(2) Isolating Writing
It is notoriously difficult to isolate and assess only writing. It is almost
impossible to separate it from skills such as listening and reading.
According to Brown and Abeywickrama (2019), “the assessment of writing
implies by definition the assessment of reading as well” since “one cannot
write a sentence in a language without being able to read the just written
sentence” (p. 228). What is more, for the students to write texts following
the exam requirement/rubrics, they need to understand the prompts given to
them; and the prompts can be either written (i.e., reading) or oral (i.e.,
listening). Therefore, writing test designers should keep in mind that
assessing writing is almost always integrated with another skill.
(3) The Aim of the Test
When preparing a writing test, the assessment objectives should be
clear. All of the aims should be represented together with their weights (i.e.,
how many points will to be allocated to each of them) in the exam rubrics.
Criteria such as handwriting, spelling, grammaticality of the sentences,
vocabulary richness, paragraph constructions, originality, logical
Language Assessment - Theory with Practice
182
development of the main idea could be part of the rubrics. While designing
the rubrics, test writers should be clear about the following points:
(i) How significant each of these constructs is in that specific exam and
for that writing task?
(ii) Are the tasks included in the exam allowing students to demonstrate
their real knowledge and skills?
3.1. Types of Writing Performance
In the past, the more popular way of assessing writing was ‘implicit
testing’ (i.e., a technique where the abilities and sub-skills that are believed
to underlie the skill of writing are tested, e.g., instead of asking test-takers to
write a paragraph, tasks eliciting information related to their knowledge of
spelling, capitalisation, punctuation, grammar, vocabulary use, discourse
etc. are created). Nowadays, however, it is widely accepted that “the best
way to test people’s writing ability is to get them to write” (Hughes, 2003,
p. 83) (i.e., direct testing). To ensure that not any other skill or type of
knowledge is assessed but the ability to write in the foreign language, test
designers have to observe the following rules (Hughes, 2003; Rakedzon &
Baram-Tsabari, 2017; Tsagari & Banerjee, 2016; Tsagari & Vogt, 2017):
(i) Set writing tasks that are proper representatives of the population of
tasks those students are expected to be able to perform;
(ii) Select or create tasks that elicit valid samples of writing (i.e., which
truly represent the students’ ability);
(iii) Train test evaluators and carefully prepare guidelines, rubrics and
other assessment tools to ensure that the samples of writing can and
will be scored validly and reliably.
The writing performance of language learners is usually classified into
four categories (i.e., imitative, intensive, responsive, extensive) depending
on how long and how creative the written language test takers are expected
to produce. The categories were introduced by Brown (2004), but they have
been widely accepted and applied for the classification of assessment
writing tasks (e.g., Heaton, 1990; Hughes, 2003). In this chapter, we follow
Brown’s (2004) categorisation of writing tasks for ease of comparison with
Assessment of Language Skills: Productive Skills
183
other resources. However, differently from some previously published work
(e.g., Brown & Abeywickrama, 2019), we try to show how each of them can
be used not as summative but as a formative assessment tool.
Dostları ilə paylaş: |