The Effect of Typographical Input Enhancement on Iranian EFL Learners ’ Accuracy in Oral Production of Narratives

Researchers suggest that language learners taught with the Communicative Language Teaching (CLT) approach tend to have trouble acquiring semantically less salient grammatical forms. This study describes an experiment in input enhancement in general, and specifically Typographical input enhancement which is one of the three main types of input enhancement and it attempts to find out which two sets of materials are better in helping students to use the past tense form of English more accurately: unenhanced texts, or enhanced materials in the form of short stories. The framework of the study was the incorporation of focused comprehension tasks. 40 female students at the age range of 18-28 participated in the study and test results indicated that the participants in the Experimental group (which received enhanced input) increased significantly in their ability to use the simple past tense verbs accurately in oral narratives on the posttest. However, inter-groups differences were not statistically significant. It was concluded that typographically enhanced input can be a valuable addition to the CLT curriculum in order to improve learners’ accuracy in their speaking skill through increasing noticing by the means of enhancing input typographically.


Introduction
Communicative Language Teaching (CLT) has been considered as the most successful approach of language teaching in recent years.The main underlying principle in CLT is enabling students to use the target language communicatively with the primary focus on meaning.Sometimes, in some conditions, this principle has made a serious problem for language teachers and students.When the primary focus is on meaning or 'message content', language form can consequently be ignored or at least it can be considered unimportant.This is a critical problem for language learners since it has been observed that even learners in advanced levels have grammatical problems in their performances or let us say they lack enough accuracy.Fluency and accuracy are both important and they are the two final goals of language classrooms; that is, the ultimate goal of language learning is to speak the language fluently and accurately.Although Larsen-Freeman (2000) emphasized fluency over accuracy, she did not state that language form (grammar which leads to accuracy) is not important.

Focus on form, noticing and typographical input enhancement
According to Gass (1998), learners need to focus on form in order to develop a more complete grammatical repertoire in the second language.He mentioned the immersion programs in Canada and their obvious success in teaching a second language such as French (quite apart from the discipline curriculum).Evaluations of these programs showed that while students seemed to show a great amount of fluency in the use of French, the range of grammatical structures that were utilized in their communication was limited (Harley & Swain, 1984;Swain, 1985).
It is a well-known fact that both the extreme interventionist focus on forms and non-interventionist focus on meaning have problems.There is a viable third option, however; which attempts to capture the strength of an analytic approach while dealing with its limitations that is called focus on form (not forms).It refers to how attention resources are allocated and involves briefly drawing students' attention to linguistic elements (words, collocations, grammatical structures, pragmatic patterns, and so on) in context whose overriding focus is on meaning (Long 1991, as cited in Gass & Torres, 2005).
The purpose is to induce what Schmidt (1990) calls noticing, i.e., registering forms in the input so as to store them in memory.Typographical input enhancement, a very useful and important technique in focus on form instruction, has been used in this study to draw learners' attention to the intended target forms.The context of the study is Task-Based Language Teaching (TBLT) since input enhancement has been used as a way in incorporating focused tasks in the classrooms with the aim of inducing students noticing of the targeted forms.Skehan (1992, as cited in Willis & Willis, 2001) argued that the learning becomes more efficient if: 1) In a task-based methodology, there is an essential need to focus on accuracy.
2) In the task-based cycle, there is a critical focus on language form.As Willis (1996) mentioned, in a task-based teaching, the post-task stage is needed to counter the danger that students will develop fluency at the expense of accuracy.One way to help learners notice important features in the input is to make those features more salient by frequency and/ or enhancement.By manipulating input, one can hope to bring about changes in the learners' sensitivity to input as well: something that the learner was previously not sensitive to and didn't notice, might become noticeable through input enhancement.That is, instead of the learner's input sensitivity being heightened, the input made more salient, to meet the learner's (lower) level of input sensitivity.Typographical input enhancement is one of the three main types of input enhancement classified by O'Bryan (2004).These three types are: 1) making linguistic characteristics salient (typographical enhancement), 2) providing elaborated input, 3) providing modified input.Chapelle (2003) goes further and states exactly how input can be made salient by use of repetition of marked input (e.g.highlighting, bolding, italicizing specific structures), that is called typographical input enhancement.
Research conducted within the last decade has revealed that students are having trouble learning grammar within the methodological framework that focuses overwhelmingly on the communication of meaning (Ellis, 1997;Celce-Murcia, 1991a,b ;Dornyei & Thurrell, 1991, 1992;Kumaravadivelu, 1993;Larsen-Freeman, 1990 ).More specifically, the aspects of inflectional morphology such as subject-verb agreement and past tense markers are notable problematic areas for learners of all proficiency levels (Larsen, 2002).Both theorists and practitioners within the area of CLT and language instruction in general will benefit from an investigation of this problem.Salaberry (1998) conducted a study on the acquisition of English past tense in an instructional setting and investigated the role of developmental sequences in the acquisition of inflectional morphology that is relevant for the analysis of adult second language acquisition, and ultimately, second language pedagogy.Salaberry (1998 p.56) states that "inflectional endings are among the most difficult features of non-native languages for adult learners' highest amount of variability and lowest degree of success".Simple past verb forms have been chosen by the researcher as the targeted forms in this study since in narrative tasks in which learners are asked to read short stories and reproduce them in the class the problem of incorrect use of the simple past form becomes more evident.

Statement of the Problem and Purpose of the Study
This study intended to investigate the role of typographical input enhancement (an implicit way of teaching grammar) on the accuracy of using simple past tense in reproduction of narrative tasks in EFL adult learners.Input enhancement in general refers to a group of techniques for focusing learner's attention on those aspects of the language that may otherwise go unnoticed and unlearned.Chapelle (2003) states exactly how input can be made salient by the use of repetition of marked input.Typographical input enhancement entails modifying the target structure(s) by highlighting, underlining, bolding, italicizing, using larger, different type or different color font, or through combination of these methods.In this study, the enhanced input provided for the treatment is bolded, italicized and enlarged font.Generally as Ellis (2003) mentioned, input enhancement involves designing tasks in such a way that the target feature is (1) frequent and/or (2) salient in the input provided.Task-Based Language Learning (TBLL) has now been considered the most effective way of learning a foreign language.Thus, tasks have a central place in language classes of Task-Based Language Teaching (TBLT).Focused tasks as one of the two general types of tasks (the other one is unfocused tasks), termed by Ellis (2003), aim to draw learners' attention to the linguistic forms of the language.Input enhancement as Ellis (2003) mentioned, is one of the two important ways in the instruction of focused tasks.Input enhancement can be seen from the focus-on-form perspective in which the primary focus is on meaning but at the same time learner's attention is drawn to certain linguistic features.In the broader meaning of input enhancement, instruction can be considered as a form of input enhancement in that the input is manipulated to force the learner to attend certain structures.Typographical input enhancement which is the written mode of input enhancement in which the intended target form is highlighted, bolded, italicized or the combination of these techniques is used (Urano, 2000) can be used as a way to help learners improve their accuracy in their language production.Simple past tense has been selected by the researcher as the intended target structure since it is used in narratives widely.The surprising point is that although learners have already studied this tense as the grammar focus in their course books, most of them are not able to use it correctly in their oral production.That is why the researcher has chosen this linguistic feature.More specifically, the two types of regular and irregular verb forms are problematic for the learners.As such, the researcher's assumption is that typographical input enhancement has a positive role in the accuracy of reproduction of narratives.

Research Question and Hypothesis
Regarding the purpose of the study the following research question is presented: Does typographical input enhancement have any role in the accuracy of reproduction of narratives?
As such, the researcher's assumption is that typographical input enhancement has a positive role in the accuracy of reproduction of narratives.To follow the requirements of the scientific studies and increase the validity of the experiment, a null hypothesis was suggested which claims that typographical input enhancement does not have any role in the accuracy of reproduction of narratives.If the results of the study can reject the null hypothesis, then the research hypothesis will be confirmed.Otherwise, it will be regarded as invalid.

Method
The study was designed to investigate the effect of typographical enhancement on the accuracy of simple past tense use in oral reproduction of narratives.It was expected that the learners' attention would be directed to the typographically enhanced forms; consequently, noticing can be increased and students' received input can be converted into intake (Hulstijn, 1997;Tomlin and Villa, 1994).Thus, the hypothesis of the study was the following.Typographical input enhancement has a positive effect on the accurate use of simple past verb forms in oral reproduction of narratives.In other words, learners exposed to typographical input enhancement of simple past verb forms would progress further in the accurate use of those forms in oral reproduction of narratives than would learners who did not get enhanced input.

Participants
The participants in the present study were 40 female students from institute in Tabriz, Iran.Their ages ranged from 18 to 28 and they were all at pre-intermediate level.Their course book was "New Interchange 2" by Jack C. Richards (2005) from unit 1 to 6 corresponding to level 6 in Jahade-Daneshgahi institute.Since there were only 2 classes available for the experiment, the researcher had no choice except using intact group design.In order to have control and experimental groups, a pretest including three short stories were given to the learners at three different stages on one day.

Instrumentation
To measure the subjects' accuracy in using simple past forms in narratives, an oral test as a pretest including three short stories from the Stories for Reading Comprehension book by J. Hill (2005) were given to the participants.For the treatment, five short stories from the same book were used.The course book which was taught for both the Experimental and Control groups was "New Interchange 2".Each unit consisted of a conversation, a reading, listening activities, a writing and two grammar focuses with related exercises.

Procedure
At the beginning of the 5-week program the pretest including three short stories was administered at three stages.The stories were of the same length and their themes were similar too.The students were asked to read the stories and narrate them orally three times on one day.All oral narratives were tape-recorded and then transcribed in orthographic format.Inter-rater reliability for coding was obtained by having another instructor as the rater.In order to score the oral pretest data the raters listened to each audio-taped recording and then transcribed them.Accuracy was measured by calculating the number of the correct simple past verb forms produced by each student as a percentage of the total number of the verb forms.The framework for measuring accuracy was generally limited to the correct form of the verbs in past tense.Previous research has found such a global measure of accuracy is more effective and sensitive than the others (Buscemi, 2003;Morris & Tremblay, 2002).The learners were asked to read the short story at each stage and narrate it orally .The learners' oral data were recorded and rated by two raters to obtain inter-rater reliability (one of them was the researcher herself).According to the scores of the pretest, learners were divided into two groups: Control and Experimental groups.The students in the Control group received unenhanced texts (the targeted feature were not typographically enhanced) in the form of short stories.They were given a short story passage and were requested to read and answer the comprehension questions after the story during five sessions, each of them lasted for 25 to 30 minutes.After that the raters asked them to narrate the story orally.Narratives (tasks) were used for two goals.One was to guide the learners to use the target language in storytelling which is one of the very important tasks in everyday life (Muranio, 2000).The other is to present learners contexts for a particular linguistic form and guide them to accurately produce a particular target form which is simple past form in this study.The Experimental Group received the same short stories but the past tense forms were typographically enhanced.The researcher enhanced the targeted feature (simple past regular and irregular verb forms, past copula and negative statement with did) through the combination of italicizing.The task was the same as in Control Group.The participants were requested to read the short story and answer the reading comprehension questions.Then they had output practice in which they were asked to narrate the story orally.
A second rater, as stated above, was required to establish inter-rater reliability for those variables requiring subjective judgment.On the basis of the scores obtained from the pretest, the students were classified in two groups: Experimental and Control Groups.
The instructional treatment was provided during five sessions, each of which lasted approximately 30 minutes.In the Control group, the students were asked to read unenhanced short story passage at each session and answer the reading comprehension questions in the form of checking the correct response at the end of the story (10 min).Then they were requested to narrate the story orally in the class (20 min).Experimental Group received the same short stories but the texts were typographically enhanced.The form of enhancement was the combination of bolding, italicizing and larger font size.The texts had a large number of sentences in simple past.The regular and irregular verb forms, past copula, and negatives with did, all were bolded, italicized and enlarged.The task was the same as in Control Group.At the end of the 5-week program, the participants in both groups were posttested.The posttest procedures were exactly the same as presetting.Three short stories were chosen from the reading comprehension book for the posttest.Each student's performance was tested on the basis of the accurate use of simple past forms and the scores were out of 20.The procedures for scoring the posttest data were the same as the pretest.The percentage scores were used in the statistical analysis conducted to answer the research question with respect to accurate oral simple past verbs production.Results from the statistical analysis of the obtained data will be presented in the following sections.

Study Design
Due the proposed research question, this study required an experimental method of research.It contained a pretest, a posttest, a control and an experimental group.The Classes were in five sessions for five weeks.The pretest was given one day before the treatment was started.The posttest was administered one day after the treatment.The time between the pretest and the posttest was long enough (5 weeks) to reduce the test-re-test effect.Typographical input enhancement in this study was the independent variable which is the major variable that is hoped to be investigated.The participants' accurate use of simple past verb forms is the dependent variable which is observed and measured to determine the effect of the independent variable.By choosing only female students, gender, was controlled.The process through which the accuracy of simple past use in participants' narrations was acquired, cannot be observed or measured and it is called intervening variable.Therefore, intervening variables are usually neither observable nor measurable.

Correlations between the Raters Inter-rater Reliability
The Pearson product-moment correlation coefficient was used to establish the inter-rater reliability through computing the correlation between Rater 1 and Rater 2 in the mean score of the three stories in the pretest.The Pearson product-moment correlation coefficient was used to establish the inter-rater reliability through computing the correlation between Rater 1 and Rater 2 in the mean score of the three stories in the posttest.
As Table 5.1.shows the computed Pearson correlation coefficient r at the level α=0.05 pretest was 0.975 (P < 0.01) which indicates that there was a significantly positive and strong relationship between the percentage scores rated by Rater 1 and Rater 2.
The Pearson product-moment correlation coefficient was used to establish the inter-rater reliability through computing the correlation between Rater 1 and Rater 2 in the mean score of the three stories in the posttest.As Table 5.2.shows the computed Pearson correlation coefficient r at the level α=0.05 in the posttest was 0.998 (P < 0.01) which indicates that there was a high positive relationship between the percentage scores rated by Rater 1 and Rater 2.

Data Analysis
A t-test analysis was run to determine if there were any statistically significant differences between the two groups' mean scores on the pretest measuring accuracy of simple past verb form use in participants' oral narrations.The difference is statistically significant.The reason behind the positive effect of input enhancement is the learners' cognitive processing system and the factor "noticing". Sharwood Smith (1993) explained the rationale for input enhancement, that is, we can facilitate the learners' selection process of input by increasing the perceptual saliency of specific targeted forms in the input.This process would appear to engage the learners' attention as a selective process as it involves directing the learners' focal attention to a specific form an array of verbal or written forms (Combs, 2006).Some studies found that only exposure to targeted language is not sufficient for SLA and that learners' cognitive process "noticing" is important.Noticing takes place when learners pay attention to certain linguistic feature in input.Gass (1988) asserted that noticing is the first stage of language learning.Cross (2002) referred to the importance of noticing by describing it as the "gateway to subsequent learning" (p.100).Similarly, Sharwood Smith (1981), Rutherford (1988, as cited in Cross, 2002) and McLaughlin (1990, as cited in Cross, 2002) advocated that noticing a feature in the input is an essential first step in language processing.
On the basis of the criticism of CLT from influential scholars such as Celce-Murcia (1991) and Ellis (1997), as well as others who call for a more direct, focus-focused teaching of grammar within CLT, this study set out to investigate if enriched input (Ellis, 1999) and enhanced input (Sharwood Smith; 1993) would be a valid option and a viable addition to CLT grammar teaching.As most learners, even relatively advanced ones, have problems with morphological forms, the English simple past tense was chosen in this study as the targeted feature to be influenced by the enhanced texts in the input provided for the Experimental group.

Conclusion
According to the results obtained from the experiment, the researcher's hypothesis which was proved to be true justified the positive effect of input enhancement on the accuracy of EFL students.The conclusions that can be drawn would be only suggestive in the sense that only one grammatical form was chosen and just oral test was administered.However, generally speaking, focus on form instruction from which input enhancement was used as a technique to overcome some accuracy problems of EFL learners has been considered to be more effective and useful in contrast with focus on forms instruction since not only it does provide learners with grammatical knowledge but also with communicative skills.
As a practical activity in focus on form instruction, focused tasks are recommended (Ellis, 2003).Focused task aims at learners' acquisition of grammatical knowledge.However, during the focused task, learners focus on meaning principally through negotiation for meaning but simultaneously they focus on linguistic feature.
Doughty & Williams (1998) stated that putting sole emphasis on forms is unsuccessful, since this approach results in learners who know about the second language but cannot use it.Putting sole emphasis on function and communication, while a far better approach in terms of the high degree of fluency and accuracy is possible, has been considered useless too.
Formal instruction, output, and feedback are all effective for SLA because they promote noticing.The results of some studies like Nagata (1998), indicated that given the same grammatical instruction, output-focused practice which was used in the present study is more effective than input-focused practice for the production of certain is significant at the 0.01 level(2-tailed)

Figure 5
Figure 5.1.Mean Difference of the Pretest and Posttest

Table 5 .
3. The means, Std.Deviation, and Std.error mean in pretest (group statistics)

Table 5 .
3. shows the means, standard deviations and standard error means of the two groups for the oral narrative tasks on the pretest.The pretest mean score was 39 for the Experimental group, and 43 for the Control group and the standard deviations were respectively 16 and 14 for the groups.It is important to note that the scores were accuracy percentages which were calculated by dividing the accurate use of target forms by the total number of target forms.As it is shown in Table5.4., the difference between the mean of the Experimental and the Control Groups in the posttest is noticeable.

Table 5 .
5. shows the means, t value, degree of freedom and 2-tailed probability in the posttest of the two groups.

Table 5 .
6. is the comparison of the pretest and posttest results of the Experimental and Control Groups.Table5.6.shows a higher mean difference between the two groups and also a higher Std.deviation for Experimental Group.

Table 5 .
7. The comparison of mean improvement in the Groups (Independent Samples Test)According to Table5.7., the participants' degree of progress in the Experimental Group is higher than the Control Group (Mean difference=25), and it is statistically significant (α=.000).