Corpus can be used as a mediational learning tool (Vygotsky, 1978). Vast compilations of authentic lines selected from a wide range of authentic texts can be created to target specific aspects of language learning (Bernardini, 2000). The terms, indirect DDL and direct DDL, have been used by Chambers (2007).
Even though both direct and indirect DDL have been shown to be useful approaches, it has been asserted that direct DDL is daunting for some students, and as a consequence, lower-level and novice learners favor indirect DDL over the direct approach (Boulton, 2010). Moreover, indirect DDL has been found to be able to be used instead of DDL without reducing the effects of corpus-based instruction on learning and promoting a low-stress and highly environment for learners (Daskalovska, 2015). It has been suggested that direct or indirect should be selected with regard to the time available in the class, learner variability, and preferences since the three approaches entail the same cognitive process (Chujo et al., 2012). From another angle, there is no single right way to use direct, indirect, or combined DDL in educational settings (Cobb & Boulton, 2015).
The effect of DDL and corpus-based instructions on various language skills and sub-skills has been investigated. Research has indicated the positive effects of DDL on phraseology learning (Ackerley, 2017), L2 lexical acquisition (Lee et al., 2019; Lee et al., 2020), collocation (Akinci & Yildiz, 2017; Chang, 2023; Saeedakhtar et al., 2020; Vyatkina, 2016), and technical words (Chang, 2023). Other studies have shown improvements in students’ writing following the use of corpora (Chen & Flowerdew, 2018; Crosthwaite, 2020; Muftah, 2023; Yoon & Hirvela, 2004; Yoon & Jo, 2014). Other DDL-based investigations have shown improvements in the use of logical connectors (Cresswell, 2007), abstract nouns (Huang, 2014), and hedging in students’ writing (Sun & Hu, 2020). Furthermore, the use of DDL has been proven to enhance students’ grammar (Chujo et al., 2012; Lin, 2021; Smart, 2014; Zhong & Wakat, 2023), and the use of adverbial clauses (Kilickaya, 2013). It has also been shown that EFL teachers had overwhelmingly positive experiences of integrating DDL into EFL teaching (Bennett & Dhonnchadha, 2023).
Notwithstanding the demonstrated efficacy of DDL, this approach has been marginally utilized (Breyer, 2009; Boulton, 2017). The lack of popularity and the neglect of DDL may stem from the fact that the use of this approach has very rarely been a part of teacher training courses. Should teacher students use this approach, they can better understand the challenges confronting their students when the students get exposed to corpora (Breyer, 2009; Boulton, 2017). On the other hand, this unpopularity can be rooted in teachers’ beliefs since many find DDL tough and challenging (Boulton, 2010). DDL has been said to be time-consuming due to two main reasons. Firstly, it takes students a long time to study corpus-based materials in schools, which is not feasible regarding school time limits (Lin & Lee, 2015). Secondly, it takes teachers a long time to prepare DDL materials in the case of indirect DDL, which encourages them to use prepared materials (Sun & Hu, 2020).
Finally, there is the detrimental effect of examinations. In some contexts, including Persian high schools, students have to take national entrance exams to be admitted to universities. The washback effect of such public examinations could be another significant factor contributing to the widespread use of the traditional deductive approach and the low popularity of novel approaches such as DDL (Lin & Lee, 2017; Messick, 1996). In other words, the washback effect fosters schools and teachers to prefer what to teach over how to teach it (Chen, 2002).
Literature review of research comparing DDL with other approaches to language learning indicates contradictory results. Some studies have corroborated the effectiveness of DDL (Ackerley, 2017; Chen & Flowerdew, 2018; Huang, 2014; Lee et al., 2019; Smart, 2014; Vyatkina, 2016), some have shown the superiority of explicit instructions to implicit inductive DDL (Akinci & Yildiz, 2017; Celik, 2011), some have demonstrated no clear or remarkable effect of DDL (Cresswell, 2007; Smart, 2012), and some have found no difference between the traditional approach and DDL concerning students’ learning (Tian, 2005; Lin & Lee, 2017). Likewise, some studies have proven TDA to be more effective than or equally effective as DDL. For instance, Akinci and Yildiz (2017) concluded that the TDA participants outperformed the DDL ones on verb + noun collocations while Celik (2011) proved that the TDA and DDL learners were not statistically different from one another regarding academic words and prepositional phrases. However, the DDL learners had higher scores on retention tests. Lin and Lee (2017) as well as Tian (2005) demonstrated that the DDL group was no different from the TDA group in terms of grammar test scores while each group made improvements during the treatment.
Perusing the literature on the instruction of cohesive devices to EFL learners, it can be indicated that even advanced EFL learners struggle with learning cohesive devices (Granger & Tyson, 1996; Altenberg & Tapper, 1998). Many researchers, studying cohesive devices, are of the opinion that Asian EFL learners find learning English cohesive devices challenging (Bui, 2022; Crossley et al., 2016; Eng et al., 2020; Hammad, 2016). Cohesive devices are a frequent source of error among EFL learners because learners usually have partial and inaccurate knowledge of the meanings and uses of connectors used in English although they might generally recognize cohesive devices upon seeing them in texts (Cresswell, 2007).
In light of the contrasting findings about DDL and TDA as shown in the extant literature, the low popularity of this approach, and EFL learners’ major difficulties with using cohesive devices, it is evident that there is an urgent need to conduct more research on this innovative approach, and its effectiveness in the instruction of cohesive devices. Accordingly, this study aims to respond to the challenge of comparing and contrasting the effects of DDL with those of TDA on the learning of cohesive devices with EFL high school students. The research questions are as follows:
1. Do both TDA and indirect DDL participants improve the knowledge of cohesive devices from pre-test to delayed post-test?
2. What are the differences between TDA and DDL groups in terms of learning at post-test and learning retention at delayed post-test?
3. What are indirect DDL participants’ attitudes toward the use of indirect DDL?