Scaffolding Summary Writing through the ‘Reading to Learn’ Pedagogy

Introduction

The past decades have witnessed widened student participation and global mobility of foreign students in higher education. The student cohort in higher education, especially in developed countries such as the United Kingdom and Australia, has become diversified to include first-in-family, mature students, students who use English as an additional language (EAL) and students who have been admitted from mixed pathways programmes (Glew et al., 2019). Due to increased access, universities offer numerous programmes to support the retention and success of the students admitted, for example, peer mentoring and orientation/transitioning programmes (Eather et al., 2022). Underpinning these support programmes is the requirement that students must adhere to rigorous academic and professional standards, and academic literacy is critical if students are to meet these standards.

However, despite the provision of these various support programmes, ample literature has documented that university students struggle to meet the academic literacy demands of set tasks. For decades, research (e.g. Holschuh, 2019; Lea and Street, 1998) has pointed out that academic staff complain about university students’ lack of ability to read and write at the appropriate level of proficiency. This lack of ability partly results from the varied and complex academic literacy demands (Holschuh, 2019). Holschuh (2019: 600) states that academic literacy tasks are ‘typically multidimensional, developmental, and goal focused’, require ‘greater discipline-specific literacy demands’ and need a familiarity with sociocultural contexts. Lea and Street (1998) explain that forms of writing required in different disciplines, such as the use of an impersonal style, can expose the students to affective and identity conflicts.

Among the academic literacy demands required of university students, summary writing has long been essential. University students need to read and extract meaning from academic texts to demonstrate understanding in their written assignments. Summary writing reflects this reading-to-writing process, in which students understand the key messages in a source text, take note of them and reproduce them in a condensed written version. Kirkland and Saunders (1991) state that this process involves reading-to-writing activity that is highly complex and recursive, and which can cognitively overload students. Research, for example, Ivanova (2020), found that university students experience various difficulties in summary writing, such as understanding the gist of the text or the interrelation of ideas/themes, referencing, paraphrasing, reducing the text and sentence making.

These difficulties are compounded for EAL students, especially those who come from overseas. International EAL students have greater difficulties in transitioning to a different environment and managing their university study and life (Taylor and Ali, 2017). Language difficulties and cultural differences in education have a direct and dramatic impact on students’ academic performance (Sawir, 2005; Taylor and Ali, 2017). Zhang and Mi (2010: 382) discussed the difficulties that international EAL students experienced in reading and writing at Australian universities. A student, like many others who participated in their research, commented, ‘The key difficulty in the use of English is my inaccurate expression and my inability to express the most crucial meaning precisely … The longer I am here, the more aware I become of the seriousness of my reading and writing problems.’

Recognising the need to support EAL students, in the past decade, the research site for this current study has offered a credit-bearing foundational unit to develop advanced academic English for successive cohorts of students enrolled in a Bachelor of Business degree programme. The difficulties experienced by students undertaking this unit over the years resemble those of EAL students that have been discussed in the literature. Therefore, the lecturers who deliver this unit have sought to find a teaching method that encompasses the reading-to-writing process that university students undertake to complete their written assignments. Consultation of the literature on summary writing reveals that many studies have investigated the features of language and strategies necessary for the reading-to-writing process (e.g. Chang and Lin, 2017; Chew et al., 2020; Segev-Miller, 2004). For example, based on Kintsch and van Dijk (1978), Chew et al. (2020) developed and assessed an online tool to assist students with summary writing that incorporated tasks to activate prior knowledge and instruction in strategies for summarising texts,. However, in alignment with Hood (2008), the authors believe that the linguistic changes necessary from source text to summary is often intuited in these studies and should be made more explicit to students due to the complexity of the task. Hood (2008: 361) suggests that the R2L pedagogy can ‘explicitly articulate, model and practise with students the ways in which content meanings can be generalised and abstracted in the process of summarising’.

Reading to Learn

The R2L methodology emerged from the genre-based approach to literacy pedagogy developed by the Sydney School and is underpinned by systemic functional linguistic theory (Rose, 2018; Rose and Martin, 2012). It is delivered over three cycles of support: preparing for reading; detailed reading; and sentence making. The cycles move from a focus on the whole text to a focus on shorter sections of the text and on sentences. Each cycle guides students through tasks starting with greater teacher control and working towards tasks in which students assume greater control.

The first cycle, that is, preparing for reading, involves an introduction to the genre and field of the text. This stage helps to reduce the cognitive load for students and assists them in gaining an overall understanding of the text. The teacher first previews the topic and assesses how the stages of the text unfold. With factual texts, as in the current study, the teacher may follow the preview by introducing and reading each paragraph and highlighting the main points and language features of the genre (De Silva Joyce et al., 2008; Rose, 2018). Preparing students for reading in this way allows the teacher to embed language learning within the content of the discipline (Rose, 2018).

In the second cycle, that is, detailed reading, the teacher reads through shorter extracts from the text sentence by sentence, focusing on language details such as lexical meanings and how cohesion is achieved across sentences. Students are guided to assign meanings to words in the sentences that the teacher can then elaborate on. Then, with guidance from the teacher, students jointly plan and write a new text based on the original extract using their notes or highlighted text. Finally, students rewrite or summarise the text individually, again based on the original text, before writing a new text, drawing on what they have learned in the preceding lessons (De Silva Joyce et al., 2008; Rose, 2018).

The third cycle is sentence making, which focuses on sentence grammar, spelling and punctuation. The teacher writes selected sentences from the text onto strips of paper, reads them with the students and then guides them in cutting up the sentences. The students then reconstruct the sentences to help them recognise features of grammar and word order in various sentence types. There is also a focus on spelling in this third cycle if necessary (De Silva Joyce et al., 2008; Rose, 2018).

The R2L methodology is flexible in that teachers can tailor lessons to the needs of the students and the curriculum. For more advanced students, a greater focus on the genre and register of the text may be more appropriate for addressing their needs, whereas for students who require greater support, it may be more suitable to select stages from each cycle, as needed (Kartika-Ningsih, 2019). For example, De Silva Joyce et al. (2008) used a six-step cycle in their project using the R2L pedagogy in an adult literacy context, the six steps being as follows: preparing for reading; detailed reading; preparing for writing; joint rewriting; individual rewriting; and independent writing. An additional cycle of sentence writing, sentence making and spelling can also precede the preparing for writing stage in the cycle if necessary.

Although originally developed to address inequalities in literacy outcomes for Indigenous students in Australia (e.g. Koop and Rose, 2008; Rose, 1999), the R2L pedagogy has been adopted in mainstream primary and secondary schooling in Australia (e.g. Acevedo and Rose, 2007; Rose and Acevedo, 2006) and overseas (e.g. Becerra et al., 2020), and in tertiary and adult education both in Australia (e.g. de Silva Joyce et al., 2008; Rose et al., 2003) and internationally (e.g. Millin and Millin, 2014; Ramirez, 2020). The close interaction between teacher and students, particularly in the earlier phases of the pedagogy, makes it ideal for teaching in face-to face settings. However, with the onset of the COVID-19 pandemic in early 2020, classes with successive cohorts had to move fully online. This may have resulted in challenges with regard to managing the very close interactions between teacher and students, and these needed to be addressed during the online delivery (2020–2022). As far as we know, no other research paper has addressed the effectiveness of R2L in the context of online teaching.

Research Question

This study aims to explore whether and how the participating students would improve their summary writing ability if given R2L pedagogy instruction in the context of online teaching and intends to address the following research question: How effective would online R2L instruction be in helping students with summary writing?

Method

Results

This study involved students who were enrolled in four iterations of a foundational unit offered by an Australian university to examine whether the R2L pedagogy instruction would have a positive impact on their summary writing ability. The examination involved the comparison of the pre- and post-test results, adopting either the paired-sample t-test or the Wilcoxon signed ranked test. The paired-sample t-test requires normal distribution of the data collected, whereas the Wilcoxon signed ranked test is applied to the skewed data set. Consequently, the paired-sample t-test was performed for the examination of the content dimension, whereas the Wilcoxon signed ranked test was performed for the examination of the other two dimensions. For each test, the significance value (two-sided) p was set to be equivalent to or less than .05. Table 1 presents the descriptive statistics of the pre- and post-test scores for the three dimensions of summary writing that were examined and the p-value results that were achieved via the analyses.

OPEN IN VIEWER

Table 1.

Average Scores and p-Values.

n = 45	Pre-test		Post-test	t-test
	M	SD	M SD	p ≤ .05
Content	2.47	.83	2.86 1.14	.019
Sentences	1.12	.34	1.32 .38	.012
Cohesion and vocabulary	.57	.19	.61 .17	.14

M: mean; SD: standard deviation.

With regard to the impact of the R2L pedagogy on the participating students’ ability to comprehend and summarise the key ideas of a source text, the paired-sample t-test showed a substantially significant improvement in test results from the pre-test (M = 2.47, SD = .83) to the post-test (M = 2.86, SD = 1.14), t (44) = -2.43, p = .024. Cohen's d (1.07) indicated a large effect size. For example, Student 06 (S06) in the academic year 2021, one of the majority of students who scored within 1 SD to the mean in the pre-test, did not fully identify the topic/thesis or outline the main points in a logical manner. The topic of the article used for the pre-test was the contribution of women to the economies of Northeast and Southeast Asia, and the thesis was that to maintain growth in prosperity and economic development, East Asian countries will have to ensure that women play a greater role in the workforce. Instead, S06 wrote: ‘This text is talk aout issues and solutions for gender equality. Nowaday, in Southeast Asia need to once again unleash and engage the female labour force to sustain prosperity and development.’ S06 introduced the topic in the first sentence with the reference nouns of ‘issues’ and ‘solutions’ and with gender equality as the focus. The second sentence referred partially to the thesis but was copied directly from the source text. Two problems were then presented to support the thesis: ‘females choose go to work, but most of the hoursework and childcare is still done by women. … In the workplace, there is also a huge gap between men's and women’ salaries, which leads to many women to give up on marriage or to make it a tool for exchange.’ However, the nature of the problems was vague due to the lack of accurate explanation of the key points. The notion of further consequences present in the original text was not summarised and the solution was not logically developed. In the post-test, S06 was able to identify the topic more accurately and outline details of the text, with only some minor inaccuracies. S06 wrote, ‘The central theme of this article is Unique Barriers to Women-Owned Businesses. The author discuss the challenges women entrepreneurs face in starting a business and how to help women overcome these challenges.’ The text continued, ‘The authors note that … There are four major factors that limit female entrepreneuship’, with each major factor being elaborated in the subsequent sentences. The text finally stated the significance of the topic and the usefulness of the article, although more detail could be expected. On the whole, the student's summary in the post-test is more polished with regard to genre, register and lexico-grammar. However, it is important to exercise caution because there is uncertainty with regard to whether the student utilised technology tools, such as translation software or similar generative artificial intelligence (AI) software, to compose the summary. Further discussion on this topic will be presented in the Implications section below.

With regard to the impact of the R2L pedagogy on the participating students’ ability to write grammatically sound and complex sentences, the Wilcoxon signed ranked test also revealed a statistically significant improvement in test performance from the pre-test to the post-test, z = 2.52, p = .012, with a medium effect size (r = .38). The median score for the test performance increased from the pre-test (Md = 1.0) to the post-test (Md = 1.4). Student 18 (S18) in the academic year 2020, whose score is also located within 1 SD of the mean in the pre-test, produced eight sentences in 150 words. The eight sentences contain a mix of simple, compound and complex sentences and show some grammatical errors, such as subject–verb agreement errors, missing articles and some verb tense errors. To illustrate, the first half of the text is as follows:

The topic of this text is … in Japanese labour market [missing article], and the treatment of women in Japanese society. The author's thesis statement is that the treatment of women in Japanese labour market [missing article] is unfair … Women have done a lot in the history [unnecessary article] and the mordern [spelling error] families [unnecessary article], but with the development of Japan's economy, Japanese women lose their working opportunities [verb tense]. I think the issue is important. There are more higher [problematic comparative form] requirements of women in the modern society [unnecessary determiner], the woman who do not have degree is difficult to jobs and get people's respect [subject–verb agreement/missing article/missing verb/run-on sentence]. Most of women are in poverty, because of the gender-unequal [wrong form of word].

Finally, for the dimension of cohesion, vocabulary and punctuation, the Wilcoxon signed ranked test indicated a non-statistically significant difference in the median score from the pre-test (Md = .05) to the post-test (Md = .06), z = 1.47, p = .14. Again, among the students located within 1 SD of the mean in the pre-test, Student 10 (S10) in the academic year 2020 used cohesion only through repetition of lexical items, employed vocabulary that was mainly copied from the original text and inserted full stops incorrectly. S10 wrote in the pre-test summary:

The central topic of the document is Japanese society remains a deeply gender-unequal one. This article mainly explains that Japan is an unequal society from several aspects such as the inequality of women in the labor market. The important point raised by the author is discrimination. The author use different examples to explain discrimination in Japan. Such as ‘dual-track’ and Non-standard employment. I think the topic is very important. Because this article explains why Japanese women are unequal in society and family.

Discussion

The positive impacts that the instruction treatment may have yielded with regard to the content dimension of summary writing could result from multiple elements contained in the R2L methodology. The first cycle, that is, topic preview and explanation of how the text unfolds, assists in gaining an overall understanding of the text, in particular a grasp of the organisation pattern. Understanding of the latter is vital for the comprehension of a text (Horowitz and Wilburn, 2017), which, in turn, assists students to recognise and outline the key points in their summary writing (Hood, 2008). The second cycle, that is, detailed reading of shorter extracts from the text, including an explanation of lexical meanings and cohesion, joint construction of a new text with guidance from the teacher and, finally, independent construction of the original text, models the process of how changed wording in summary writing still conveys the meaning of the original text (Hood, 2008) and provides opportunities for applying what has been learned. The current study used two texts for the instruction and for each text the instruction followed the R2L pedagogical cycles. In the independent construction stage, the students wrote a summary of each text, which was part of the assessment for the unit, and received feedback from the lecturer. Having received the instruction, which offered various methods for improving summary writing, according to the post-test results, the participating students seemed to improve their summary writing ability. However, this improved post-test performance could not be purely a result of the instruction treatment. Such improved performance could also, to a lesser extent at least, result from the students’ familiarity with the topic and the test procedure.

The positive impacts that the instruction treatment may have yielded with regard to sentence writing could have resulted from the third cycle of the R2L methodology. The instruction focused on sentential features of academic writing, such as use of nominalisation, passive voice and modality. With regard to complex sentences particularly, the instruction isolated elements of the sentences and offered reconstruction opportunities. Given that the participating students had met the English language requirement of the university and had sufficient grammatical knowledge of sentence construction, such explicit instruction should further enhance their ability to apply their knowledge base and write grammatically sound sentences of various types. In a review of the then-existing empirical literature, Ortega (2015) states that instruction containing sufficient writing practice contributes to the production of various diverse and elaborated syntactic structures. The instruction in the current study, which followed the R2L pedagogy as described previously, provided the participating students with abundant opportunities for writing practice and likely resulted in the improved performance of sentence writing in the post-test.

Surprisingly, the instruction treatment seems to be ineffective in improving the participating students’ performance in relation to cohesion, vocabulary and punctuation. Although cohesion and vocabulary learning are distinct concepts, they are interconnected because a large vocabulary enhances a person’s ability to recognise and utilise cohesive devices effectively in writing. Taking a broad and inclusive approach to cohesion and vocabulary learning, the authors believe that this ineffectiveness could partly result from the fact that vocabulary learning occurs incidentally during comprehension-oriented activities (Hulstijn, 2013) and gain from incidental vocabulary learning is limited (Teng, 2022). It is possible that the limited vocabulary that has been learned incidentally from the texts during the instruction treatment may not be applicable to the production of another text. Unlike the learning of vocabulary, learning how to construct sentences is rule based, and the organisation pattern across the texts used in this study (e.g. topic sentences and the introduction and conclusion) presents similarities. The participating students had opportunities to apply what they noticed or what their consolidated learning had taught them when isolating text elements in the post-test, which might then have contributed to their improved performance in the content and sentence writing dimensions.