![]() |
2019年度 異文化言語教育評価論 |
Chapter 4
Investigating Stakeholder
Perspectives on Interact (4.4 to 4.8, pp.87-98)
This part evaluates the
fundamental principles of interact against the theoretical framework for
usefulness and presents the methodology for the 2-year study that has sought
stakeholder views (both teachers and students) during the initial phases of the
implementation of interact (2012-2013).
According
to Bachman and Palmer’s six qualities of test usefulness (1996), it seems that interact
is a particularly valid and useful measure of spoken communicative proficiency:
Construct validity |
Assessments => a
more comprehensive range of facets of spoken communicative proficiency (≠ single-candidate
interview test) |
Reliability |
-
Quantity:
collection of evidence over time -
Quality:
clear instructions about preparation (eg. no pre-scripting and rote-learning)
ensure candidate’s own work, IC -
Grading:
requirement for clear identification of candidates for scoring and external
moderation purposes |
Interactiveness |
Opportunities for
feedback Move away from the
requirement to use specific grammar structures |
Impact |
|
Practicality |
Reasonably short
evidence of interaction for assessment purposes Evidence assessed
holistically on one occasion |
Authenticity |
Both situationally and
interactionally authentic scenarios Requirement for
fluency and spontaneity (≠ converse) |
Counter-arguments
to a claim to usefulness:
Construct validity & Reliability |
Collection of evidence
over time -
Different
levels of proficiency throughout the year -
What
evidence may be included to demonstrate proficiency? -
When
should the evidence be collected? |
Feedback and
feedforward -
When
is the work clearly the candidate’s own? -
When
is that work unfairly influenced by feedback? |
|
Interactiveness |
Informing candidates
when an assessment is to take place focuses the attention on the interaction
as an assessment |
Impact |
|
Practicality |
Teacher’s early
reactionary feedback to the proposal to introduce interact |
Authenticity |
Spontaneous
interactions recorded beyond the classroom -
What
evidence does constitute a genuinely authentic interaction? |
Because theoretically, interact
has both good points and bad points, there is a need to see what is happening
with interact in practice.
The
study investigates stakeholders’ perspectives on the assessment reform during
the period of its initial roll-out (2012-2013).
Research
questions:
-
What
are teachers and students making of the innovation?
-
What
is working, what is not working, what could work better?
-
What
are the implications, both for on-going classroom practice and for on-going
evaluation of the assessment?
Why
investigate perspectives at an early stage of the implementation? To focus on
the comparison of interact with converse.
Data
collected during the 2 stages of the empirical study:
1.
2012
anonymous teacher survey
2.
2012
and 2013 interviews with teachers
3.
2012
and 2013 anonymous student surveys
Target
of the survey = teachers of the 5 principal FLs taught in New Zealand (Chinese,
French, German, Japanese and Spanish)
Construct
measured by the survey = ‘perceived usefulness’ of both interact and converse
Perceived
usefulness measured according to 4 sub-constructs of the usefulness construct:
1.
Perceived
validity and reliability
2.
Perceived
authenticity and interactiveness
3.
Perceived
impact
4.
Perceived
practicality
The survey
Section
I Some close-ended questions about the
perceived usefulness of interact.
Prompting
more precise and nuance responses (≠Likert scale):
Section II 4
open-ended questions:
1.
Comment
on the perceived advantages of interact in comparison with converse
2.
Comment
on the perceived disadvantages of interact in comparison with converse
3.
If
you are using interact, describe briefly your experiences with
introducing the new assessment at levels 1 and/or 2. If you are not using interact,
explain briefly why.
4.
How
might interact be improved?
10
teachers => comment on their understanding of the statements + how long did
it take to respond? Conclusion: the survey can be regarded as a reliable
measure of teachers’ perceptions of the usefulness or fitness for purpose of converse/interact
as assessments.
What
did the pilot survey tell us?
-
Internal
consistency of the survey is higher for perceived validity and reliability, but
lower for perceived impact, and considerably lower consistency for perceived
practicality
-
The
fact that students might enjoy the opportunities to speak does not necessarily
make the experience less stressful
-
The
survey could be completed rather quickly
Interviews
(November – December 2012) with 14 teachers using interact about:
-
Interviewees’
understanding of the main purposes of interact, and opinions about interact
-
Comparisons
and contrasts between interact and converse
-
Advantages
and disadvantages of interact in comparison with converse
-
Experiences
and challenges with the implementation of interact
-
Types
of assessment tasks used and perceived student reception of these
-
Advice
for others about how to implement interact successfully
Three
key domains of concern arose from the survey and interview data:
-
The
importance of the task
-
The
concept of ‘spontaneous and unrehearsed’ (a focus on fluency)
-
A
de-emphasis on grammar (the place of accuracy)
Interviews
(November – December 2013) with 10 teachers who had not taken part in the Stage
I interview + 3 teachers who had already participated in stage I interview.
Focus
on the 3 key issues of concern from Stage I interviews at interact level
3.
2
student surveys:
-
Year
13 students who had taken level 3 converse in 2012 (last year of
operation)
-
Year
13 students who had taken level 3 interact in 2013 (first year of
operation)
Similar
as teacher surveys but simplified and focus on students’ perception about interact
rather than comparison interact/converse.
Analyses
of students’ responses focused on the 3 key issues identified from the teacher
surveys from Stage I.
This
book proposes two contrastive means of assessing spoken communicative
proficiency: which of the 2 assessment formats realized in converse and interact
better reflects assessments of spoken communicative proficiency that are valid,
useful and fit for purpose?
This
book is an attempt to address this issue by taking account of stakeholder
perspectives in a study that is largely qualitative (independent and
complementary data sources such as surveys and interviews), but also
quantitative.
ð The study is robust, and
its findings will be discussed in the remaining chapters.
Discussion
questions
1.
What
kind of interact assessment can lay the focus on fluency? Provide
examples of appropriate tasks for interact and explain why you think
they are appropriate.
Where should be the place of accuracy (or grammar) in interac