Investigating the Effect of Process-based Instruction of Writing on the IELTS Writing Task Two Performance of Iranian EFL Learners: Focusing on Hedging & Boosting

Abstract

The instruction of metadiscourse markers to L2 writers has been recommended by some scholars to assist them in employing a certain tone in persuading readers. Nonetheless, there is a dearth of research investigating the effect of process-based instruction on hedging and boosting devices to L2 learners. To fill this gap, the present study aimed at investigating the effect of explicit instruction of hedging and boosting devices through a process-based approach on the achievement of L2 learners in academic writing. By comparing the results obtained through teaching 115 Iranian IELTS candidates who benefitted from (a) a process-based writing instruction with its focus on hedging and boosting, (b) mere process-based writing instruction, and (c) conventional writing instruction, the researchers found that the process-based approach to teaching writing with its focus on hedging and boosting resulted in a significant effect on L2 learners’ academic writing achievement. The findings of this study may have implications for EAP teachers, syllabus designers, and decision makers in designing EAP writing courses.

Keywords:

• Metadiscourse
• hedging
• boosting
• forms and functions of hedges
• EAP writing
• process-based instruction

PUBLIC INTEREST STATEMENT

This research aimed at testing the effect of a process-based approach on the achievement of two groups of Iranian IELTS candidates in the essay writing section of IELTS compared to another group of learners who were instructed via the conventional product-based approach. As for the processes-based group, there was a particular emphasis on metadsicourse instruction of hedging and boosting devices that are considered a characteristic feature of argumentative essays. The researchers hypothesized that the process-based instruction of writing through hedging and boosting devices does not have any significant effect on the IELTS writing Task 2 performance of EFL learners at the high-intermediate level. The results rejected this main hypothesis and supported such process-based instruction as beneficial.

1. Introduction

Nowadays, academic discourse is no longer considered as a mere objective presentation of informational content but a socially situated practice through which authors interact with readers. The interaction between the writers and their audience is possible through stance (Hyland, 1999, 2017). By having the essential knowledge of the conventions of academic writing, authors can develop a persuasive text which can not only gain the acceptance of discourse community but also raise persuasion in the audience which may be viewed as the ultimate goal of academic texts. Having a tentative approach in academic writing is often crucial to developing a persuasive text. Hedging and boosting have been generally accepted as useful metadiscourse devices and integral pragmatic features of academic writing to be used by writers in organizing their ideas in an efficient and acceptable manner. By using such tools, writers project their commitment or detachment in academic texts. Academic writers should learn to be “confidently uncertain” (Biook & Mohseni, 2014, 775) in presenting their ideas to academic community. However, the literature in the international context seems to indicate second language learners’ inclination to express unnecessary amounts of certainty and confidence while making claims. In addition, unlike their interest in using boosters, L2 students use fewer hedging devices in their academic writings (Davoodifard, 2006; Hyland, 2000; Hyland & Milton, 1997). This is particularly of concern to Iranian IELTS candidates who need to have a high level of performance in the Task 2 of writing, involving an argumentative essay (Alavi et al., 2020).

Exposing L2 learners to conventions of research writing and process-based instruction of hedging and boosting have been suggested by many scholars in the field (Kondowe, 2014; Nivales, 2010; Pritchard & Honeycutt, 2007) in an attempt to make students aware of the effectiveness of metadiscourse devices in presenting and organizing their claims. Teacher-designed lessons in process-based approach to writing instruction can motivate students to write and affect their emotions (Pritchard & Honeycutt, 2007).

From the theoretical perspective, a number of scholars in L2 writing have called for the integration of process-oriented and genre-based approaches to teaching writing to students in EFL contexts (e.g., Badger & White, 2000; Hyland, 2003b, 2004). By using process-genre approach, it is expected that students gain the necessary knowledge of textual features, process of writing, and the context to produce successful texts to deal with writing as a complex activity (Archibald & Jeffery, 2000). However, to the present researchers’ best knowledge, there has been insufficient empirical research investigating the effect of the process–genre approach to teaching writing in L2 contexts, and more specifically the process-based instruction of IELTS writing Task 2 to Iranian EFL learners. At the same time, one of the main features of academic writing, which is particularly highlighted in the context of proficiency tests, encompasses the use of hedging and boosting elements. To this end, the present study aimed at investigating the effect of process-based instruction of writing with its focus on hedging and boosting on the IELTS writing Task 2 achievement of Iranian Intermediate EFL learners.

2. Review of literature

To have a more profound perception of hedging, it is necessary to clarify the concept of metadiscourse. Williams (1981, p. 226) considers metadiscourse as “writing about writing”. Following the same idea, Vande Kopple (1985) proposes that “discourse about discourse or communication about communication” (p. 83) is what metadiscourse includes. Vande Kopple (1997) further defined metadiscourse as the discourse members of a community choose to avoid pointing to a specific reference yet to help the audience to combine, interpret, and asses certain attitudes in the direction of some references themselves. Hyland (2005) defines metadiscourse as “the cover term for the self-reflective expressions used to negotiate interactional meanings in a text, assisting the writer or speaker to express a viewpoint and engage with readers as members of a particular community” (p. 37). Metadiscourse can be interpreted in a variety of ways. The range of different categories of metadiscourse ranges from mere marks such as punctuations and parentheses to pronunciation features such as tone and stress. It further includes sentence and clause level elements (Hyland, 1999, 2005).

Metadiscourse can be interpreted in a variety of ways, and hedging tools are subscribed under some metadiscourse classifications (Crismore et al., 1993; Hyland, 2005; Vande Kopple, 1985). The concept of hedging has been defined significantly differently by various researchers (Hyland, 1998c), and despite the frequent use of hedging over the past four decades or so, no clear-cut definition seems to be found in the textbooks and seminal articles (Varttala, 2001). Assigning a linguistically oriented analysis to hedges, Lakoff (1973) focused on the issue of relating natural phenomena to natural linguistic ideas and defined hedging as “words whose job is to make things fuzzier or less fuzzy. Crompton (1997, p. 281) suggests this definition of hedging as “items of language which a speaker uses to explicitly qualify his/her lack of commitment to the truth of a proposition he/she utters”. Literally, conveying a sense of restrict or limit, hedging refers to those statements that decrease the strength of what speakers and authors utter (Heng & Tan, 2000). A more commonly accepted definition by Hyland (1998c) describes hedges as linguistics devices to show the degree of commitment to a claim or basically evading it. Yet, the term hedging is comprehensive, multi-functional, and complex, as it is connected with other aspects such as politeness, indirectness, modality and vagueness. The complexity of its definition shows itself in the related research. Schro¨der and Zimmer (1997) considered the research on hedging as a multifaceted cross-disciplinary area that covers philosophical, pragmatic, logical, linguistic, and semantic issues. Nowadays, hedges are viewed more in a pragmatic sense rather than a purely semantic way.

In writing academic papers, researchers also apply boosters, as crucial tools, to share their results with readers and convince them to support the stance of the writer. Boosters strongly support writers’ statements, and this can contribute to readers’ conviction to the issue (Yagiz & Demir, 2015). In fact, writers use boosters in order to make an impact on the response of the audience to whom the paper is addressed and persuade them to support the conclusions drawn from the study (Vázquez Orta & Giner, 2009). According to Hyland (1998b), boosters are used by writers to prove their engagement and commitments to their claims. He further asserts that boosters are certain devices in the service of persuasion to highlight and modify mutual points of view among the same discourse groups

Researching hedging and boosting in written EAP/ESP discourse has gained a lot of prominence over the past three decades. A review of the related literature reveals that researchers have merely attempted to conduct cross-cultural, cross-disciplinary, or cross-linguistic studies in order to investigate the use of the types and frequency of hedges and boosters in research articles written by native and non-native speakers of English (e.g., Aptoula & Bayyurt, 2020; Crismore et al., 1993; Dafouz-Milne, 2008; Hyland, 1998a, 1998b, 1998d, 1999; Serholt, 2012; Yang, 2013). Other researchers have also investigated the effect of teaching metadiscourse markers on EFL students’ writing development and examined the effect of instruction on the use of hedging and boosting devices in academic papers (e.g., Algi, 2012; Alward et al., 2012; Hyland, 1998b, 2004; Kim & Lim, 2015; Nikula, 1996; Sarani & Talati-Baghsiahi, 2017). All these studies demonstrate the importance of exposing novice writers to the conventions of academic writing and emphasize the direct instruction of hedging and boosting lessons in academic writing courses, which can be significantly helpful in developing learners’ writing skills in academic settings.

Writing is a recursive and not a linear, process of creating a text (White & Ardnt, 1997). It entails all stages from prewriting and drafting to publishing. Writing instruction, then, is recommended to take a process approach which, according to Zamel (1987), is suggested to be a process which entails finding out and reformulation of writers’ ideas to convey meaning. As a creative process which offers many techniques, process approach to writing is a practice that gives more importance to meaning than to form and engages learners in sharing and collaborating while being guided by a facilitator’s constant evaluation and feedback (Ariza Martínez, 2005; Zhou, 2015).

A review of related research reveals the effectiveness of process approach to writing on learners’ writing achievement in general (Allen et al., 2019; Ferretti & Graham, 2019). By comparing process and product approaches to writing instruction, Tong (2007) concluded that the process approach was more effective in developing the writing proficiency of the learners. In a similar study Bayat (2014) investigated the effect of the process-based approach on writing success and anxiety of university students. The results indicated that the learners made significant progress both in higher level skills of developing ideas and organization and in written expressions such as cohesion and grammar. The results also demonstrated that the process-based approach decreased students’ level of writing anxiety in a significant way. This approach also fostered learners’ enthusiasm to engage in writing task (Westervelt, 1998).

As this review indicates, the bulk of research on hedging and boosting has focused either on hedging and boosting type and frequency (Bacang et al., 2019; Bondi, 2020; P. Crosthwaite et al., 2017; Nekouizadeh et al., 2020) or the effect of teaching them on learners’ achievements in academic writing (Ho & Li, 2018; Salmani-Nodoushan, 2016). It also indicates that no similar study has ever focused on the effect of teaching hedging and boosting through a process-based approach on the achievement of IELTS candidates in writing Task 2. The present study aimed to fill this gap by using Hyland (2004) model of hedges and boosters as the theoretical framework of the study.

This research relies on Hyland model of hedges and boosters Hyland (1996b) as the reference for conducting this research. Hyland (2004) categorizes five types of hedges and boosters. Type 1, also called Tentative Verbs and Modals, includes hedges such as may, might, could, seem, suggest, appear, and phrases that contain hedges such as “seems to suggest”. Should is treated as a type 1 booster. Type 2, also called Tentative Adjectives and Adverbs, includes hedges such as, possibly, likely, and probably. Certainly and definitely are treated as Type 2 boosters. Hyland (2004) also suggests that nominalized verbs be categorized under Type 2. Distancing phrases, the use of impersonal third person, and the Unnamed are categorized under Type 3. Solidarity features such as “it is a fact” and “it is well known” are categorized under Type 4. Also, Hyland (2004) suggests that rhetorical questions signified by words such as “can, may, would” fall under Type 4. Under Type 5 of hedges and boosters, Hyland (2004) suggests Self-mention reference. This category includes pronouns and nouns such as I, we, researcher, writer or author. The model of hedges and boosters proposed by Hyland (2004) has been the resource to be considered in the process-oriented writing instruction in the present study.

3. Research questions

Hedging and boosting devices find special place in writing Task 2 of IELTS which requires candidates to write an argumentative essay. Given the hypothesis that process-based approach can help EFL learners develop writing skills, this study strove to compare the Iranian EFL learners who were taught writing through a process-based approach with those who were taught writing through the same approach with its focus on hedging and boosting through the following questions which were next changed into null hypotheses.

1. Does the process-based instruction of writing have any effect on the IELTS writing Task 2 performance of EFL learners at the high-intermediate level?

2. Does the process-based instruction of writing through hedging and boosting devices have any effect on the IELTS writing Task 2 performance of EFL learners at the high-intermediate level?

3. Is there a statistically significant difference between the IELTS writing Task 2 performance of EFL learners taught through the process-based instruction of writing and those instructed through the process-based instruction with its focus on hedging and boosting devices?

4. Is there a statistically significant difference between the IELTS writing Task 2 performance of EFL learners taught through the process-based instruction of writing and those instructed conventionally for IELTS?

5. Is there a statistically significant difference between the IELTS writing Task 2 performance of EFL learners taught through the process-based instruction of writing with its focus on hedging and boosting devices and those instructed conventionally for IELTS?

4. Method

4.1. Research design

A quantitative research design was used to explore the research questions of the study. In order to test the null hypotheses, a quasi-experimental model was utilized due to such limitations as randomization or sample size. In order to approximate the standards of quasi-experimental design, this study benefitted from (a) the presence of a control group beside the experimental groups and (b) the administration of a pretest for the recognition of differences among groups at the outset of the study. There were two experimental groups and one control group in this study. The participants of each group were selected according to their scores on a reliable, valid pretest. One experimental group benefitted from the process-based writing instruction, and the point of departure for the other experimental group was the explicit instruction of hedging and boosting devices. The control group, on the other hand, was taught writing conventionally through topic discussion and the feedback they received on their essays.

4.2. Participants

The sampling of the present study was based on availability. It initially comprised 142 Iranian EFL learners who had registered for IELTS preparation courses in Farazmon IELTS institute, the center that exclusively holds Academic IELTS courses in Tehran. One hundred and fifteen learners out of 142 candidates were ultimately selected as the participants of the study. One of the researchers of this study and two official IELTS examiners served as raters in this study.

4.3. Instruments

A number of instruments were used for pretest, posttest, and instruction phase during the course of this research. First, the researchers used a complete, two original IELTS mock tests as the pretest and posttest of the study. In both pretest and posttest, the standard time limit of the IELTS test were observed which is 40 minutes for Task 2. Since the focus of the study was on the performance of the participants on Task 2 of the writing module of IELTS exam, the examiners were required to score Task 2. Task 2 in the writing section encompassed an argumentative topic on a recent issue of interest over which the candidates were supposed to write at least a 250-word essay.

The researchers then utilized researcher-designed lesson plans for each of the two experimental gro and control groups. Due to the point of focus of this research, the researcher could not use any previously available lesson plans or materials since they did not include parts on hedging and boosting to meet the demands of the study. The lesson plan for each group was compiled out of the following sources: a) The Official Cambridge Guide to IELTS for Academic & General Training (Cullen et al., 2014), b) Cambridge Practice Tests for IELTS (2019), c) Cambridge Mindset for IELTS Series (Archer, 2018), and d) Advanced Writing (Birjandi et al., 2004). In addition to the researcher-designed lesson plans for each group, the first experimental group was provided with a booklet based on Hyland (2004) model of hedges and boosters. Furthermore, the researchers provided a pamphlet for the control group which included a variety of samples collected from different IELTS authentic sources. The samples were roughly categorized into five types including (a) opinion or argumentative essays, (b) discussion essays (opposing opinions), (c) advantage/disadvantage essays, (d) direct questions, and (e) solution essays. The samples were collected from the Official Cambridge guide to IELTS (Cullen et al., 2014), and the other above-mentioned sources, which can provide enormous valid samples for IELTS examination.

4.4. Data collection

The procedure of the present study began with taking an original mock test from IDP IELTS center in Australia. This step assured the homogeneity of participants in terms of their language proficiency. The topic for the writing section was not publicly available in the practice material available in the market to assure its originality and authenticity. The pretest was taken by the learners who had registered for IELTS preparation courses in Farazmon language institute. The criteria including the time limit and format of the test booklet were closely followed. The time limit for the whole test was almost 2 hours and 43 minutes, out of which 40 minutes were allocated to Writing Task 2 section.

Next, using the band descriptors of IELTS test for the writing sections (Task 2), the researcher and the two official IELTS examiners scored the essays of the participants to establish inter-rater reliability of the scores. In this way, the researcher was provided with three scores for each candidate, the average of which was used to establish homogeneity in terms of scores. In this research project, the intermediate level was defined in terms of the band scores of 4–5.

Once the scores of the pretest were obtained, 115 learners out of 140 candidates were selected as the participants. The age range of the final group of ESL learners ranged from 20 to 44 including both males and females. Their level of education ranged from diploma to master’s degree. These learners were later required to provide a detailed profile about their English learning backgrounds. This extra procedure along with the result of the pretest ensured the homogeneity of participants’ level of language proficiency. It should be noted that the participants were informed that their collected essays were going to be used for research purposes, and they expressed their agreement by signing the prepared written consent forms. Then, the participants were initially assigned randomly to one control and two experimental groups. Complete demographic information was collected for each group in terms of age, sex, level of education, and major. The groups were later counterbalanced according to the obtained profile. The number of participants for each experimental group was 38 and 39 for the control group.

The instruction phase of this study for the experimental groups was based on process-based writing instruction in the form of two weekly sessions, each lasting 90 minutes, which totally amounted to 30 sessions. The first experimental group received the treatment that included researcher-designed lesson plans. The lesson plans comprised various sections including different types of argumentative writing, a wide range of sample questions and texts, and the required formats of writing for each specific types of writing. In addition, the group benefitted from the booklet designed based on Hyland (2004) model of hedges and boosters. The participants were introduced to the concepts of hedging and boosting. They were given elaborate short writing tasks on using different hedging and boosting devices and were required to ultimately use them in their essays. The steps of prewriting, writing, revising, and editing were subsequently as the systematic process writing procedure.

Similar to the first experimental group, the second group received process-based writing instruction. The only point of departure was the elimination of teaching hedging and boosting devices from their syllabus. Pre-writing, writing, and post-writing phases were meticulously followed. This group also benefitted from the same lesson plan designed for the first experimental group with hedging and boosting booklet not included in the syllabus. The instructor (one of the researchers herself) undertook the instruction for both experimental groups. At the same time, the writing instruction of the control group included a variety of samples collected from different IELTS authentic sources.

In the post-instruction phase, the participants were required to sit for a mock IELTS test. The researcher made sure that the issues of time limit, venue and administration were under strict control mimicking the real situation of IELTS examination. The assessment procedure was the same as the one for the pretest. Three IELTS raters scored the writing papers, and the average of the three scores was used in the data analysis.

It should be noted that the researcher shouldered all the responsibility for the instruction in this research for all the three experimental and control groups. This was to ensure the same quality of writing instruction for all groups so that the results would be eventually comparable.

5. Results

The present study aimed at investigating the effect of the process-based writing instruction, the process-based writing instruction with its focus on hedging and boosting, and the conventional writing instruction on the improvement of the EFL learners’ performance on IELTS writing Task 2. In order to explore the research questions raised in this study, repeated measures ANOVA was run. Since the absolute values of the ratios of skewness and kurtosis over their standard errors were lower than 1.96, it was concluded that the data collected through this study did not show any severe departure from normality. Repeated measures ANOVA requires that the groups’ variances on pre- and posttests of writing be roughly equal. In other words, it assumes homogeneity of variances.

Based on Levene’s Test of Homogeneity of Variances for the pre- and posttests of writing (Table 1), it was concluded that there were not any significant differences between the three groups’ variances on pretest (F (1, 111) = .092, p = .913) and posttest (F (1, 111) = .166, p = .847) of writing. The non-significant results of the Box’s test for the homogeneity of covariance matrices (Table 2) also indicated that the assumption of homogeneity of covariance matrices was retained. Repeated measures ANOVA also assumes that the variances of the differences between any two dependent variables are roughly equal, i.e. the assumption of sphericity. If there are two tests in a study, as is the case here, the probability for the Mauchly’s test of sphericity cannot be computed (Table 3) (Field, 2018).

Table 1. Levene’s test of homogeneity of variances: pretest and posttest of writing

		Levene statistic	df1	df2	Sig.
Pretest	Based on mean	.162	2	111	.851
	Based on median	.092	2	111	.913
	Based on median and with adjusted df	.092	2	107.340	.913
	Based on trimmed mean	.099	2	111	.906
Posttest	Based on mean	.377	2	111	.687
	Based on median	.166	2	111	.847
	Based on median and with adjusted df	.166	2	94.692	.847
	Based on trimmed mean	.342	2	111	.711

Table 2. Box’s test of equality of covariance matrices; pretests and posttest of writing

Box’s M	5.846
F	.949
df1	6
df2	307,077.231
Sig.	.458

Table 3. Mauchly’s test of sphericity

Within- subjects effect	Mauchly’s W	Approx. Chi-square	df	Sig.	Epsilon
					Greenhouse–Geisser	Huynh–Feldt	Lower bound
Writing	1.000	.000	0	.	1.000	1.000	1.000

A repeated measures ANOVA plus simple effect analysis was run to compare the means of the process-based writing instruction, the process-based writing instruction through hedging and boosting, and the conventional method on the pre- and posttests of writing in order to (a) prove that the groups were homogenous in terms of their writing ability prior to the administration of the interventions and (b) probe the five research questions raised in this study.

Table 4 demonstrates the main results of between-subjects effects. These results cannot be used to answer any of the research questions, which can only be probed through simple effect analysis. Based on the results displayed in Table 4, it can be concluded that there was a significant difference between three groups’ means on overall writing test (pretest and posttest).

Table 4. Tests of between-subjects effects; pretest and posttest of writing by groups

Source	Type III sum of squares	df	Mean square	F	Sig.	Partial Eta Squared
Intercept	5973.197	1	5973.197	12,013.325	.000	.991
Group	32.612	2	16.306	32.795	.000	.371
Error	55.191	111	.497

Table 5 displays the results of within-subjects effects. Based on the results, it can be concluded that there was a significant difference between overall means on pretest and posttest of writing, irrespective of types of treatments, and there was a significant interaction between types of treatments and pretest and posttest of writing.

Table 5. Tests of within-subjects effects: pretest and posttest of writing by groups

Source		Type III sum of squares	df	Mean square	F	Sig.	Partial eta squared
Writing	Sphericity assumed	74.123	1	74.123	585.761	.000	.841
	Greenhouse–Geisser	74.123	1.000	74.123	585.761	.000	.841
	Huynh–Feldt	74.123	1.000	74.123	585.761	.000	.841
	Lower bound	74.123	1.000	74.123	585.761	.000	.841
Writing × Group	Sphericity assumed	33.831	2	16.916	133.677	.000	.707
	Greenhouse–Geisser	33.831	2.000	16.916	133.677	.000	.707
	Huynh–Feldt	33.831	2.000	16.916	133.677	.000	.707
	Lower bound	33.831	2.000	16.916	133.677	.000	.707
Error (Writing)	Sphericity assumed	14.046	111	.127
	Greenhouse–Geisser	14.046	111.000	.127
	Huynh–Feldt	14.046	111.000	.127
	Lower bound	14.046	111.000	.127

Table 6 displays the descriptive statistics for the means of groups who were taught writing through process-based, process-based with its focus on hedging and boosting and control groups on pre- and posttests of writing. The results indicated that all the three groups had almost the same means on the pretests of writing, whereas the group receiving the process-based writing instruction with its focus on hedging and boosting had the highest mean on the posttest of writing (M = 6.55), which was followed by the process-based (M = 5.81) and the control (M = 4.60) groups.

Table 6. Descriptive statistics; overall means on pretest and posttest of writing by groups

Group	Writing	Mean	Std. error	95% Confidence interval
				Lower bound	Upper bound
Process-basedwith hedgingand boosting	Pretest	4.513	.078	4.360	4.667
	Posttest	6.553	.102	6.351	6.755
Process-based	Pretest	4.592	.078	4.438	4.746
	Posttest	5.816	.102	5.614	6.018
Control	Pretest	4.539	.078	4.386	4.693
	Posttest	4.697	.102	4.495	4.899

Table 7 demonstrates the results of the first simple effects analysis. The results compare the three groups on the pretests of writing first and then on posttests of writing. Based on these results, it can be concluded that the three groups were homogenous in terms of their writing ability prior to the administration of the interventions. The descriptive statistics shown in Table 6 and the simple effect analysis displayed in Table 7 can be used to probe the first three research questions as follows: (a) the experimental group 1 receiving the process-based writing instruction with its focus on hedging and boosting significantly outperformed the group that received process-based instruction only on the posttest of writing, (b) the experimental group 2 receiving the process-based writing instruction only significantly outperformed the control group on the posttest of writing, and (c) the experimental group 1 receiving the process-based writing instruction with its focus on hedging and boosting significantly outperformed the control group on the posttest of writing.

Table 7. Simple-effect analysis; comparing groups on pretest and posttest

Writing	(I) Group	(J) Group	Mean difference (I − J)	Std. error	Sig.	95% Confidence interval for difference
						Lower bound	Upper bound
Pretest	Process-Based	Hedging	.079	.110	.473	−.138	.296
		Control	.053	.110	.632	−.165	.270
	Control	Hedging	.026	.110	.811	−.191	.244
Posttest	Hedging	Process-based	.737*	.144	.000	.451	1.023
		Control	1.855*	.144	.000	1.569	2.141
	Process-Based	Control	1.118*	.144	.000	.833	1.404

*Mean difference is significant at the .05 level.

Table 8 displays the results of simple effect analysis investigating the improvements in the writing skills of each group from pre- to posttests. These results and the descriptive statistics displayed in Table 6 can be used to answer the last two research questions. Thus, the group receiving process-based instruction only showed a significant improvement in their means from pretest to posttest (mean difference = 1.22, p = .000). In addition, the group receiving process-based instruction with its focus on hedging and boosting showed a significant improvement in their means from pretest to posttest (mean difference = 2.03, p = .000). Although of no concern in this study, the results illustrate that the control group did not show any significant improvements in their means from pre- to posttests (mean difference = .158, p = .056).

Table 8. Simple effect analysis; mean improvement from pretest to posttest by groups

Group	(I) Writing	(J) Writing	Mean difference (I − J)	Std. error	Sig.	95% Confidence interval for difference
						Lower bound	Upper bound
Hedging	2	1	2.039*	.082	.000	1.878	2.201
Process-based	2	1	1.224*	.082	.000	1.062	1.385
Control	2	1	.158	.082	.056	−.004	.320

*.Mean difference is significant at the .05 level.

6. Discussion

The present study aimed to examine the effect of process-based approach to writing and the direct instruction of hedging and boosting on the performance of Iranian IELTS candidates in writing Task Two. The results of the study indicated that L2 learners who benefitted from the process-based approach to writing instruction performed better on writing Task 2 of IELTS exam than those who received conventional product-oriented instruction. Additionally, the L2 learners who experienced the explicit instruction of hedging and boosting through process-based approach to teaching writing obtained higher scores in writing Task 2 of IELTS exam than those who were merely subjected to process-oriented approach.

The results of this study are in line with certain findings by other researchers in the field. Research has shown that process-based writing instruction has significant effects on writing achievements of L2 learners (e.g., Goctu, 2017; Tong, 2007). Product-oriented approach to writing instruction is believed to be a problem-solving task, which inhibits learners’ creativity for writing various contents in different contexts (Sarhady, 2015). Process-based approach, on the other hand, considers writing as a social reader-based activity (White & Ardnt, 1997), moving from ego-centrism to target audiences (Pritchard & Honeycutt, 2007) and assists L2 learners to develop a personal approach to writing (Ariza Martínez, 2005; Goctu, 2017). As Pritchard and Honeycutt (2007) put it, “in the process approach, not every prewriting activity will lead to a final draft, but students’ understanding of the movement from first idea to finished product is an essential feature” (p. 30). This approach not only develops L2 learners’ higher level skills such as planning and organizing ideas but also assists them in advancing spelling, punctuation, choice of vocabulary, capitalization, and grammatical structures (Sarhady, 2015). The results of the present study can consolidate the claim that process-based approach to teaching writing enhances L2 learners’ academic writing achievements.

The findings of this research indicated that exposing L2 learners to the direct instruction of hedging and boosting devices could enrich their language repertoire and result in great enhancement in their academic writing. The participants of this study benefited from a mixture of hedging and boosting devices, leading to significant effects on their academic writing achievements. This is justifiable in the light of the fact that one of the distinguishing features of academic language, especially as it pertains to written language, is the extensive use of hedging devices. It could be mainly due to the flexibility and non-commitment mood to the propositions expressed in a context that such devices provide for the researchers and academicians (Crismore et al., 1993). Through hedging devices, academic writers imply caution and uncertainty of the claims they make and they present their statements without being precise (Hinkel, 1997; Hyland, 1996b, 2004). As Kondowe (2014) puts it, writers use more hedging in order to “reduce the risk of opposition, being precise in reporting results, but also as a means of being polite and accommodative in their attempt to get their thesis approved and have them passed” (p. 220). Hedging can assist writers with taking a cautious stance in criticizing other research and showing concerns to their works (Samaie et al., 2014).

Conversely, boosters strongly support writers’ statements and aim at convincing the readers to support the conclusions drawn from the study by the researcher (Vázquez Orta & Giner, 2009; Yagiz & Demir, 2015). As Hyland states, boosters “allow writers to project a credible image of authority, decisiveness, and conviction in their views” (Hyland, 1998b, p. 238). While being assertive demonstrates authors’ stance in the claims they make in academic writing, some researchers argue that exaggerations need to be avoided in scientific writings as such expressions give an unnatural tone to the text (Hinkel, 1997; Kazemi, 2016). Some other researchers (e.g., Demir, 2017) acknowledge that discourse community respects a balance between being tentative and assertive in writing academic prose; hence, illuminating appropriate use of boosting devices in academic writing is necessary for L2 writers in order to convey the required sense of persuasion and conviction.

The significant improvement in the writing achievement of the ESL learners who benefited from the explicit instruction of hedging and boosting devices is congruent with a number of previous research reports. Such research findings have revealed that compared to native English-speaking writers who express less certainty and commitment in writing academic texts, non-native English language learners use considerably less hedging devices in L2 written texts (e.g., Davoodifard, 2006; Hu & Cao, 2011; Hyland & Milton, 1997; Samaie et al., 2014). Native English-speaking writers markedly utilize more hedging tools to weaken and mitigate their claims, while non-native English speakers use metadiscourse markers in favor of strengthening their statements. Given that academic discourse community requires a writing style which incorporates a balance between being tentative and assertive (Demir, 2017), researchers recommend L2 teachers to include lessons of hedging and boosting devices in order to acquaint novice writers with the conventions of research writing (Alward et al., 2012; Hyland, 1998c; Kim & Lim, 2015; Kondowe, 2014). Moreover, explicit instruction of hedging devices has been found to be an effective method in increasing ESL learners’ both pragmatic and linguistic knowledge (Sarani & Talati-Baghsiahi, 2017).

In writing academic and scientific prose, authors and readers may share different norms and expectations. This is the reason that socially and culturally specified conventions for being tentative have been considered for academic writing. However, the existing literature has merely shown that non-native English writers are inclined to show a stronger degree of commitment to their statements, while native English-speaking writers use proper frequency and types of hedging devices to mitigate their claims in academic writing (Hyland & Milton, 1997; Yang, 2013). Language learners vary in the amount of using hedging and boosting devices in terms of types and frequency due to linguistic, sociocultural, and pragmatic reasons. Some researchers believe that there is high level of correlation between the level of proficiency and the use of hedging devices among ESL learners as L2 learners employ a very limited range of hedges due to limited lexical repertoire (Hyland, 2002; Nikula, 1996). In addition to the level of language proficiency, some other researchers argue that culturally based rhetorical conventions can result in differences in the occurrence of hedges and boosters in native and non-native English writers (Hu & Cao, 2011). Scientific journals of each culture have specific rhetorical norms and discursive preferences rooted in their respective culture, thus, leading writers to employ their culturally preferred rhetorical strategies (ibid). Different paradigms of scientific inquiry in each culture encourage writers to write in certain ways, i.e., to write in convincing and assertive way of making scientific claims or favor the use of hedges to moderate knowledge claims and lessen the author’s commitment toward the statements (Yang, 2013). Further research may clarify the way cultural background, language proficiency, epistemological beliefs, and culturally based rhetorical conventions may affect the use of hedging and boosting devices in academic writing. The present research, in its limited scope, concluded that intense instruction of hedges and boosters through a process-based approach to writing instruction could improve L2 writers’ achievements. Avid researchers can design investigation that focus on other metadiscourse devices that are of relevance in academic writing.

7. Conclusion

The findings of the present study highlight the effect of the process-based approach to writing instruction on learners’ academic writing achievements. Furthermore, the explicit instruction of hedging and boosting devices in a sequenced and integrated way through a process-based approach to writing has been proven to be significantly helpful in developing L2 learners’ academic writing skills. Employing a direct and explicit method of teaching hedges and boosters through a process-based approach not only enhanced the linguistic knowledge of L2 learners but also acquainted them with the expected rhetorical conventions of discourse community in academic writing and highlighted the importance of hedging their statements and claims to convince their readers. Employing hedging devices shows uncertainty, caution, objectivity, and politeness in academic writing (Sarani & Talati-Baghsiahi, 2017). The explicit instruction of hedges and boosters can differentiate between formal and informal hedges and provide the L2 learners with the formal intensifiers and hedges that are more appropriate to be used in academic texts (Kazemi, 2016). Additionally, it helps L2 learners to differentiate between intentional hedges and hedges that are inherent in the speech acts (Lewin, 2005).

The results of this study have implications for English for academic purposes course designers to include explicit instruction of hedges and boosters in their curriculum. Additionally, material developers may take advantage of the findings of the present research and add sections of lessons on the types of hedging and boosting devices and highlight their significance in persuading readers. Furthermore, L2 teachers could provide the students with a teacher-made pamphlet of metadiscourse markers and enjoy learners’ improvements in writing academic texts.

Additional information

Funding

The authors received no direct funding for this research; This study was self-funded. [No Number];

Notes on contributors

Mojgan Firoozjahantigh

Mojgan Firoozjahantigh is a Ph.D. candidate in TEFL at Islamic Azad University, Qaemshahr, Iran. She has taught different EFL courses over the past 18 years. Her research interests include teaching second language skills, cultural studies in EFL and teacher training.

Ebrahim Fakhri Alamdari

Ebrahim Fakhri is an assistant professor at the Islamic Azad University. He has published papers in international journals including Francis & Taylor, Routledge: International Journal of Listening and ReCall journal published in Cambridge University Press. His research domains are bilingualism, listening comprehension, and listening in relation to other language skills.

Amir Marzban

Amir Marzban is an associate professor of TESOL at Islamic Azad University, Qaemshahr branch. His research interests include conversation analysis, L2 reading & writing, CALL, and teacher education. He has published in both Iranian and International journals and also has presented in many international conferences.