Building the Ideal Relationship:

An assessment plan to test academic vocabulary and writing

Introduction

Language testing poses a moral dilemma, in that its intended purpose is to discriminate

between people. Consequently, test designers have a responsibility to be discriminating in

discrimination (Hamps-Lyon 1989, cited in Lynch 1997, p. 315) by considering a tests validity,

reliability, washback, authenticity and practicality. This paper discusses two progress tests,

namely a discrete-point vocabulary test, including selected and constructed-response tasks,

and a writing portfolio. The tests are designed for the Academic Reading and Writing (ARW)

unit, which is part of the English for Tertiary Studies (ETS) programme at UOW College. It will

be argued that this combination of objective and subjective measurements builds the ideal

relationship and thus discriminates discriminately. Firstly, the context and student cohort will

be discussed. Secondly, the ARW curriculum will be outlined. Thirdly, language constructs and

test specifications will be described. Fourthly, both tests will be evaluated in relation to the

aforementioned criteria. The assessment plan, the tests, and scoring and grading scales can

be found in the Appendices (pp. 18 - 52).

Context

UOW College, which is an integral part of University of Wollongong (UOW), provides direct

entry programmes for domestic and international students. ETS is a ten-week non-credit

carrying EGAP programme for international students planning to undertake tertiary courses

in business, medicine, law, humanities, creative arts, engineering, science, IT, or social

sciences at UOW. Traditionally, ETS programmes comprise a heterogeneous student cohort,

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 1 of 52
who not only intend to specialise in disparate disciplines upon conclusion of dissimilar

pathways, but also speak in a babel of native tongues. Students from more than thirty L1s

have participated in the programme (UOW College 2015). Future students require

intermediate English language proficiency or higher for admission, which needs to be

evidenced by appropriate results from an accepted test. This includes either the IELTS or

TOEFL standardised test, whereby students need to achieve a minimum of an IELTS 5.5 overall

band score with modest user competence in reading and writing, or 525 in the TOEFL paper

test or 195 in the computerised equivalent. Alternatively, students can undertake a

placement test or participate in a preparatory programme at UOW College.

ARW Curriculum

ETS is an integrated skills-based programme, comprising three core units: Critical Literacy;

Academic Listening and Speaking; and, Academic Reading and Writing. The ARW component

carries the majority weighting of sixty percent. Learning objectives include computer literacy;

time management; and, essay writing. The unit focusses on writing an argumentative essay,

whereby teaching and learning includes several microskills, namely rhetoric, process writing,

writing from sources, and, academic lexis.

The present paper is based on the 2003 ETS draft syllabus and proposes two alternative

formative assessments for the ARW unit. As the argumentative essay appears to be the

principal learning objective, an assessment plan that promotes beneficial washback for

writing in this genre is proposed (Appendix A, p. 18). It entails an academic vocabulary test

and a process portfolio that includes compositions, essay plans, and writing journals. Their

purpose is to identify what students have learned and still need to learn (Brown 1998, p. 15)

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 2 of 52
in regards to mastering control of the target task, i.e. an academic argumentative essay. The

assessments fulfil a triadic function, namely to offer feedback; provide scaffolded learning;

and, assess the curriculums appropriateness (Ibid.). Both tests are criterion-referenced,

whereby criteria are based on theoretical constructs and ARW learning objectives. All

objectives are addressed, except for note-taking, time management, and writing abstracts

(Appendix A, p. 18). Before test specifications can be described, underlying constructs need

to be discussed, in that tests are operational definitions of constructs (Brown &

Abeywickrama 2010, p. 33).

Language Constructs

The language model to be tested involves two distinct constructs, namely, argumentative

essay conventions, and the skills required to write in this genre. lvarez (2001, cited in

Bejarano and Chapetn 2013, p. 129) defines it as an interactive text, whereby the author

defends a perspective in order to convince or persuade. Andrews (1995, cited in Winsgate

2012, p. 146) emphasizes the genres cohesion, in that it consists of a connected series of

statements intended to establish a position. According to Toulmin, Reike and Janik (1984,

cited in Winsgate 2012, p. 146) these statements contain claims and reasons. Furthermore, a

carefully constructed argumentation includes qualifiers, i.e. a form of hedging, and rebuttals

of counterarguments (Toulmin 1958, cited in Liu & Stapleton 2014, p. 118). In sum, the

academic argumentative essay construct is defined as a text type that aims to convince or

persuade, which is achieved through interrelated moves, including claims, reasons, qualifiers

and rebuttals. This construct formed the premise for test development and marking rubric

design.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 3 of 52
Additionally, the design process was informed by a pragmatic amalgam of writing constructs.

Firstly, the test is informed by genre theory. Practicality considerations preclude a dedicated

genre approach, in that the context comprises a student cohort who plan to specialise in

diverse disciplines, whereas argumentative texts are highly discipline-specific (Winsgate

2012, p. 147). However, Swaless move structure concept has been incorporated. This is

evidenced in the writing journals, which stipulate that students elaborate on moves and

communicative purposes incorporated in compositions (Appendix C, pp. 45 - 47). Secondly,

the tests incorporate process writing concepts, which emphasize intrapersonal cognitive

skills. For instance, the first test includes an editing task (Appendix B, pp. 34 - 43), whereas

the second test requires students to develop essay plans (Appendix C, pp. 45 - 47). Thirdly,

writing from sources, which is a fundamental aspect of real-life scholarly writing (Gebril 2009,

p. 508), has been included as a criterion in the writing assignments, which is demonstrated in

the marking rubric (Appendix D, p. 49). Finally, the tests are informed by a product-oriented

approach, which emphasizes textual conventions such as paragraph structure, syntax and

lexis. This construct is evidenced in the first test, which measures academic vocabulary

(Appendix B, pp. 19 - 43), and elaborated in the second test, whereby compositions are scored

according to paragraph cohesion; sentence structure variation; and, appropriate TL use

(Appendix D, pp. 48 - 51). Thus, the academic writing construct is defined as an amalgamation

of cognitive skills, in order to communicate effectively, according to the linguistic conventions

of the TL domain.

Vocabulary Test Specifications

The first measurement comprises a triadic discrete-point test of academic vocabulary,

whereby each component measures distinct vocabulary knowledge and abilities (Appendix B,

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 4 of 52
pp. 19 - 44). The tasks were sequenced in order to increase the skills complexity level,

progressing from recognition to production skills. The test measures high-frequency academic

vocabulary, which was selected since many EFL students perceive academic lexis as

particularly challenging (Li & Pemberton 1994, cited in Hyland & Tse 2007, p. 326). Test items

originate from Coxheads (2000) Academic Word List (AWL), which consists of 570 word

families, including 3,112 individual items (Hyland & Tse 2007, p. 327). The selection

predominantly consists of the most frequently occurring members of AWL word families,

although occasionally less frequently occurring words had to be selected for practical reasons.

Furthermore, the tests third task, which comprises authentic texts, required occasional

deviations from the AWL framework.

The first part of the vocabulary test comprises a multiple-choice (MC) instrument, which

measures the ability to recognise denotations of Latinate content words (Appendix B, pp. 20

29). All test items include four options to reduce the effect of guessing (Hughes 2003, p. 77),

which are furthermore homogenous in length and content, and contain plausible distractors.

For example, Item 8 tests recognition, or recall, of the word interpretation, whereby the

distractors include translation, inducement, and solution. The items were designed

according to Bothels (2001) guidelines, whereby the aforementioned item is realised as

follows:

8. This paper offers an alternative interpretation of Manchesters football history, arguing that it was a

minor form of football in a city dominated by a rugby code.

What is the purpose of a paper that offers an interpretation?

a. to offer an explanation

b. to offer a translation

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 5 of 52
 c. to offer an inducement

d. to offer a solution

The item adheres to a memory-plus-application-format (Bothell 2001, pp. 1 2), whereby

the prompt is framed as a question, preceded by an authentic reading passage containing the

token in question. This format encourages students to recall principles, rules or facts in a real-

life context thus emphasizing higher-level thinking skills (Ibid.). All reading passages were

selected from authentic articles, sourced from Google Scholar and UOW library databases. A

few sentences were adapted from Swales and Feaks (2012) Academic Writing for Graduate

Students.

The second part comprises a single-word gap filling test, which focusses on accuracy and

measures productive vocabulary knowledge of collocations (Appendix B, pp. 30 - 33). Previous

research found that even advanced NNSs have great difficulty with native-like collocations

(Ellis, Simpson-Vlach & Maynard 2008, p. 378); however, fluent and accurate usage of

formulaic language signals competence to a given discourse community (Hyland 2008, p. 42).

The test contains discrete sentences, whereby students need to fill in the correct preposition

following a verb or noun provided. All prompts were selected from authentic texts according

to the procedure described above.

The third part entails a MC editing test, which focusses on pragmatic meanings and measures

both receptive and productive abilities (Appendix B, pp. 34 - 43). This task type was selected

in order to promote proofreading skills (Brown & Abeywickrama 2010, p. 247) in preparation

of the skills required for the second assessment, i.e. the portfolio. The task contains two

separate texts, which require students to proofread each sentence; identify inappropriate

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 6 of 52
vocabulary; and, provide an accurate and appropriate alternative. It measures knowledge of

lexical phrases, discourse markers and content words. The two texts were adapted from

authentic academic articles on general knowledge topics, i.e. the resources curse and

modern slavery. As such, they do not require specialist knowledge thus cultivating fairness

for test-takers (Bachman & Palmer 1996, pp. 106 107)

The vocabulary test will be scored according to an absolute grading scale, whereby one point

is awarded for each question answered correctly. The MC test is scored dichotomously and

the gap-filling task follows an exact word-scoring procedure, whereby a point is awarded for

the correct preposition only. The MC editing task is scored according to appropriate word-

scoring and partial credit-scoring procedures. Credit is awarded for grammatically correct and

appropriate language choices whereby acceptable answers do not necessarily need to

originate from the AWL. Half a point is awarded for selecting the key and another for

providing a suitable alternative. The entire test contains one hundred items and aggregate

scores map onto the grading scale (Appendix E, p. 52) as provided in the ETS syllabus, e.g. a

total score of sixty-eight points translates into a credit grade.

Writing Test Specifications

The second test is a performance-based assessment, namely a process portfolio, which

includes compositions, essay plans and writing journals (Appendix C, p. 45 - 47). According to

Brown and Abeywickrama (2012, p. 131), portfolio development has various pedagogical

benefits, including individualisation of learning; encouragement of critical thinking and

revision processes; and, promotion of intrinsic motivation, responsibility and ownership.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 7 of 52
The portfolio is based on three separate themes, or topics, whereby each requires a

composition, an accompanying writing journal, and an essay plan. Students do not have a

choice of topics and need to undertake all tasks. Each composition focusses on one aspect of

the essay macrostructure, for instance, the first composition task requires students to write

an introduction; the second requires body paragraphs; and, the third requires a conclusion.

Additionally, three writing journals are required (Appendix C, pp. 45 - 47), whereby content is

restricted to thinking and writing processes related to each specific composition, with the

purpose of raising awareness of logical argumentation, communicative purposes and

language choices. The essay plans are not discussed in this paper, as they are intended to be

integrated into daily classroom activities.

The test contains three separate prompts, which were designed according to Kroll and Reids

(1994) guidelines and include contemporary social issues, namely genetically-modified food;

constant public surveillance; and, human cloning. For example, Task Three (Appendix C, p.

47) prompts students to consider the following:

Topic: Twenty years ago, the first mammal, Dolly the Sheep, was cloned successfully.

To date, no human clone has been born. Nonetheless, the topic is a rich source for

media and fiction narratives alike. What opportunities, or threats, do you think

human cloning presents? Relate the response to your intended field of study.

The framed prompt above does neither require specific cultural schemata nor discipline-

specific knowledge. Furthermore, it allows multiple interpretations and content is based on a

body of knowledge that is equally accessible to all students (Horowitz 1991, cited in Kroll &

Reid 1994, p. 235) thus avoiding cultural biases.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 8 of 52
The compositions will be scored according to an analytical scale (Appendix D, p. 48 -51), based

on the theoretical constructs outlined above. Criteria included relate to logical

argumentation; textual coherence and cohesion; language choices; source writing; and,

presentation. The performance levels include four gradations, ranging from excellent,

good, pass to unsatisfactory. The rubric is scaled, with points assigned for each level of

performance, whereby the maximum score amounts to one hundred points, so that

aggregate scores translate directly into a grade as per the ETS syllabus scale (Appendix E, p.

52). The scales purpose is to provide individualised feedback and pinpoint the microskills not

mastered yet (Bloom et al. 1971, cited in Perkins 1983, p. 656). The writing journal will not be

scored or graded, but instead functions as an additional instrument for providing

individualised feedback in that it provides insights into students language awareness and

thinking processes. Process portfolios and writing journals will be discussed during one-on-

one conferences. When compositions have been revised, they become part of the final

presentation portfolio.

Reliability

The vocabulary test represents a relatively reliable instrument. Firstly, it includes a substantial

number of items, based on the premise that the more items there are on a test, the more

reliable it will be (Hughes 2003, p. 44). In addition, the first sixty items each provide a fresh

start thus further improving test reliability. Secondly, it entails a comparatively reliable

scoring procedure. The first two tasks represent objective tests, in that there is only one

correct answer for each question. The MC editing task is more subjective, in that multiple

alternatives might be possible, and scoring requires a judgement call on behalf of the raters.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 9 of 52
Rater reliability could be improved by introducing a second rater for this specific task (Hughes

2003, p. 50).

Several safeguards were implemented to increase the reliability of the writing portfolio.

Firstly, it comprises three independent, untimed compositions thus preventing the snap-shot

approach, which creates unreliable and unrepresentative impressions (Hamp-Lyons & Kroll

1996, p. 53). Secondly, students do not have a choice of topics, in that too much freedom in

topics is likely to have a depressing effect on test reliability (Hughes 2003, p. 45). Thirdly,

scoring should be conducted according to a detailed marking rubric, which may increase intra-

rater reliability, prevent accidental construct under-presentation (Hughes 2003, p. 102), and

improve trustworthiness and auditability (Brown & Hudson 1998, p. 655). In sum, several

steps were taken to improve the writing tests reliability.

Nonetheless, both tests should be trialled to assess other plausible threats to test reliability.

Firstly, the vocabulary test is relatively lengthy. Although it is not timed, it needs to be finished

within one period. Consequently, test takers could become fatigued, which may cause

inaccurate test results (Brown & Abeywickrama 2012, p. 29). Perhaps it needs to be

administered across several periods to address this possible threat. Secondly, MC items need

to be tested to determine distractor efficiency, item facility and item discrimination. Thirdly,

the composition prompts need to be appraised to establish whether content is indeed

unbiased and appropriate for this specific context (Kroll & Reid 1994, p. 241).

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 10 of 52
Validity

The portfolio demonstrates several types of validity. Firstly, it demonstrates face validity, in

that it entails a direct test of writing skills. Secondly, it demonstrates content validity, in that

it addresses the majority of ARW learning objectives and requires performance of the target

task, namely an argumentative essay. Finally, it demonstrates construct validity, in that it is

premised on research-based frameworks for argumentation and academic writing.

Contrarily, the vocabulary test may have construct validity issues in that the existence of a

common core of academic vocabulary is contested. Hyland and Tse (2007, p. 248) argue that

the AWL has variable usefulness for various disciplines since many items are

underrepresented in some fields. The AWL has most utility for IT students and least for those

studying biology (Ibid.). This suggests that caution may need to be exercised in

implementation. Notwithstanding, the AWL could be used as a threshold concept (Clarke &

Hernandez 2011, p. 66) to develop discipline-specific language awareness through in-class

readings and reflective activities. Furthermore, aside from EAP courses, students have limited

opportunities for common core vocabulary development, in that sub-technical lexis is unlikely

to be taught by content teachers (Flowerdew 1993, cited in Hyland & Tse 2007, p. 236). Thus,

a vocabulary test may have a signficant role to play by motivating students to study and

review (Bachman 1990, p. 22) the building blocks of academic language.

Authenticity

The process portfolio could be considered an authentic assessment, in that there is a strong

correspondence between the test and the skills required in real-life academic contexts.

Arguing a case is one of the most frequent and important types of assessment tasks set at

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 11 of 52
university (Lee 2008, p. 240), whereby logical argumentation is considered a key writing skill

(Lee & Street 1998, cited in Winsgate 2012, p. 145). Academic essays must both carry

appropriate authority and engage readers in ways they are likely to find credible and

persuasive (Hyland 2002, p. 215). However, many students demonstrate language lacks in

discussing, arguing and evaluating competently, logically and persuasively (Lee 2008, p. 240;

Winsgate 2012, p. 145). The process portfolio provides step-by-step support in constructing

an argumentative essay, from planning and revising, to explaining and rebutting. Thus, this

assessment addresses both potential language lacks and authentic language demands.

However, the authenticity of the vocabulary test is more tenuous. Brown and Hudson (1998,

p. 659) state that this test type lacks authenticity, in that real-life language is not multiple-

choice. Bothell (2001, p. 4) emphasizes that MC questions should be avoided when other

item types are more appropriate. Nonetheless, constructed-responses are the only test items

that can address avoidance strategies, whereby NNSs evade using challenging lexis (Brown

2000, p. 149). Although other constructed-response tests could likewise address the

undesirable strategy, the MC test was selected in that it is more reliable than a true-false test

(Brown & Hudson 1998, p. 659) and may have more construct validity than a matching test,

in that the latter can become more of a puzzle-solving process than a genuine test of

linguistic comprehension (Brown & Abeywickrama 2012, p. 218).

Washback

Both tests could potentially promote beneficial washback effects. Although the vocabulary

assessment may seem a bolt-on test to discriminate between those who have memorised

sufficient lexis and those who have not, its inclusion is based on sound theoretical principles.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 12 of 52
Mitchell, Miles and Marsden (2013, p. 143) reiterate research that found that explicit

vocabulary learning may lead to proceduralisation of this knowledge. Thus, it is hoped that

vocabulary learning required for the first test, leads to fluent use in future writing tasks.

Beneficial washback is promoted by turning the traditional test into a feedback opportunity,

which aims to identify students strengths and weakness and provide suggestions for further

development (Appendix B, p. 44).

The portfolio assessment could promote beneficial washback by integrating similar tasks into

daily teaching and learning. Ideally, test content will be taught according to a genre-based

reading-writing paradigm, whereby students explore and analyse argumentative essays in

class perhaps with the use of the AWL - and sequentially apply and transform new

understandings during classroom writing activities. As students are required to relate their

compositions to own disciplines (Appendix C, p. 45- 47), it would be most effective if they

source own discipline-specific readings. Additionally, students could work on the portfolio in

class, whereby learning to write is scaffolded by peer reviews and teacher feedback.

Practicality

The two tests balance each other in regards to practicality. The vocabulary test could be re-

used; implementation is straightforward; and, scoring is comparatively effortless, in that sixty

per cent of the test is scored objectively. Practicality could be further increased by

computerising the test, whereby current items form the basis for an item bank. By contrast,

performance based assessments such as portfolios are relatively time-consuming and

potentially costly to administer, in that colleagues may require training, rating sessions may

need to be conducted (Brown & Hudson 1998, p. 662), and, sufficient time needs to be

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 13 of 52
allocated for student conferences (Brown & Abeywickrama 2012, p. 134). Consequently,

portfolios and conferences have a low practicality rating (Ibid.).

Conclusion

The two tests were carefully constructed by considering reliability, validity, authenticity,

washback and practicality. Reliability was improved by adding objectively-scored tasks and

developing a detailed analytical scoring rubric. Validity was addressed by designing the

assessments according to relevant theoretical constructs and ARW learning objectives.

Authenticity criteria were met by aligning the task with real-life language demands in

university settings. Washback was addressed with the suggestion of integrating assessment

tasks into daily learning activities. Finally, time and effort required for portfolio assessment

and conferencing is balanced by the practicality of the vocabulary test. In conclusion,

favourable conditions were created to foster the ideal relationship between tests as well as

between test and context.

References

Bachman, LF 1990, Measurement, Fundamental Considerations in Language Testing, Oxford University Press,

Oxford, UK, pp. 18 52.

Bachman, LF & Palmer, A 1996, Describing, identifying, and defining: test purposes, tasks in the TLU domain,

characteristics of test-takers, and the construct to be measured, Language Testing in Practice:

Designing and Developing Useful Language Tests, Oxford University Press, Oxford, UK, pp. 95 132.

Bejarano, PAC, Chapetn, CM 2013, The Role of Genre-Based Activities in the Writing of Argumentative Essays

in EFL, Profile, vol. 15, no. 2, pp. 127 147.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 14 of 52
Bothell, TW 2001, 14 Rules for Writing Multiple-Choice Questions, 2001 Annual University Conference, Brigham

Young University, pp. 1 5.

Brown, HD 2000, Principles of language learning and teaching, 5th edn, Pearson Longman, White Plains, NY.
Brown, HD & Abeywickrama, P 2010, Language Assessment: Principles and Classroom Practice, 2nd edn, Pearson

Longman, White Plains, NY.

Brown, J 1998, Language testing: purposes, effects, options, and constraints, TESOLANZ Journal, vol. 6, pp. 13

30.

Brown, JD & Hudson, T 1998, The alternatives in language assessment, TESOL Quarterly, vol. 32, no. 4, pp. 653

675.

Clarke, IL & Hernandez, A 2011, Genre Awareness, Academic Argument, and Transferability, WAC Journal, vol.

22, pp. 65 78.

Coxhead, A 2000, A New Academic Word List, TESOL Quarterly, vol. 34, no. 2, pp. 213 238.

Ellis, NC, Simpson-Vlach, R & Maynard, C 2008, Formulaic Language in Native and Second Language Speakers:

Psycholinguistics, Corpus Linguistics, and TESOL, TESOL Quarterly, vol. 42, no. 3, pp. 375 396.

Gebril, A 2009, Score generalizability of academic writing tasks: Does one test method fit it all? Language

Testing, vol. 26, no. 4, pp. 507 531.

Hamp-Lyons, L & Kroll, B 1996, Issues in ESL Writing Assessment: An Overview, College ESL, vol. 6, no. 1, pp. 52

72.

Hughes, A 2003, Testing for Language Teachers, 2nd edn, Cambridge University Press, Cambridge.

Hyland, K 2002, Directives: Argument and Engagement in Academic Writing, Applied Linguistics, vol. 23, no. 2,

pp. 215 239.

Hyland, K 2008, Academic clusters: text patterning in published and postgraduate writing, International Journal

of Applied Linguistics, vol. 18, no. 1, pp. 41 62.

Hyland, K & Tse, P 2007, Is There an Academic Vocabulary?, TESOL Quarterly, vol. 41, no. 2, pp. 235 253.

Kroll, B & Reid, J 1994, Guidelines for Designing Writing Prompts: Clarifications, Caveats, and Cautions, Journal

of Second Language Writing, vol. 3, no. 3, pp. 231 255.

Lee, SH 2008, An integrative framework for the analyses of argumentative/persuasive essays from an

interpersonal perspective, Text and Talk, vol. 28, no. 2, pp. 239 270.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 15 of 52
Liu, F & Stapleton, P 2014, Counterargumentation and cultivation of critical thinking in argumentative writing:

Investigating wash-back from a high-stakes test, System, vol. 45, pp. 117 128.

Lynch, B 1997, In search of the ethical test, Language Testing, vol. 14, no. 3, pp. 315 327.

Mitchell, R, Myles, F & Marsden, E 2013, Second Language Learning Theories, 3rd edn, Routledge, London.
Perkins, K 1983, On the Use of Composition Scoring Techniques, Objective Measures, and Objective Tests to

Evaluate ESL Writing Ability, TESOL Quarterly, vol. 17, no. 4, pp. 651 671.

Swales, JM & Feak, CB 2012, Academic Writing for Graduate Students, University of Michigan Press.

UOW College, Diversity and Equity, viewed 10 February 2015, http://www.uowcollege.edu.au/

about/diversity-equity/index.html.

Winsgate, U 2012, Argument! helping students understand what essay writing is about, Journal of English for

Academic Purposes, vol. 11, no. 2, pp. 145 154.

Bianca van de Water 5037712 EDGT983 Assessing & Evaluating Rationale Page 16 of 52