Teaching Summary Writing as a Transferable Academic Skill in Ukrainian University Context

Abstract

Teaching and assessing summary writing remains in the focus of research into academic literacy, the issue revisited by every new generation of researchers in the light of the emerging affordances and plurilingual reality. In higher Ukrainian education, where teaching academic skills has gained momentum only in the past 15–20 years, summary writing pertains mostly to teaching Ukrainian to international students, teaching English for professional communication and training post-graduates. A 15-year practice of teaching academic English to students of linguistics has provided the author with extensive empirical data which, upon analysis conducted within action research, have enabled maximized effect on teaching summary writing. Along with the teaching methodology, the assessment technique was also optimized, based on rating scales with the criteria accepted by the students and detailed feedback provided to them via the Google Classroom platform. The comparison of pre-and post-tests of writing summaries of Ukrainian research articles verifies the hypothesis about the transferability of summarization skills built in English to academic L1. The author formulates the limitations of the action research project, implications and prospects of further research.

1. Introduction

Summarizing is considered one of the most complex cognitive and meta-cognitive activities [1], both in L1 and especially in L2 [2]. Hence, in the CEFR [3], the summarisation skill is for the first time referred to on level B2 and, with the introduction of mediation scales in the CEFR Companion Volume, it received much more attention in the descriptors on this and higher levels [4] (p. 207).

Grabe and Zhang [5] argue that summarizing is the core skill of academic writing, which consists of processing and transforming knowledge obtained through reading. Normally, researchers provide brief summarized pieces from scholarly papers, which have shaped their vision of the researched issue, in order to underpin their hypotheses, and argue for or against the views of other scholars in the section “Literature review”. This section is mandatory in articles and theses as it reveals the scholarly background of researchers, provides foundations for research paths, clarifies their premises while organizing the ideas logically and in a critical and persuasive manner [6].

Apart from selecting the ideas worth referring to in order to support own point of view, authors need to decide in which way to do this—making direct citations or putting in-text references in order to avoid academic dishonesty—plagiarism [7,8]. So, an academician writing a summary initially decides which of the ideas are indispensable for a literature review and further employs the language which should not be verbatim of a source text (ST), terse and clear. The summarized excerpts should also be properly organised to reproduce the ST’s line of argumentation.

Summarisation skills are quite challenging to develop, and according to research [2,9,10], they should be systemically taught. In Ukrainian senior secondary and higher education, students are expected to be able to summarise sources while doing library research either in L1 or L2. Despite the declared objectives, it appeared futile for the author of this article to discover empirical research offering a robust methodology of teaching/assessing summary writing.

The purpose of this paper is to bridge the existing gap and present the methodology of teaching summary writing to university students majoring in linguistics. It was developed by the author of this article back in the 2010’s [11] and has been practiced each year ever since. While progressing in academic English, the students consistently expressed their contentment with the ability to summarise sources for a literature review in their course/bachelor’s papers in Ukrainian, too. The idea to check if summary writing skills in L2 could be transferred into summarising skills in L1 prompted this research study. So, this paper will present the methodology, procedure and results of this comparative action research.

Literature Review

Reviewing the literature that addresses summary writing, it seems reasonable to specify the nature of this complex skill. Knoch and Sitajalabhorn [12] regrettably concede that the study of the cognitive processes of reading-to-write is still in its infancy. However, they find it worthy to quote Grabe [13] who lists the decisions that a summary writer needs to make, such as “how much and which information should be selected from the text, how the information will fit task demands and writer goals, how accurately the information from the source texts should be represented and what formal mechanisms should be used for transforming the textual information to fit the final written product” [12] (p. 303). Similarly, Yu calls summarization “a discourse in its own right” [14] (p. 97), ascribing to its cognitive processes the comprehension of the ST, reduction and reconstruction of the main ideas.

Within the interaction between an ST, a reader, who eventually becomes a writer, and a transformed text (TT), the cognitive processing of a reading text insensibly merges with the cognitive processes of writing a text. The ability to read with full understanding or, at least for understanding the main ideas and gist, enables effective completion of the reading stage. Critical reading skills, i.e., the ability to distinguish the crucial information from supporting details, examples, enumerations, as well as explicitly or implicitly stated information, in most cases, depends upon the learners’ range of vocabulary and grammar control as well as awareness of relationships between the ideas expressed in parts of the text [15,16]. Obviously, insufficient L2 resources hinder the adequate processing of a ST and especially summarizing it [9,17]. L2 summary writers use excessive language borrowings, “Exact copies” or “Near copies”, from the ST much more frequently compared with L1 writers [2,18]. Yet, having determined 70 strategies for processing a reading text and producing its summarised version, Li [10] argues that writing ability contributes more to effective summarisation than reading skills. Researchers [2,5,10,19,20,21] have also provided evidence that a writer’s capacity appears critical for producing an effective summary.

Chin [22] examined the performance of L2 summary writers at two different levels of proficiency. According to the study, the more proficient students tend to do better in paraphrasing and reformulating the content of the ST, while the lower-level learners copy more frequently verbatim from the ST. Less proficient writers tend to produce longer and linguistically inaccurate summaries with abundant insignificant details. More advanced students write shorter, more content-rich and coherent TTs using their own language—lexis, syntactical structures, and linking devices.

Having considered the issues relevant to teaching summary writing, such as revealing the content of an ST, transforming language, and reflecting the rhetoric in a TT, let us focus on the assessment of academic summary writing.

In the area of standardized testing, the issue of assessing summarization has achieved significant progress [23], whereas in classroom settings, the methodology of both teaching and assessing summary writing remains under-researched [24,25]. A noteworthy contribution to bridging this gap was made by Chin [22] who elaborated a framework for assessing TT, which includes three sets of criteria: content-related criteria, writing-related criteria and a holistic scale for evaluating the overall quality of the summaries (both content coverage and writing quality).

The in-depth scrutiny of summary writing performances of students with different levels of proficiency in L2 led the researcher to formulate some interesting implications for teaching and assessing summary writing ability. In terms of enhancing linguistic transformation skills at the high–intermediate level, Chin recommended more practice on the use of synonyms while learning to paraphrase, combining sentences within and across paragraphs, and training students in the accurate use of syntactical structures. With respect to text organization, the students should receive advanced training in the analysis of an ST macrostructure so as to be able to follow the logical and rhetorical development of ideas in the ST and recreate them in the TT with the help of appropriate connectors or transitions.

From the abovementioned study by Li [10], it can be implied that in assessing summaries, more weight should be assigned to writing skills to make the total scores more precise. The researcher also proposed beginning by writing summaries of the texts of familiar genres and encourage students “to read journal articles that offer excellent examples of proper source use to familiarize themselves with the real-world context and conventions of academic writing” (p. 20). In addition, making a reference to Yang [26], Li suggests that training should be provided on selecting the most important chunks of the text and paraphrasing them, then focus should be made on a higher level of summarization skills, such as condensation and discourse synthesis thus leading to a maximal paraphrase of the text.

Finally, relevant to this action research is a comparative study of summarization skills of students in L1 (Chinese) and L2 (English) conducted by Yu [27]. He discovered that the summaries written in L1 were longer probably because trainees understood the texts much better, and probably because of the deficit of linguistic resources to write longer summaries in English. Importantly, the author refers to the studies by Cumming et al. [28] and Corbeil [29], who identified the transfer of summarization skills between English and French, which belong to the same language family. However, Yu stated it was impossible to detect the transferability of summarization skills in too-distant languages such as English and Chinese.

These implications seem quite worthy to conclude the literature review and move on to the description of the current research, which, to a great extent, resonates with the analysed studies.

2. Materials and Methods

2.1. Research Framework

The action research was conducted by a teacher of the course of Academic English (AE), the author of this article, to graduate students of linguistics at the Taras Shevchenko National University of Kyiv The materials have been updated and improved several times during the 15-year implementation of the course [11], whereas the action-oriented approach [30] to its design and delivery was intact. This paper presents the procedure and outcomes of the latest version of the course. Additionally, the author aimed to verify the hypothesis of the transferability of summarization skills developed in English to L1 academic writing.

This action research, therefore, posed two questions:

• How does the developed methodology of teaching summarising affect the students’ ability to write summaries of academic English texts?
• Does learning summary writing in English enhance skills in writing summaries of L1 academic texts?

The participants in the research were selected based on the scores of an entry test administered to the students majoring in Oriental languages. Upon testing, 15 students (two males and 13 females) aged from 20 to 23, all Ukrainian-Russian bilinguals were ranked according to the test results and formed three focus groups: advanced, upper-intermediate and intermediate.

The pre-test consisted of two parts: (1) a test of English reading and lexicogrammar; (2) summarising an excerpt from an L1 academic text (about 400 words long).

Part 1 contained five test tasks which were selected from B2 First Handbook for teachers [31] based on the tasks’ relevance for testing summarisation skills: multiple choice cloze (knowledge of vocabulary), open cloze (grammar control), keyword transformation (syntactical structures), multiple choice questions (understanding tone, purpose, main idea, gist, text organisation features), as well as gapped text (cohesion, coherence, and text structure). The total number of questions was 34, with a total score of 52. The students were allotted 80 min to do the test. The total score was converted into the 5-point grade scale to fit the approach to scoring adopted in this research.

Part 2 tested the summarisation of an excerpt from an L1 research article (RA) about using electronic teaching platforms during the COVID-19 pandemic. The excerpt was 377 words long (four paragraphs, 14 sentences in total). The task was to write a summary of 260 words which makes 70% of the length of the ST. The criteria announced to the testees included (1) coverage of major content points, (2) TT organization, (3) use of fresh language, and (4) word count. The maximal score for the summary was 5 points.

The post-test in L1 offered the students an excerpt from an RA, which was 519 words long, consisting of 10 paragraphs, including bullet point lists, an in-text citation, statistics, reporting words and discourse markers. The topic was using messengers in distant learning. The criteria to assess the summary were the same as those to assess the English language summary.

Across the course, the abilities to paraphrase and write summaries in English were tested upon covering each module. To gauge the skill development, the researcher used the rating scales developed and utilised for several previous academic years, which appeared quite effective. (See Appendix A, Appendix B and Appendix C). All tests were conducted online synchronously, with the teacher proctoring the process.

2.2. The Academic English Course Design

The course [11] aims to develop oral reception, oral production and skills in mediating a text at level C1 [4]. The four modules of the course, which focus on paraphrasing, summarising popular science texts, and summarising RAs, are arranged as a scenario [30], the tasks of which are steps towards creating a visible, meaningful product—an academic presentation based on a summarised RA. The course covers ten weeks with four contact hours a week, which in the period described were conducted online.

Module One of the course envisages students’ mastering general academic and field-related vocabulary. Modules Two, Three and Three (advanced) focus on academic reading and writing, whereas Module Four focuses on preparing an oral scholarly presentation as a final product of the scenario.

2.3. Teaching and Assessment Methodology

The methodology of teaching academic writing encompasses three consecutive stages.

Table 1. The Academic English course profile.

Module	Writing	Learning Objectives	Materials	Assessment
Module Two	Paraphrase	Retain the key vocabulary, alterthe rest vocabulary and structure	Paragraphs from RAs (three-five composite sentences)	An analytic rating scale (two criteria, five bands)
Module Three	Summary	Convey the gist, use fresh language and reporting verbs, compress the ST by up to 75%.	Popular science texts, excerpts and abstracts of RAs (500–1300 words)	An analytic rating scale (four criteria, five bands)
Module Three Advanced	Research report	Convey the core information, use reporting verbs and compress the ST by 75–85%.	Section Results/Findings of RAs	An analytic rating scale (four criteria, five bands)

presents the learning objectives, teaching materials and assessment methods employed.

The methodology of assessing students’ L2 writing was developed by the author of this article, who had undertaken advanced training in designing rating scales. The scales were updated as the teacher grew expertise in rating scale design.

2.3.1. Stage One: Teaching and Assessing Paraphrase

Objective: paraphrasing short excerpts of academic texts, assessing students’ writing, and providing detailed feedback.

Teaching materials: paragraphs from RAs consisting of three to five composite sentences related to linguistics (cultural studies, CALL, EMI), i.e., the topics related to the majors’ specialisations and familiar to them.

Procedure:

(1)

Enhancing English academic vocabulary through reading area-related texts, supplying relevant equivalents to the English terms in L1;

(2)

Reading about stages of writing a literature review of an MA thesis; focusing on the representation of knowledge obtained through the study of scholarly sources in a concise and clear way; comparing and contrasting excerpts of STs and TTs; and drawing insights about possible paraphrasing strategies, including those of avoiding plagiarism.

Assessment: Paraphrase is generally valued for the transformation of wording while rendering the meaning clearly and in a coherent way. To assess paraphrase, therefore, the author developed an analytic rating scale containing two criteria and five bands. (See Appendix A).

2.3.2. Stage Two: Teaching and Assessing Summary Writing

Objective: summarising academic texts beginning with less specialised ones, i.e., belonging to popular science genre up to the excerpts of RAs.

This stage is of special significance in terms of the scope of skills that are consecutively developed: (1) skills of careful/critical reading; (2) skills of writing from the ST; (3) ability to introduce direct quotations and use other referencing types.

Teaching materials: texts of a popular science genre; excerpts and abstracts of research RAs on the issues of summary writing and academic dishonesty.

Procedure:

(1)

Reading about paraphrasing, summarising and synthesising, discussing the purpose of each type of academic writing; reading about different types of referencing, discussing the meaningfulness of quotations in the literature review section of a research paper; analysing excerpts of the literature review for the use of reporting verbs that introduce in-text citations, refer to what the reviewed authors stated, argued, etc., express the authors’ stance, etc.

(2)

Skimming and discussing an abstract to the RA. The use of paraphrase in summary writing: a comparison of L1 and L2 writers to discuss the key questions highlighted in the article [2]; reading an abridged version of the paper by Writing from Sources, Writing from Sentences [9] to obtain first-hand information about copying, summary, paraphrase, or patchwriting; examining some excerpts of summaries to define the techniques used by their authors (copying, patchwriting, etc.); and evaluating the quality of referencing in the discussed excerpts.

(3)

Careful reading of an excerpt from an academic text of informative character (107 words) and five summaries of this text; discussing the quality of the summaries; drawing inferences about what makes a summary successful.

(4)

Examining the use of reporting verbs and discourse markers/linking elements in several excerpts from RAs; filling in two tables: one with reporting verbs, the other with linking elements, complemented with their translation into Ukrainian.

(5)

Analysing a popular science text (1292 words) following the algorithm: identify the introduction, the main body and the conclusion; analyse the rhetoric of each paragraph (is it description/narration/comparing/contrasting/providing arguments for and against, etc.); highlight the discourse markers that help identify the rhetorical functions performed in the paragraphs; distinguish facts from writer’s opinions; think of useful language to introduce the author’s opinion; ensure that summary is at least twice as short as the ST.

(6)

Familiarizing with three techniques of summary writing: (a) taking notes; (b) analysis of the text, such as identifying keywords and main information, the author’s stance and rhetoric, and (c) a frame for using prompts—reporting words and discourse markers—to introduce new ideas in paragraphs; discussing the pros and cons of each method.

(7)

Writing a summary (500–650 words) using one of the algorithms (running ahead, it is interesting to note that none of the students chose a very demanding way of writing a summary based on notes and expanding them to sentences, and the ratio of preference between the rest two options was 55:45); the fragments of the TTs were shared on the screen and discussed by the class.

(8)

Consolidating the skills of summary writing by editing summaries written by the students collected by the teacher in the previous years of her work; doing a rhetorical analysis of the popular science text (844 words), choosing one of five text summaries, and editing it. At this stage, the students were offered the use of the analytic scale presented in Appendix B to fully understand the criteria for assessing a summary and enhance knowledge about the features typical of an effective summary.

Assessment: while deciding upon the construct of the test to gauge the ability to write a summary at Stage Two, it was assumed that the crucial feature for a summary is accurately conveying the ST’s message, distinguishing the main thesis from supporting ideas. Secondly, it is important, especially for a summary, to evaluate the language that enables the task completion. From this perspective, the second criterion considers the relevant paraphrase of the gist and sufficient contraction of supporting information, plus abiding by the rules of academic integrity. In addition, sophisticated lexicogrammar is used in the TT, the latter should necessarily be as coherent as the ST through the use of relevant reporting verbs and discourse markers. Last but not least, a good summary should not exceed 50% of the ST’s length. (See Rating scale to assess summary writing in Appendix B).

2.3.3. Stage Three: Teaching and Assessing Summarising a RA

Objective: summarising the information contained in such sections of RAs, as “Findings/Results”, “Discussion” and “Conclusion”. It is from these sections that researchers generally obtain meaningful novel information, which they further use while writing a literature review to underpin their hypotheses. The major task of reviewers is to distinguish the core data that emphasise the accomplishments of the reviewed author, elicit their stance on the data provided by research on similar topics and clearly and coherently summarise the findings, making appropriate references. The completed summary, therefore, is in actual fact, a research report (RR).

Teaching materials: a choice of RAs published in English-language journals about the issues related to language education, EMI, as well as linguistic and rhetoric features of academic writing in the target oriental language. Apart from the topic, the texts are selected by the teacher on the grounds of an analysis of the conceptual and linguistic complexity of the text, the moderate length (4000–6000 words), presence of all the mandatory sections of an RA, but only a few graphs and tables. The students may choose an article according to their research interest, but very few take advantage of this opportunity after they make sure that the teacher’s choice is the most reasonable in terms of content, length and linguistic features.

Procedure:

• skimming the initial sections of the chosen articles to obtain a general idea of the research framework and questions;
• writing a summary of the abovementioned sections paying special attention to (a) content, (b) reporting the author’s standpoint, (c) accurate in-text citations and references, (d) compressing the summarized sections up to 85%—1–2 sentences per a paragraph (appr. 400–450 words).

Assessment: a rating scale to assess RR is based on the scale for summary writing with some relevant alterations. (See Appendix C).

3. Results and Discussion

The results in this section present the scores obtained by the focus groups for the following:

• Pre-testing (in English and L1);
• Stage One Paraphrase;
• Stage Two Summary writing;
• Stage Three Writing a summary of a RA;
• Post-testing (in L1).

3.1. Pre-Testing and Placing Students in Focus Groups

The tests of proficiency in English and the ability to write a summary in L1 are described in detail in Section 2.1.

Table 2. Results of pre-testing.

Students	Test of English					Score (Converted)	Score in Ukrainian	Total Score
	MC Cloze	Open Cloze	Key Word Transformation	MCQ	Gapped Text
VK	8	8	12	12	12	5	5	10
MS	8	8	10	12	10	4.6	5	9.6
EO	8	8	10	12	10	4.6	5	9.6
MG	8	8	12	10	10	4.6	5	9.6
DM	8	8	12	10	10	4.6	5	9.6
Mean	8	8	11.2	11.2	10.4	6.6	5	9.7
DP	8	8	10	10	10	4.4	5	9.4
KS	7	7	8	8	12	4	4	8
OS	8	8	8	10	8	4	4	8
DS	8	7	8	12	6	3.5	4	7.5
VS	6	6	8	8	6	3.2	4	7.3
Mean	7.4	7.2	8.8	9.6	8.4	4.4	4.4	8.2
AT	5	4	8	8	8	3	4	7
TK	5	5	8	8	6	3	4	7
AF	4	4	6	6	8	2.7	3	5.7
BH	4	4	6	6	6	2.5	3	5.5
AB	3	3	6	6	4	2.3	3	5.3
Mean	4.2	4	6.8	6.8	6.4	2.7	3.4	6.1

shows the scores which allowed the placement of the students into three focus groups, with students’ full names replaced with their initials.

The mean scores for the entry test in English of each focus group suggest that in terms of lexicogrammar, the situation is quite typical of the relevant levels of proficiency. The task Key words transformation, which required making changes to the syntactical structures, appeared more demanding in all groups, as well as Gapped text, the task which focuses on cohesion, coherence and text structure. It may be presumed that these aspects will need more effort when learning to write a summary.

The pre-test in L1 was completed by students quite effectively, although, in the majority of cases, the assigned word count (260) was exceeded by 40–50 words, with three summaries containing below 200 words. The other very serious flaw was that the students attempted to copy some sentences, just leaving out the words that they had found not meaningful, i.e., they patchwrote. However, they managed to supply the correct equivalents to L1 discourse markers, such as since/because; it is necessary to note/point out, besides, because, as a result, although, however, according to the research. Nevertheless, the texts appeared somewhat disproportionate in conveying information, as if the writers recollected the word count at the very last moment and, in an attempt to abide by it, left out some ideas that were worth mentioning. None of the students used reporting verbs; therefore, these isomorphic summaries looked rather like shortened versions of the ST than paragraphs of a literature review.

3.2. Stage One: Paraphrase

Paraphrasing in this study means conveying the main and supporting ideas, retaining the key vocabulary of the ST, and replacing the supporting vocabulary with synonyms or definitions. Syntactically, the sentences could be split or joined, turned into subordinate clauses or infinitive phrases. Cohesive devices could be replaced by synonymous means.

Testing materials: a paragraph from an RA consisting of six composite sentences about the use of English as a medium of instruction (EMI) (130 words).

As is seen in

Table 3. Results of paraphrase.

Student	Scores		Mean
	Criterion 1	Criterion 2
VK	5	5	5
MS	5	5	5
EO	4	5	4.5
MG	5	5	5
DM	4	4	4
Mean	4.6	4.8	4.7
DP	4	5	4.5
KS	5	5	5
OS	4	5	4.5
DS	4	3	3.5
VS	4	4	4
Mean	4.2	4.4	4.3
AT	4	3	3.5
TK	4	4	4
AF	2	3	2.5
BH	3	3	3
AB	1	2	1.5
Mean	2.8	3	2.9

, paraphrasing appeared successful (i.e., >2) in 14 incidences out of 15, which suggests that the students adopted the approach, followed the rules and had sufficiently practised paraphrasing. The quality of paraphrasing differed, which reflected the English proficiency of the writers.

The advanced students used more sophisticated vocabulary, made reasonable syntactical changes accurately, and used relevant cohesive devices, keeping up the coherence of the ST. The upper–intermediate students did not feel confident in transforming the ST, doing only moderate revision such as replacement of adjectives in the function of predicative (e.g., was critical → was crucial), occasionally using wrongly chosen synonyms (additionally → correspondingly) which, consequently, distorted the message and doing insignificant changes in sentence structures.

The intermediate students attempted to replace the vocabulary wherever their proficiency allowed them (mostly adjectives and cohesive devices, e.g., for example → for instance); syntactical alterations were minimal (e.g., splitting lengthy sentences); and in some cases, they resorted to sheer copying quite lengthy chunks of the ST (6–8 words). Obviously, the conclusions made with respect to students’ paraphrases resonate with the findings of some previous research. First of all, the participants confirmed Keck’s [2] conclusion about the multiple challenges the non-native speakers face when paraphrasing; this results in more Exact copying than the native speakers would have done. Secondly, in line with Chin’s findings [22], the group which is less proficient in L2 used more Exact copying and paraphrased less than the more proficient group.

Cho and Brutt-Griffler [32] have also detected slower progress in paraphrasing made by lower proficiency level students as compared with their more advanced colleagues since they “lacked writing skills that needed lexical and syntactical knowledge” [32] (p. 252). The researchers assert that “English learners’ lack of vocabulary knowledge (or their inability to use it) results in using words redundantly and ambiguously when paraphrasing” (p. 254).

The insights drawn from this research echo what Cho and Brutt-Griffler [32] suggest regarding the enhancement of students’ lexical, semantic, and syntactic knowledge, as well as the self-awareness of plagiarism. Obviously, teaching synonyms for less vague nouns or it-cleft structures should occupy a proper place in teaching. In line with these ideas, the methodology presented in this article envisages, primarily, the enrichment of academic English vocabulary and provides clear explanations of academic integrity, including reading authentic RAs about paraphrasing and summarizing ‘to learn from’ [2,9].

3.3. Stage Two: Summary Writing

The module test task was to render the main idea and the gist, attending to essential supporting details and creating a coherent TT. Unlike with the paraphrase, the requirement was to be terse and significantly compress the ST (up to 50%), use fresh language, relevant reporting verbs to introduce key sentences/paragraphs as well as proper discourse markers/connectors.

Testing material: an excerpt from an RA (336 words) about the challenges of academic writing in Chinese. The TT was to be 155–170 words long,

Table 4. Results of summary writing.

Student	Scores				Mean
	Criterion 1	Criterion 2	Criterion 3	Criterion 4
VK	5	2	5	5	4
MS	5	5	5	5	5
EO	5	2	4	5	4
MG	5	2	4	4	3.75
DM	5	5	5	5	5
Mean	5	3.2	4.6	4.8	4.4
DP	5	4	5	5	4.75
KS	4	4	5	4	4.25
OS	5	4	5	5	4.75
DS	4	4	4	5	4.25
Mean	4.5	4	4.75	4.75	4.5
AT	4	4	2	5	3.75
TK	2	4	4	4	3.5
AF	3	3	4	4	3.5
BH	3	3	3	3	3
Mean	2.8	3.4	3.2	4.2	3.4

.

The outcomes were utterly surprising: most advanced students (three out of five) violated the rules and made up to 50% of copying or near copying, with the rest of the text undergoing moderate structural changes. Despite this, they managed to use various relevant reporting words and discourse markers to maintain the TT’s coherence.

Three out of five upper-intermediate students attempted to use fresh language and compress the text focusing on the main ideas and being accurate with in-text references, although not without occasional near copying. They demonstrated quite appropriate use of reporting verbs and discourse markers. Unfortunately, academic dishonesty was detected, which led to a reduction in the number of students in this focus group.

The results in the group of intermediate students were varied in several respects. First, one of the summaries met almost all the requirements with the exception of ignoring the use of reporting verbs, which was required by the task. The second summary from the intermediate group was also written independently, with minimal copying from an ST, but inadequately conveying the initial paragraph’s message. The third summary was an attempt to use own language: it resulted in paraphrasing a topic sentence from each paragraph by providing synonyms wherever possible, with a consequent lack of coherence between paraphrased sentences and the absence of reporting verbs. Otherwise, the author used relevant reporting verbs and quite coherently organised the TT. Summary Four was relatively accurate in terms of using fresh language and text organization, although too short to cover both the main idea and some essential supporting details. The last of the summaries was almost fully copied, lacked coherence, and was inadequately short.

Thus, the intermediate and upper-intermediate groups, excluding one person from the upper-intermediate group and the three from the advanced group, have progressed in terms of improving their paraphrasing skills and showed sufficient development in using reporting verbs and discourse markers.

However, it seems interesting to discuss the results of summarization in the advanced group. Upon receiving the teacher’s feedback on their multiple verbatim borrowings, two of the students provided an excuse which seemed lame in light of their actual ability to write summaries, especially when compared to the significant progress made by the less proficient groups. The third of the students confessed that she had not thought of copying 6–8 long chunks from the ST as plagiarism.

Here, it is reasonable to mention scholars’ dual positions on what should be considered plagiarism and what not. Some researchers [33] consider shorter copied strings and/or deleting words in patchwriting as a step toward learning to write not from sentences but from sources. In fact, the idea of using a “copy-paste-delete” strategy seems more attractive to those who would not wish to strain their cognitive resources to process an ST and do its maximal revision scrupulously.

In the case of the “defaulters” in the advanced and other groups described here, it might be that, comparing own TT with ST they notice a big gap which leads them to verbatim borrowing instead of using fresh language. Nevertheless, it is undoubtful that more advanced students easily compress the ST as was earlier established in [2].

Before moving on to describing Stage Three, it is worth noting that according to the recent modification of the course curriculum, the students were taught how to write a synthesis essay. This aspect of the author’s methodology is beyond the scope of this article; however, it is clear that the students kept enhancing their summarization skills during this additional practice.

3.4. Stage Three: Writing a Summary of an RA

The final stage, according to the curriculum, envisages teaching/testing the ability to summarise the information contained in such sections of RAs, as “Findings/Results” and “Discussion” to fit these brief summaries in the literature review. This summary is termed in the article as an RR.

Testing material: the students were offered an excerpt of RAs published in English-language journals about linguistic and rhetoric features of academic writing in the target oriental language. The excerpt was 550 words long. The TT was to compress the ST by 75% and contain 140–160,

Table 5. Results of writing a summary of a RA.

Student	Scores				Mean
	Criterion 1	Criterion 2	Criterion 3	Criterion 4
VK	5	5	5	5	5
MS	5	5	5	5	5
EO	5	5	4	5	4.75
MG	5	5	4	5	4.75
DM	5	5	5	5	5
Mean	5	5	4.6	5	4.9
DP	5	5	5	5	5
KS	5	4	5	4	4.5
OS	5	5	5	5	5
DS	5	4	5	4	4.5
Mean	5	4.5	5	4.5	4.75
AT	4	4	4	5	4.25
TK	4	4	4	4	4
AF	3	3	4	3	3.25
BH	3	3	3	3	3
AB	3	3	3	3	3
Mean	3.4	3.4	3.6	3.6	3.5

.

The results demonstrated the following: the students from the advanced group fully and accurately completed the RRs, and they reflected the core information that was worth referring to, appropriately wording and substantially compressing it. They adequately used a variety of reporting verbs to introduce the author’s stance, discourse markers were not abundant but used to the point. Notable is that a similar outcome was achieved by two persons from the upper–intermediate group who managed to enhance their summary writing skills, which is definitely a great achievement.

The other two students from the upper–intermediate group were nearly as successful, although their linguistic repertoire was not equally sophisticated as that of the more advanced students. Additionally, it appeared too difficult for them to compress the RR to meet the word limit. Nevertheless, in all other respects, their RRs deserved quite high scores.

The issue of exceeding the word count detected in the upper–intermediate group was also present in the intermediate group, thus showing the inability of all students to compress the ST as required. Although these groups adequately distinguished the core information, their language proficiency did not allow them to write RRs faultlessly, which mostly concerned the syntactical aspect of writing. To give them credit, there were some rather successful attempts to use fresh language, relevant reporting words and discourse markers. Incidences of near copying were found only in two works out of five. In-text citations were made with precision not only in this but in two other groups as well. Overall, the progress of less advanced students appeared more visible and, therefore, significant.

The more successful use of reporting words and discourse markers by almost all students implies that the work on the rhetorical analysis of an ST, which began when learning summary writing and continued during the stage of writing a synthesis essay not described here, was fully relevant and effective. Irrespective of the quality of the employed language, sophisticated in the advanced group and rather plain in the intermediate group, the students mastered the ability to single out the core information and appropriately shape it to fit a literature review.

In Rivard and Gueye’s article [34], it is stated that “[r]esearch on writing confirms the importance of explicitly teaching textual features characterizing academic texts, including connector words for both L1 and L2 learners”. The authors also found it beneficial for the students to compare connector use in L1 and L2 in order “to understand the particularities of each language more explicitly” (p. 218). In the above description of the methodology of teaching summary writing it has not once been pointed to the necessity to explore the use of L1 equivalents to the discourse markers used in the English language RA. Rivard and Ndeye R. Gueye, for instance, claim that the repertoire of English connectors is limited. This can be said not with respect to the L1 connectors but to the reporting verbs, which are few and too general, so the translation of English RAs into L1 without repeating the same verb too frequently is rather demanding. This issue seems to be of some relevance to the transferability of summarization skills into Ukrainian academic writing.

To conclude this sub-section, the scores obtained by the focus groups while doing Stages One-Three and the interpretation of data allow us to answer research question 1: How does the developed methodology of teaching summarising affect the students’ ability to write summaries of academic English text? Despite the significant variance of linguistic features of summaries written by the students of the focus groups, all students attained a level enabling them to write summaries of English academic texts of appropriate quality.

3.5. Post-Training Testing—Writing A Summary of an RA in L1

Testing material: a 506-word long excerpt of an RA in L1 (section “Results”), including 12 paragraphs and three bullet point lists of outcomes and conclusions. The topic covered the comparison of different messengers in terms of their use in distance learning. The TT was to include 50% of the information and be 245–255 words long. To assess the final product, the rating scale developed for testing the research report was used (see Appendix C).

Table 6. Results of post-test in writing a summary in L1.

Student	Scores				Mean
	Criterion 1	Criterion 2	Criterion 3	Criterion 4
VK	5	5	5	5	5
MS	5	5	5	5	5
EO	5	5	5	5	5
MG	5	5	5	5	5
DM	5	5	5	5	5
Mean	5	5	5	5	5
DP	5	5	5	5	5
KS	5	5	5	4	4.75
OS	5	5	5	5	5
DS	5	5	5	4	4.75
Mean	5	5	5	4.5	4.87
AT	5	4	5	4	4.5
TK	5	4	5	4	4.5
AF	4	4	4	4	4
BH	5	3	4	4	4
AB	4	3	3	3	3.25
Mean	4.6	3.6	4.2	3.8	4.05

presents the results of the post-test in writing a summary in L1.

Replicating the data by Keck [2], the testees revised the ST in L1 more effectively than the ST in L2. As is seen in Table 6, the advanced students, including one student from the second focus group, repeated their top scores. Of more interest in this respect is the level attained by the other eight students. They all improved their summary writing skills from a minimal value of 0.25 (3 → 3.25 in the intermediate group) to a maximum of 1.0. (3 → 4; 3 → 4; 4 → 5 in the upper–intermediate and intermediate groups). Two more students in the upper–intermediate group could have also attained the top score (5 points) but failed to manage the word count. Similarly, all students of the intermediate group failed to compress the ST sufficiently. However, the scores for conveying the major content point varied in that group from four to five; three out of five students had a rather low number of near copying, and four out of five quite effectively used reporting verbs and discourse markers.

The comparison of the pre-and post-tests in summarizing in L1, presented in

Table 7. Comparison of results of pre-and post-test in summary writing in L1.

Student	Pre-Test	Post Test	Difference
VK	5	5	1
MS	5	5	1
EO	5	5	1
MG	5	5	1
DM	5	5	1
Mean	5	5	1
DP	5	5	1
KS	4	4.75	1.18
OS	4	5	1.25
DS	4	4.75	1.18
Mean	4.25	4.87	1.14
AT	4	4. 5	1.12
TK	4	4.5	1.12
AF	3	4	1.33
BH	3	4	1.33
AB	3	3.25	1.08
Mean	3.4	4.05	1.19
Mean			1.11
Standard deviation   0.3385

, shows that the students of upper-intermediate and intermediate groups enhanced their summary writing skills. The students in the advanced group confirmed their top scores received in the pre-test although, given the rating scale to assess, the latter did not require the use of reporting verbs, whereas in the post-test, using them was a must, the quality of their work tended to be more in the scholarly fashion.

Based on the data shown in Table 6 and Table 7, our hypothesis of the effective impact of summarization skills built in L2 on those in L1, i.e., transferability of such skills into academic L1, appears verified. Therefore, the second research question of this study is resolved too.

3.6. Limitations of the Study

The limitations of this action research lie in the quite moderate sample of the trainees, which prompts the prospects for further research—making the scope of the research broader and involving other teachers and/or researchers in it. The interdisciplinary character of the study suggests that the teachers of academic English could contribute to the development of teaching/assessment methodology of summarizing sources for a literature review in L1. Moreover, it would be beneficial to engage teachers of academic L1 in the pre-and post-testing to ensure the accuracy of the L1 TT evaluation.

3.7. Implications of the Study

The action research conducted by one teacher/researcher over quite a long period of time provides multiple insights into the processes of teaching/learning and assessing summary writing skills at the graduate level. The first of them echoes the implications made by Holmes [35], who states that “[g]raduate preparation programs should emphasize the critical importance of academic writing development for successful matriculation in advanced study” (p. 13).

Student mobility and internationalization of higher education corroborate the necessity of learning AE to excel in studies and future professional careers. Moreover, judging from the linguistics majors’ responses to the course, intensive work on academic skills should begin as early as the education programs permit.

Given a certain number of scholarly minded graduate students in Taras Shevchenko National University of Kyiv, it is also worth considering the development of a more advanced course of AE for the cohort of students engaged in doing and making public their research. In addition, if taught in parallel from the perspective of the relationship between two academic languages and cultures, the courses of AE and academic L1 could achieve a double effect.

It also seems crucial to consolidate the interdisciplinary links of the AE course with the academic cultures of the students’ target language countries (China, Japan, and South Korea). It is desirable that new research foci and novel research methodologies evolving in these regions should be considered in the course objectives and teaching materials.

Finally, the AE course has already been adapted to the training needs of student teachers. As a result, teaching L1 and L2 summary writing in senior secondary school is enthusiastically researched by them. Hopefully, these prospective teachers trained in AE will be able to promote the sustainability of the course discussed in this paper.

4. Conclusions

The action research project conducted by the author of this article has verified the efficacy of summary writing methodology consisting of three stages—from Stage One (teaching/learning to paraphrase) to writing summaries of popular science articles (Stage Two) and Results/Findings sections of RAs (Stage Three). Authentic texts of several academic genres are used across the consecutive stages of teaching and assessment, with the use of authentic academic tests being called by Cho and Brutt-Griffler [32] a helpful learning strategy.

All students engaged in the study developed summary writing skills, with the majority of them revealing the ability ‘to write not from sentences but from sources’ to rephrase the title of the article of Howard et al. [9]. The summaries written by the trained students demonstrate an effective use of reporting verbs and discourse markers to signpost the development of the TT as well as accurate referencing and a decrease in Near copying from STs. Obviously, the linguistic quality of summaries is dependent on the English language proficiency of the learners. Nonetheless, the acquisition of principles to summarize RAs helped the students to improve summary writing skills in Ukrainian, in other words, the skills built in English appeared transferrable into academic L1.

The prospect of further empirical research is to develop a feasible methodology of teaching and assessment by synthesising several sources in the university classroom. As was mentioned in Section 3.2, this work is already in progress. Although testing integrated (source-based) skills is being successfully researched in standardised testing [15,16,19], it is obvious that classroom-based methodology needs a specific focus on practicality beginning with selecting texts of ‘synthesisable’ content, establishing realistic word count of STs and TTs to updating rating scales.