Elsevier

Learning and Instruction

Volume 10, Issue 1, February 2000, Pages 55-71
Learning and Instruction

Building models of adult second-language writing instruction

https://doi.org/10.1016/S0959-4752(99)00018-3Get rights and content

Abstract

This paper first reviews recent attempts to describe formal models of second-language writing; we then identify three major issues in need of clarification to advance model-building in this domain. We discuss these issues in respect to findings from a preliminary attempt to develop a process-product model of ESL writing instruction based on analyses of antecedent, process and outcome variables for 108 ESL learners from diverse cultural backgrounds in 6-week intensive courses at a Canadian university. Findings from multivariate, regression, correlation and exploratory factor analyses indicated that the tentative model had little explanatory power overall but pointed toward interesting interactions in achievement for (a) different aspects of L2 writing among (b) learners at different levels of L2 proficiency and with (c) differing orientations to learning L2 writing.

Section snippets

Recent approaches to model-building in educational contexts

What analytic approaches have been attempted to date? What information do they provide for building models of second-language writing instruction? Four analytic techniques have reported notable (albeit preliminary) success in developing and evaluating models of second-language writing that are relatively comprehensive and grounded in empirical data from natural educational contexts: meta-analysis, cross-sectional studies, longitudinal studies, and explanatory modeling.

Silva's (1993)

Building models of second-language writing instruction: Three key issues and examples from a preliminary attempt

The studies reviewed above highlight three major issues in need of clarification for the purposes of building models of second-language writing relevant to education:

  • 1.

    What are appropriate indicators of achievement in second-language writing?

  • 2.

    What are the principal variables that influence learning to write in a second language in an educational context? What data should be gathered to describe them?

  • 3.

    How do these variables interrelate (as processes of learning) to produce students' achievement in

Indicators of achievement in second language writing

Developing a model that tries to explain second-language writing requires a precise definition of the construct. To determine what people might learn in respect to second-language writing we have to know what changes in their writing performance are significant and valid indicators of learning (Cumming & Mellow, 1996). This is a complex issue because virtually all existing tests of second-language writing aim to group or sort learners for the purposes of placement or admissions into educational

Relevant variables

A second major issue in developing models of second-language writing instruction involves determining which variables to gather data on, which instruments to use to do this, and how to reduce the resulting data into empirical forms suitable for analyses. As noted above, process-product models conventionally distinguish (a) antecedent (or presage) variables related to students' personal characteristics and backgrounds prior to beginning an educational program and (b) process variables, including

Interrelations between variables and achievement scores

The goal of this kind of model-building research is of course to evaluate whether trends in the antecedent and process variables relate significantly to the outcome variables, in this case, students' achievement in their second-language writing. Although the project we have been describing produced numerous interesting findings, it ultimately fell short of having very much explanatory power. That is, the antecedent and process variables did not explain very much of the variance in the outcome

Multivariate analyses

Three sets of findings are worth discussing for their relevance to future model-building. The first involved multivariate analyses, which we used (a) to assess students' pre-post term achievement in the five traits of second-language writing then (b) related to the antecedent variables documented. MANOVAs comparing student's writing scores at the beginnings and ends of their six-week terms of ESL study showed statistically significant differences (N =73) for all but one of the five traits of

Exploratory factor analyses

For a second approach to analyses we used exploratory factor analyses to identify trends in four logically grouped sets of the antecedent and process variables. Although of value only for descriptive rather than explanatory purposes, these analyses produced interesting profiles of the previous and current learning experiences of these students in respect to second-language writing. Using Principal Axis Factoring and OBLIMIN rotations until satisfactory factor solutions were reached, we

Concluding remarks: Whither models?

Overall, the analyses presented here failed to produce an adequate explanatory model of the relations between adult ESL students' backgrounds, processes of learning and using classroom instruction, and their achievement in ESL writing. Perhaps this aim was overly ambitious or maybe even not feasible in the absence of an explicit, prior theoretical formulation of learning to write in a second language, given the relatively small number of students assessed and their very diverse cultural,

Acknowledgements

We thank the Social Sciences and Humanities Research Council of Canada (standard grant 410-91-0722) for supporting the research reported in this paper; Sue Elgie for advice on analyses; as well as Michael Lessard-Clouston, Jiang Li, Susanna Lo, Kara Moscoe, Hiroko Saito, Ling Shi and Sufumi So for assistance in data collection. Certain findings from this research were reported in October, 1994 under the title “Learning variables and achievement in ESL written composition” at the Second Language

References (42)

  • A. Cumming

    Expertise in evaluating second-language compositions

    Language Testing

    (1990)
  • A. Cumming

    Metalinguistic and ideational thinking in second-language composing

    Written Communication

    (1990)
  • A. Cumming

    Recent trends in research on biliteracy

  • Cumming, A. (1997). The testing of writing in second languages. In C. Clapham, Language testing and assessment (Vol 7...
  • A. Cumming

    Theoretical perspectives on writing

    Annual Review of Applied Linguistics

    (1998)
  • A. Cumming et al.

    Learning ESL literacy among Indo-Canadian women

    Language, Culture and Curriculum

    (1991)
  • A. Cumming et al.

    An investigation into the validity of written indicators of second language proficiency

  • J. Cummins

    Interdependence of first and second language proficiency in bilingual children

  • R. Ellis

    The study of second language acquisition

    (1994)
  • N. Gage et al.

    Process-product research on teaching: A review of criticisms

    Elementary School Journal

    (1989)
  • L. Hamp-Lyons

    Reconstructing “academic writing proficiency”

  • Cited by (43)

    • Effects of teacher-scaffolded and self-scaffolded corrective feedback compared to direct corrective feedback on grammatical accuracy in English L2 writing

      2019, Journal of Second Language Writing
      Citation Excerpt :

      Nonetheless, this can leave little time for other areas of study. L2 learners with weak L1 writing skills may be at a disadvantage when learning L2 writing (Cumming & Riazi, 2000). With little writing experience and little experience receiving feedback, it is possible that learners with test-centric educational backgrounds might benefit from being prompted to engage in languaging as a way of scaffolding their use of CF.

    • Effects of different language environments on Chinese graduate students' perceptions of English writing and their writing performance

      2017, System
      Citation Excerpt :

      The reliability of the questionnaire was satisfactory (r = 0.87), using Cronbach alpha. Most evidence suggests that using a single, brief composition alone provides an “inadequate representation of the domain of second-language writing and thus inadequate for research purposes” (Cumming & Riazi, 2000, pp. 60–61). Therefore, in the current study we requested the participants to complete two writing tasks: argumentative essay writing and letter writing (see Appendix B).

    • Teachers' practices in EAP writing instruction: Use of models and modeling

      2014, System
      Citation Excerpt :

      However, there is general agreement among both teachers and scholars (e.g. Swales, 1990; Turner, 2004) of the need for all stakeholders to recognise that far from just providing remedial assistance for students from non-English-speaking backgrounds, EAP instruction actually endeavors to develop knowledge and skill in a range of sophisticated literacies. Recently published scholarly texts on L2 writing (e.g. Ferris & Hedgcock, 2005; Galbraith & Rijlaarsdam, 1999; Grabe & Kaplan, 1997; Hyland, 2002, 2003; Matsuda, 2003) advise that instruction needs to blend textual, cognitive and social dimensions, and that the curriculum needs to be conceptualised less simplistically than in terms of “process” or “product” (Cumming & Riazi, 2000; Matsuda, 2003). Cummings, Erdösy, and Cumming (2006) speculate that the exact combination of curriculum components might well be determined by course goals and students' level of writing expertise, with grammar and lexis more likely to be emphasised in pre-university courses, and genre knowledge and critical thinking more prominent in in-sessional undergraduate courses.

    • Teaching and assessing academic writing via the portfolio: Benefits for learners of English as an additional language

      2011, Assessing Writing
      Citation Excerpt :

      Recent studies into the academic writing (AW) of students of English as an additional language (EAL) have focused increasingly on how learners “actually learn to write” in a second language and how teaching contributes to learning (Cumming & Riazi, 2000, p. 57).

    View all citing articles on Scopus
    View full text