1. Introduction

Some of the researchers in the field of second language acquisition (SLA) drew attention to the essential role that input plays in the process of second language learning. For instance, Krashen (1987) pointed out to the crucial role that comprehensible input plays in the process of second language acquisition.  In Input Processing Model, Van Patten (2004) categorizes input processing as the early stage of SLA. This stage involves the transition process from input to intake. By the way, Swain proposed the notion of the comprehensible output hypothesis in 1985. According to this hypothesis, comprehensible input may not be sufficient for developing communicative competence. Therefore, there is a need to push second language learners toward producing comprehensible output. After that, abundant studies were done in this area to scrutinize the fundamental role that output plays in second language acquisition (e.g., Nasaji & Tian, 2010; Mahmoudabdi et al., 2015; Alabdullah, 2021). 

Three benefits of output which are listed by Gass and Selinker (2001) are: 

1. receiving feedback for the confirmation of hypotheses in the course of conversation; 
2. testing hypotheses about the grammatical structures and meanings of the L2;
3. increasing automaticity in interlanguage production; 

As for, research on the efficacy of output-based tasks was limited to some types of tasks, for instance, oral and written output tasks (Rastegar & Safari, 2017), reconstruction editing task (Ganji & Nejad Ansari, 2018), summarizing and generating comprehension questions (Rassaei, 2017) and dictogloss, reconstruction cloze task, and jigsaw (Roohani et al. 2017). What is worth mentioning here is that comparing the impact of editing task and dictogloss as two types of output tasks received scant attention from researchers in the field. 

     In a meta-analysis, Douglas (2018) has pointed out that so far, the relationship between output-based tasks and grammatical knowledge has been scrutinized regarding questions, phrasal verbs, conditionals, and articles. It shows that the passive form has remained underexplored till now (Rahemi, 2018). Investigating the actual role that output tasks play in the process of second language acquisition in general and in grammatical development mainly is a beneficial and valuable area of research. It can have implications for material developers, instructors, and learners.  

     The primary aim of the current study is to help Iranian EFL learners develop their grammatical knowledge via output-based tasks. To achieve this goal, the influence of two types of output-based tasks, i.e., editing task and dictogloss, on the development of their grammatical knowledge will be explored. It is expected that the findings of the present study will shed light on the unexplored aspects of this issue and provide an impetus for future research. The actual impact of multiple kinds of output tasks on the development of various types of grammatical structures is an underexplored issue in the field of second language acquisition, especially in the context of Iranian educational systems. 

     To deepen our understanding of the relationship between output-based tasks and grammatical knowledge regarding passive voice, this study was set out to provide the answers to the following research questions:

1. Is dictogloss, as an output task, effective in improving Iranian EFL learners' grammatical knowledge regarding passive voice?
2. Is text editing, as an output task, effective in improving Iranian EFL learners' grammatical knowledge regarding passive voice?
3. Is there any difference between dictogloss and text editing in promoting Iranian EFL learners' grammatical knowledge regarding passive voice?

2. Literature Review

2.1. Theoretical Overview 

Swain (1985) talked about the role of pushed output in developing the automaticity of L2 learners. She proposed three critical functions for comprehensible output. The first one is the noticing/triggering function, whereby learners' attention to the form and structure of output is triggered. Hypothesis testing is the second function that Swain suggested for output. By producing output, learners can test the hypotheses that they have formed in their interlanguage system.  It is also mentioned that output can have a metalinguistic function in the process of language acquisition. Hence, learners can gain information about the language that they have produced. According to Lee and VanPatten (2003), output-based activities should be delivered to second language learners in a structured manner. This structured output should have some characteristics such as presenting things one by one, using oral and written output, answering to the content of output collaboratively and using activities that are meaning-focused.   

     Research on output is not irrelevant to form-focused instruction (Collin, 2012; Adloo & Rohani, 2019), and focus-on-form (Long, 1998). These approaches emphasize drawing students’ attention to particular language forms within the context of communicative and meaning-based activities. In this way, instructors can encourage students to heed to both meaning and the form. In this framework, language form can refer to grammatical structures, spelling, punctuation and so on. Focus on form is a necessary element for inverting input to intake because otherwise, it may be impossible for learners to notice new structures presented in tasks. 

     It should also be noted that traditional instruction, which is categorized as form-only activities ordinarily called “drills,” “mechanical practice” or “pattern practice” are different from output tasks in that the latter contains the inclusion of focus-on form in activities that are primarily meaning-based.

     One of the most relevant concepts to output and production of L2 is the noticing. Noticing Hypothesis, as proposed by Schmidt (1994, 1995), maintained that learning calls for awareness at the level of noticing. Schmidt proposed that consciousness of input in the form of noticing causes target language items to become more available for acquisition. Nonetheless, he believed that noticing does not assure acquisition. It is only the necessary and sufficient condition for converting input to intake for learning. Robinson's (1995) definition of noticing is “detection plus rehearsal in short-term memory, prior to encoding in long-term memory” (p. 296).

2.2. An Overview of Empirical Studies 

The bulk of the studies carried out concerning output measured the effectiveness of input-based instruction as opposed to output-based instruction. In one such study, Farley and Aslan (2012) investigated the impacts of output-based instruction and input-based instruction. The structure at foci in this study was English present subjunctive form. Participants were Turkish EFL learners. The results of the study showed the enhanced performance of both instructional groups.  Kavianpanah, Alavi and Ravandpour (2020) conducted not a dissimilar study. Of interest was exploring the influence of input-based and output-based tasks. The researchers, especially, cared about the similarity of involvement load of the tasks. They measured the vocabulary learning of Iranian EFL learners.  The results advocated the significant positive impact of input- and output-based tasks with equal involvement loads regarding vocabulary learning. 

     In another study on the comparison between input and output, Namaziandost, Saberi Dehkordi, & Shafiee (2019) were concerned with Iranian EFL learners. They intended to investigate the relative impacts of output-based and input-based tasks. The results indicated that both experimental groups outdid the control group. Nevertheless, another finding of the study was that no significant difference between the performances of the input-based and output-based groups in two post-tests existed.  In a more recent study, Alabdulah (2021) scrutinized the short-term effects of output and input. He targeted English causative forms. He used interpretation and production tasks to measure the efficacy of two instructional treatments. The study showed that input tasks are a beneficial pedagogical intervention and help L2 learners interpret, process, and produce English causative structures accurately. 

     Further research on the same issue was pursued by Mahmoudabadi et al. (2015). They investigated the role of output-first arrangement of tasks in improving vocabulary understanding. The results indicated that the output-input group outdid the input-output group. They also recommended order for the presentation of vocabulary tasks. In this order, the output is presented to learners before input to enhance noticing gaps and progress vocabulary knowledge. In the same line, Amiryousefi (2021) compared the effectiveness of input-based versus output-based instruction. The foci of this study was vocabulary development. The findings uncovered that both input-based and output-based instruction are helpful in improving participants' vocabulary knowledge, both receptively and productively. 

     Another trend of research compared the efficacy of multiple types of output tasks on the development of language skills of EFL learners. For instance, Khatib and Alizadeh (2012) carried out a study to inspect the impact of using two different types of output tasks. Their focal attention was the English past tense. The results revealed that only one experimental group, namely, the reconstruction group improved in target feature as far as noticing is concerned. On the other hand, both experimental groups equally promoted their learning regarding past tense. In the same line, Nassaji and Tian (2010) compared the effectiveness of two output tasks, i.e., the reconstruction cloze task and reconstruction editing task. They were interested in English phrasal verbs. The results showed more significant advantage for collaborative completion of tasks compared to individual completion of tasks as far as accuracy is concerned. Conversely, participants did not depict greater advances in vocabulary knowledge in collaborative tasks compared to individual tasks. 

     This impact of output-based task is also well evidenced in the study done by Sun (2016). He scrutinized vocabulary knowledge and acquisition. The researcher compared three instructional modes: picture-book reading-only, picture-book reading plus vocabulary instruction, and picture-book reading plus reading-based collaborative output activity, which he named PRO, PRVI, and PRCOA, respectively. The results showed that the most successful group was PRVI group in instant word learning. As far as word retention was concerned, the PRCOA group was the most successful group. This group was also the most successful one regarding productive knowledge. In another study, Etemadfar, Namaziandost, and Barani (2019) examined the effects of different output-based task reiteration conditions. The scope of their research contained pragmatics and the production of speech acts. The groups under investigation were programmed as: (1) the explicit task-repetition group, (2) the implicit task-repetition group, and (3) the no-input task repetition group, which were named ETR, ITR, and NTR. All the groups under study were engaged in reiterating output generation tasks. The outcomes demonstrated huge gains of participants in the ETR and ITR groups from the pretest to the posttest, but not in the NTR group. 

     Likewise, Teng (2020) scrutinized the effectiveness of the individual, pair and group output tasks. The target structure was phrasal verbs. The researcher concluded that group work led to a more considerable advance in the knowledge of the phrasal verbs in comparison with pair output task. On the other hand, pair forms of tasks led to superior performances than the individual type of task. Using a different kind of output task, Park (2020) made use of the cloze task. He investigated the effects of collaborative output tasks. Like Teng (2020), he put learning of L2 phrasal verbs under investigation. Participants experienced either a cloze (C) or editing (E) conditions. The findings revealed that the P group outdid the I group. Notwithstanding, significant differences were depicted only in the delayed posttests. An additional result of this study was that the learners showed superior performance in cloze tasks than in editing condition. 

     As it becomes apparent from the aforementioned studies, the comparison of various kinds of output tasks regarding passive form has received scant attention from researchers in the field of second language acquisition. Taking a contextualized perspective, it has been revealed that the actual role of output tasks in language learning in Iran is not a deep-rooted matter. To our dismay, a vast array of output tasks has not yet been explored across various grammatical structures, various proficiency levels, and different age groups. Hence, the current study tries to fill this recognized lacuna by testing the impact of utilizing two different kinds of output tasks on Iranian EFL learners.

3. Methodology

This study has a quasi-experimental design due to the nonprobability sampling used for choosing the study participants.  

3.1. Participants 

The participants of this study were 24 female Iranian EFL learners studying in the last year of high school (grade 12) in one of the high schools of Iranshahr, Sistan and Balouchestan province, Iran. Their first language was Persian. They were 17-18 years old. According to their English score for the previous semester, they were intermediate learners. The researcher also interviewed their English teacher. She also confirmed that they are at intermediate level of proficiency based on her qualitative analysis and their scores from the outset of the academic year. The years of their formal education were identical, i.e., they had received five years of formal instruction at the beginning of the research. They were not learning another language except English. They claimed that they were not exposed to English outside the school. From the information mentioned above, it became apparent that the participants of the current study shared identical demographic features. 

     The participants were already familiar with passive voice due to the English curriculum designed by the Ministry of education and taught in schools of Iran. Nonetheless, the treatment administered via this study aimed at fortifying their knowledge regarding passive voice and contributing to their accurate application of the target structure. 

      Before the treatment, a pretest was given to the participant to ensure their homogeneity. The pretest results showed no statistically significant differences between the participants. In other words, as far as their knowledge of passive voice was concerned, there was no disparity between them. They were divided into three groups; namely the control group, the dictogloss group, and the editing text group. Their demographic information is delineated below. 

Table 1

Demographics of the participants   

3.2. Sampling

For the sake of sampling, purposive (judgment) sampling was employed. Ary et al. (2017) pointed out that this is a nonrandom procedure for selecting the sample members, which has economy and convenience as its main advantages (p.177). The participants were judged to be representative of the target population (female high school learners), to which the results are going to be generalized. 

3.3. Target structure

The grammatical structure determnied for examination in the present study was passive voice delimited to the simple present and past tenses. This structure was chosen due to some reasons. At first, the passive voice was an underexplored structure regarding output tasks. For instance, Rahemi (2018) is one of the few studies exploring the effect of output tasks in the context of Iran. Another reason for choosing this structure was the participants's familiarity with it. They were taught passive voice in grade 12. Hence, it was supposed that the selected output tasks could assist them in the mastery of passive voice. The last reason for choosing passive voice is its complexity and difficulty for nonnative learners of English (Moreb, 2016; Elmadwi, 2015).

3.4. Data Collection Tools

Dictogloss, jigsaw, information gap, text editing, and text reconstruction tasks are instances of output tasks. Two types of output-based class activities were utilized in the present study. Dictogloss and text editing were chosen and used as output task in this study on the ground that few studies compared their influence on the learning of a second language in general and on the learning of passive voice in particular. Dictogloss is put under collaborative output task classification (Dehqan & Mohammadi Amiri, 2017). Collaborative output tasks have a number of advantages. They encourage participants to produce output in collaboration with others. They assist learners to reflect on and negotiate the correctness of their language use, whereby both meaning and forms is emphasized (Swain, 2005).

Wajnryb (1990) has stated that the dictogloss procedure consists of four stages: 

1. Preparation: During this stage, learners are informed about the topic of the text. They are given the necessary vocabulary to handle the task through a number of warm-up discussions. Another activity during this stage is that the learners are organized into groups. 
2. Dictation: During this stage, for the first time learners hear the text at a natural speed. they are not allowed to take any notes. The second time is when learners can jot down keywords to assist them in reconstructing the text and remembering the content of it. 
3. Reconstruction: This stage takes place in small group when learners work together to reconstruct the text with as correct grammar and content as possible. 
4. Analysis and Correction: During this stage, learners are expected to analyze, compare and correct their texts. This stage is done with the help of the teacher and the other groups.

Wajnryb's (1990) stages were followed during the administration of the dictogloss in the current study. Firstly, the participants in dictogloss group were given the title of the text. Secondly, their English teacher dictated text to them twice. In the third step, they reconstruct the text in pairs. Lastly, they corrected their text using the cooperation of classmates and their teacher.

     Regarding editing task, it is a type of output task, whereby learners correct a text to enhance its accuracy. Depending on the way participants edit the text, the editing task can be regarded as a collaborative or individual task.  In the current study, the editing task was operationalized as an individual task. The participants in the editing text group were asked to edit and correct tasks individually. They were given a text containing eight errors pertinent to incorrect usage of passive voice. They were asked to identify and correct the erroneous parts of the text. The texts used for the editing text group and the dictogloss group were precisely identical. The focal structure in these texts was the passive voice. The texts contained 100-115 words. 

3.5. The Data Analysis

 Data were analyzed quantitatively. For the sake of analyzing the data of this study, mean scores were calculated for the learners’ performance in dictogloss and editing tasks and on the pretests and posttests as well. The means were then compared using post hoc Tukey and t-test. 

3.6. Research Design

The study took four sessions and involved a pretest, four treatment sessions, and a posttest. When the participants were selected, information sheets were distributed among them. This sheet provided general information about the purpose of the study, and the rights of the participants. Nevertheless, the participants were neither told of the specific focus of the study, nor were they informed about the procedures conducted with the other groups under investigation not to jeopardize the internal validity of the research inadvertently (Clark & Creswell, 2015). The participants’ email addresses and phone numbers were taken for possible additional follow-up clarification. To gain insights into the participants’ demographic information and to ensure that the participants shared similar features, they were asked to complete a survey. Along with questions regarding their major, age, gender, and experience living in other countries, the participants were questioned about their estimation of the hours on the average day they spent listening to English radio programs, watching English TV programs, or conversing in English. 

     Participants were divided into three groups: one control group and two experimental groups. The first experimental group received treatment in the form of dictogloss. The second experimental group was engaged in editing a text. The text contained some errors regarding passive voice which required the students to edit and correct them. The third group, present in the study as the control group, read a text in active voice. The text for the two experimental groups was the same each session. The three intact groups were taught by the same instructor. They attended English classes two days weekly. The researcher sat in the classroom as an observer during the administration of tasks to ensure that the progression of tasks was as pre-planned. To reduce the experimenter effect and threats to the internal validity of the study (Ary, et al. 2017, p. 279), the researcher had attended participants' classes for two sessions before the commencement of the study. Upon the completion of the tasks, a posttest was administered to the three groups under investigation. Pretest and posttest were at the same level of difficulty. Pretest and posttest contained measured participants' receptive and productive knowledge concerning passive voice.  

4. Results

To examine the effectiveness of the interventional program, the mean scores of the three groups, one control group, and two experimental groups, were compared. Figure 1 illustrates the mean scores of the three groups. According to Figure 1, all groups have started from similar points in the pre-test. After the intervention program and in the posttest, only experimental groups showed progress and showed changes in their performances, while the control group’s performance did not change too much. 

Figure 1

Groups’ Mean Scores in Pre- and Posttest

To investigate whether the observed changes are statistically significant, the Post-Hoc Tukey test was used to allow for multiple comparisons. Table 2 depicts the descriptive statistics of the performance of the three groups on both pretest and posttest. What is more, Table 2 indicates that, in the pretest, the best performance belonged to editing group (M = 2.38), the control group (M=2) came second, and the weakest performance belonged to the dictogloss group (M = 1.50).

Table 2

Descriptive Statistics of the three groups of the study

Post-Hoc Tukey test was run to ensure that the assumption of homogeneity of variances was met in the pretest. The results of post-hoc Tukey (see Table 3) revealed no statistically significant difference at the p < .05 level in the pretest scores for the three groups. Consequently, this guaranteed homogeneity of the three groups regarding their grammatical knowledge of passive voice. Whereby they are comparable with each other. 

Table 3

 Post- Hoc Tukey for all three groups on pretest

What is more, the comparison of the results of the pretest and posttest for the control group did not depict any significant change (p > 0.05). Ergo, there was no improvement in grammatical knowledge of the control group as far as passive voice is concerned.  

Table 4

Within the Group Comparison for the control group in pretest and posttest

Now, the research questions are examined one by one. 

4.1. Answering the First Research Question

The first research question of the present study was:

Is dictogloss, as an output task, effective in improving Iranian EFL learners' grammatical knowledge regarding passive voice? 

Given that, the first null hypothesis of the current study is:

Dictogloss, as an output task, is not effective in improving Iranian EFL learners' grammatical knowledge regarding passive voice.

To answer the first research question, the participants' performance in the dictogloss group was examined in terms of differences across pretest and posttest. The intention was to see if there were any statistically significant improvements in the participants' performance from the pretest to the posttest. An independent sample t-test was applied. As depiced in Table 5, the performance of the dictogloss group improved from the pretest to the post-test (0.00 ≤ p ≤ .05).

Table 5

Within the Group Comparison for the dictogloss group

As it became apparent, the first null hypothesis of the study is rejected since dictogloss does have a positive impact on the improvement of Iranian EFL learners' grammatical knowledge regarding passive voice.

     A further post-hoc Tukey was run to test the differences between the dictogloss group and the control group. The results are tabulated in Table 6 below. 

Table 6

Comparison between the control and dictogloss group on the post-test

The results indicated that there was a statistically significant difference between the control group and the dictogloss group on the posttest (0.00 ≤ p ≤ .05). Thus, the findings of the previous table, i.e., Table 6, are fortified. 

4.2. Answering the Second Research Question

The second research question of the current study is:

Is text editing, as an output task, effective in improving Iranian EFL learners' grammatical knowledge regarding passive voice?

Consequently, the second null hypothesis of the current study is:

Text editing as an output task is ineffective in improving Iranian EFL learners' grammatical knowledge regarding passive voice.

To answer the second research question of the present study, the participants' performance in the text editing group was examined in terms of differences across pretest and posttest. Similar to the previous calculation for the dictogloss group, the purpose was to see if there were any statistically significant improvements in the text editing group participants' performance from the pretest to the posttest. An independent sample t-test was run. As depicted in Table 7, the performance of the text editing group improved from the pretest to the posttest (0.00 ≤ p ≤ .05).

Table 7

Within the Group Comparison for text editing group

To boost the findings obtained from the independent t-test, a posthoc Tukey was run between the text editing group and the control group. The results, tabulated in Table 8, indicated that there was a statistically significant difference between the editing group and the control group (0.00 ≤ p ≤ .05) on the post-test. 

Table 8

Between the Group Comparison for text editing and control groups on posttest

Ergo, the second null hypothesis of the study is rejected due to the finding that text editing does have a positive impact on participants' grammatical knowledge regarding passive voice. 

4.3. Answering the Third Research Question

The third and last research question of the present study is: 

Is there any difference between the dictogloss and the text editing in promoting Iranian EFL learners' grammatical knowledge regarding passive voice?

Accordingly, the third null hypothesis of the study is:

There is no difference between the dictogloss and the text editing in promoting Iranian EFL learners' grammatical knowledge regarding passive voice.

To answer this question, a posthoc Tukey was run between the performance of the two experimental groups, i.e., the editing text group and the dictogloss group, on post-test. The results are tabulated in Table 9.

Table 9

Post Hoc Tukey for comparison between editing and dictogloss groups on pretest and post-test

The results showed no significant difference between the two interventional programs (p > 0.05). In other words, both programs are effective in improving the English language learners’ grammatical knowledge of passive voice, but no program is more effective than the other one. 

     Thus, it can be concluded that the third null hypothesis of the current study is accepted due to the finding that there is no statistically significant difference between text editing task and dictogloss as the two types of output tasks as far as improvement in knowledge about passive voice is concerned. 

5. Discussion

The current study sheds light on the effectiveness of two types of output tasks, namely dictogloss and editing text, on improving EFL Iranian learners' grammatical knowledge. The focal attention was on passive voice. The results indicated that the two experimental groups significantly outperformed the control group on the post-test. In addition, it was shown that there was no statically significant difference between the two experimental groups on the post-test. Hence, the study demonstrated the positive impact of dictogloss and editing text in the development of EFL leaners' grammatical knowledge regarding passive voice.

     The first point to be discussed is the potential utility of output tasks in improving grammatical knowledge. These results are consistent with the findings of Farley and Aslan (2012) for present subjunctive form, Khatib and Alizadeh (2012) for past tense, Rahemi (2018) for passive voice, Nassaji and Tian (2007), Park (2020) and Teng (2020) for phrasal verbs, Alabdulah (2021) for causative form. Output tasks push learners to apply their previously-acquired knowledge in their output. This opportunity provided by output tasks in using and practicing newly-acquired knowledge results in the efficacy of output tasks.

     What can be added to the studies mentioned above is the noticing-the-gap function of output tasks (Swain, 1985) and the pertinent notion of noticing the holes in interlanguage by learners of a foreign language (Doughty & Williams, 1998). The production of output via output-based tasks corroborates learners' awareness-raising about the deficiencies in their interlanguage in general and grammatical knowledge in particular.  It can be said that these kinds of output-based tasks stimulated participants to focus on certain grammatical aspects which were not readily available to them via the present state of their interlanguage. Nassaji and Fotos (2007) declared, “if the goal of L2 classroom activities is to develop both accuracy and fluency, it is clear that meaningful activities must be integrated with form-focused activities, particularly those requiring output” (p. 15).

     The results of the current study can be explained in the light of the Noticing Hypothesis. In this hypothesis, Schmidt (1990, 1993, 1994, 1995) argues that learning requires awareness at the level of noticing. He proposes that consciousness of linguistic items in the form of noticing makes target language items more available for acquisition. It can be argued that output tasks push the participants to consciously notice passive voice and utilize it in their further production.

     The efficacy of these kinds of output-based tasks can also be explained in the light of Dekeyser's (2007) notion of 'skill specificity' concerning the effect of output practice. DeKeyser (2007) pointed out that particular kinds of practice have the potential to result in the development of the specific skills. Likewise, this study has shown that editing and dictogloss type of output tasks lead to the development of dealing with specific grammatical structures, namely passive voice. In line with Dekeyser's (2007) argument, it should be considered that the ability gained from practice (output-based tasks) may be skill-specific. Hence, there is a need for knowledge of every grammatical structure to be corroborated by the help of output-based tasks. 

     Additionally, in line with Benati's (2017) claim who suggested that “grammar instruction should move from input to output practice” (p. 391), this study indicates that output-based tasks that focus on grammatical structures should be designed in a way that paves the way for learners' noticing and processing of those grammatical structures. After that, the instructor can use output-based tasks to encourage language production and consequently, the development of grammatical structures. 

     The next point is the prior knowledge of the students with a passive voice that create an appropriate situation leading to the efficacy of output-based tasks. Ahmadi, Ghaemi, and Birjandi (2016) prudently mentioned that for effective learning to happen, it is a requirement for learners to use the newly received language in their production. The participants' prior familiarity with passive voice in English high school books fortified the influence of output tasks in this study.

     Regarding the efficacy of dictogloss, it can be said that this type of task assists the learners in paying synchronic attention to both meaning and structure. In this way, lexical and grammatical knowledge are boosted synchronously. This strengthening of form-meaning connection is well-evidenced in the outperformance of the dictogloss group on the post-test in comparison with the pretest. Another feature of the dictogloss is its requirement for EFL participants to pay attention to form while allowing them to solve linguistic problems.

     Another factor that mediates the effectiveness of output tasks, especially the text-editing task, can be found in the Direct Contrast Hypothesis (Saxton, 1997). This type of output-based tasks explicitly indicates to learners that there are some errors in the text in need of correction. Such cognizance of the existence of errors in the text can form the basis for trying to find errors and correcting them. When learners recognize the contrast between incorrect and correct grammatical forms, the first step in modifying the interlanguage to L2 norms occurs.

     The viability of text editing as an output-based task can be explained in the light of Vygotskyʼs (1978) sociocultural theory. Based on this theory, knowledge is socially constructed and internalized via interaction. Interaction provides the opportunity for L2 learners to develop their linguistic and problem-solving repertoire. As they perform the editing text tasks, they build knowledge through metalanguage on the errors that contain in the text. What facilitates this regulation process is the provision of support by other learners. That is to say, the participants of the current study examined their grammatical structures in the course of interaction, regulated or restructured their grammatical knowledge, made an adjustment, and received feedback. This, in turn, led to better grammatical gains. 

     The intriguing issue is that all of the participants who participated in this study were female. Ergo, another reason can be explored in gender factor. Females are in general more sensitive. Therefore, they make every effort to discard errors from their interlanguage to speak and write correctly. It can be claimed that they make better use of output tasks to be accurate in the second language. This finding is commensurate with Thijittang and Lê’s (2008) research which demonstrates that women tend to be more sensitive and attentive.  

     The last point to be mentioned here is that although this study was about output tasks, integrating both input and output in the actual context of the classroom is highly recommended. Both input and output are vital for learning of a second language. VanPatten (2002) accentuated the necessity of incorporating output activities into grammar instruction. In this way, the learners will notice structures that fill gaps in the system of their interlanguage. This, further, prompts fluency and accuracy in accessing their interlanguage system.  

6. Conclusion

In this study, the effectiveness of output-based instruction was scrutinized. It was found that dictogloss and text editing as two types of output-based tasks effectively improve EFL learners' grammatical knowledge as far as passive voice is concerned. Further, it was found that there is no difference between dictogloss and editing task in developing knowledge of the passive voice. That is to say, both dictogloss and editing task are equally effective in developing grammatical knowledge.

     We should keep in mind that one should be cautious in drawing implications from a single research due to the limitations inherent in it. Nonetheless, the findings of the current study uphold the use of output-based and output-oriented teaching of a foreign language as a superb vehicle to adopt a learner-based approach. Output tasks can be applied as an alternative in foreign language teaching to enhance opportunities and chances for practice.  They can be looked upon, especially in contexts and educational settings where an input-based approach toward teaching fails. The Ministry of education can hold roundtables and workshops for language teachers at national and local levels to inform L2 teachers about the findings of this study and similar studies on output tasks. 

     The findings have implications for L2 learners and instructors, as well. One of the indirect advantages of output tasks is their potential to raise the level of self-confidence of EFL learners and lessen their stress while dealing with tasks because learners are pushed to learn via their production and practice. At the time of their self-study, EFL learners can benefit from output tasks. They can also use dictogloss and editing when they are practicing newly taught structures outside of the classroom. In addition, teachers can best apply the findings of this study and similar studies to enrich their instruction of grammatical structures of a foreign language. 

     It is also highly recommended that material developers at any level, mainly those providing textbooks to be taught in high schools, consider the findings of this study and provide a wealth of output practices in the materials they are going to develop. To put it differently, careful inclusion of activities such as editing texts in textbooks can assist the materials developers in attaining better results.

     Limitations are unavoidable and inherent to any study. This study is by no means an exception. The inclusion of just two types of output-based tasks, paucity of the participants, consideration of one level of proficiency, the inclusion of just native speakers of Persian, and measurement of just short-term effect of treatment are some of the shortcomings of the current study.

     Some possible and promising avenues for further research are extending similar studies across multiple contexts, concentrating on other types of output tasks such as cloze task, scrutinizing other grammatical structures such as conditionals, the inclusion of participants with various levels of L2 proficiency, presence of participants of different age and gender and,  exploring the relationship between the efficacy of output-based tasks, and such factors as personality, individual differences, developmental readiness, motivation and anxiety. 

          At last, the researcher hopes that what has been reported herein be successful in covering one of the gaps in the reign of instruction of grammatical structures and paving the way for further research to come. It is also hoped that other researchers learn from and build upon what has been proposed in the current study, thereby revolutionizing the field of SLA based on the innovations suggested here and those which will grow out of prospective studies.

Acknowledgement 

The authors express their gratitude toward all people who assisted them during the data collection phase of the study. 

Declaration of Conflicting Interests

The authors of the present article do not have any conflicts of interest to declare.