially correct answer =
6.5%
• Correct answer
4.4%
=
32
Biology: on locusts
33
Biology: on locusts
34
Biology: on locusts
Questions:
a) In which picture do you think the locust will/may
die?
b) Why do you think it will/may die?
35
Original item modified:
• Greater contextualisation
• Simplification of instruction
• 2 structured parts: A & B
• Visual clues to support information retrieval
• Rephrasing of the item
• Altering item layout
36
Results
• Original item
– Only 32.6% of students wrote an answer
– Just under 11% gave a partially correct or
correct answer
• Modified item
– 100% responded to this item
– 42% gave a partially correct answer to Part A
– 53% gave a partially correct answer to Part B
37
Nzige (locust): Changes in student response
OR
Sc
Written responses on
Modified Item
MOD
Score
Comment
H3 0
In picture A the locust may/will
die
I think it will/may dies because
the locust get its breathing by
using its body
2+2
Understands that
locusts breathe using
the body
C2 1
I think it will/may die because
the boy is dipping the locust in
the water to all the bodies with
its trachea that used to
respiration as a respiratory
surface of a locust
2+2
Getting very close to
a very complex
answer “I think a very
able pupil indeed”.
The right answer but
with great difficulty in
expressing this in
English
38
Nzige (locust): changes in response
G2
OR
Written responses on
MOD
Comment
Sc
Modified Item
Score
(Neil Ingram: biologist)
0
I think its because of its
2+1
Test taker locates a locust as not
body covered (immersed)
naturally aquatic and therefore unlikely
completely in the water,
to survive if immersed in water.
and its terrestrial not an
Evidence of a learner using everyday
aquatic. It can’
(informal) knowledge beyond the
curriculum to answer the question but
this level of insight would not be
recognised in the original scoring
39
Means of Form II Exam Results (MoEVT)
KISW
Islamic
BIO
CHEM
MATH
PHYS
studies
2004
49
37.1
18.4
30.8
21.1
25.8
2005
46
44.3
22.6
34.2
15.3
25.4
2006
36
47.9
21.5
34.3
15.5
25.1
2007
35.5
44.5
21.5
32.0
14.3
23.9
2008
50.6
47.5
24.5
34.8
16.4
29.1
40
ENG and MATH
Total: 64.717, school=29.18%,
pupil=70.82%
ENGLISH explains (90.464.717)/90.4=28.41% of the
maths total variance
Total=90.4, School=30.75%,
Pupil=69.25%
41
ENG and BIO
Total=65.646, school=18.17%,
pupil=81.83%
ENGLISH alone explains
(114.968-65.646)/114.968=42.90%
of the total variance in BIOLOGY
Total=114.968, school=18.79%,
pupil=81.21%
42
English and CHEM
Total=158.281, school=23.06%,
pupil=76.94%
ENGLISH explains (275.993158.281)/275.993=42.65% of the
total CHEM variance
Total=275.993, school=22.05%,
pupil=77.95%
43
How about KISWAHILI & ARABIC?
•
Although other two languages (Kiswahili and Arabic) are also
significant predictors of the students’ performance in maths, biology
and chemistry, it is noted that they are less capable of explaining the
variances than ENGLISH.
•
KISWAHILI explains (275.993-88.954)/275.993=31.54% of CHEM total
variance, (114.968-76.585)/ 114.968=33.39% of BIO total variance,
(90.4-75.678)/90.4=16.29% of MATH total variance.
•
ARABIC explains (275.993-203.452)/275.993=26.28% of the CHEM
total variance, (114.968-88.569)/114.968=22.96% of the BIO total
variance, (90.4-71.885)/90.4=20.48% of the maths total variance.
44
Summary of the multilevel models (a)
•
It is very clear that ENGLISH is a significant
and substantial predictor of the students’
performance in MATH, BIO & CHEM.
•
The school-level variances explained in the
cons models as well as in the models
including ENGLISH as the single explanatory
variable demonstrated a substantial
proportion of the variance is attributable to
school factors
45
Summary of the multilevel models (b)
• Not much improvement in terms of fitness of the models
(measured by the change of % of school-level variance in the total
variance).
• Therefore, essential to collect further school- and pupil-level data
to examine what factors (e.g. English language learning
opportunities at home and at school, academic English proficiency)
and how much they account for the variances (in the tradition of
school effectiveness studies) – our plan for nationwide data
collection using pupil and headteacher questionnaires and
vocabulary knowledge test
46
High stakes classroom assessment: the realities
Impact/Potential
Disadvantage
Consequences/Injustice
•Teachers do not use full range
of LTA procedures & processes
•Use of inappropriately
constructed assessment
frameworks
•Inaccurate CWA of learners
•CWA implemented as a series
of tests
•Learners not fully supported in their
language & content knowledge
development & fail to reach potential
•Test performance valued over
learning
•Leave school with poor educational
outcomes
•Inadequate d-base for decision
making about student & learning
progression
47
Using English in high stakes examinations (SSA, UK,
TIMSS, PISA): the realities
Impact/Potential
Disadvantage (examples)
Consequences/Injustice: (examples)
•Learners do not engage or
respond poorly in
examinations
•Learners are not given a
fair chance to show their
abilities
•Subject area (e.g. Biology,
maths) construct can only be
assessed once the linguistic
construct has been
successfully negotiated
•Loss of self-esteem & motivation for
learning
•Learners do not achieve their potential –
glass ceiling effect
•Learners leave school as unsuccessful
learners at end of Basic Education (in
some countries end of Primary Phase)
•Unskilled individuals unable to join the
workforce in turn leading to social &
economic deprivation
48