Effects of Task Complexity, Task Conditions, and Task Difficulty on the Grammatical Accuracy of EFL Learners in Written Discourse

Effects of Task Complexity, Task Conditions, and Task Difficulty on the Grammatical Accuracy of EFL Learners in Written Discourse

SaeidehAhangari*,¹Mahlagha Akbari²

1, 2.Department of English, Tabriz Branch, Islamic Azad University,Tabriz, Iran

*Corresponding author: Ahangari@iaut.ac.ir

Received: 2011.5.25

Revisions received: 2013.8.17

Accepted: 2014.12.9

Abstract

Different methods of language teaching have tried to help EFL learners to develop good language skills based on their various perspectives. Research findings have underscored the effect of using task types in promoting language skills in terms of accuracy in written discourse. Therefore, this study set out to investigate whether there is an evidence of correct use of simple past tense (Accuracy) based on Task Complexity (Task type :Here-and now & There-and-then),Task Conditions (Gender: Male & Female), and Task Difficulty (Proficiency: Lower-intermediate & Intermediate). Sixty Iranian English learners in a language institute participated in the study and were assigned to four groups of lower-intermediate male, lower-intermediate female, intermediate male and intermediate female. Initial homogeneity of the groups was verified using two general proficiency tests; KET for lower-intermediate and PET for intermediate. All groups in here-and-now task type were asked to write a story using simple past based on a picture strip while for there-and-then task type the participants were supposed to write about their last birthday. The results from paired samples t-test, independent samples t-test and two-way ANOVA analysis of the written data revealed significant differences in performing task types, at different proficiency levels and interaction between them. The findings have significant pedagogical implications for EFL learners to understand the relationship among Task Complexity,Task Conditions, Task Difficulty and L2 written production leading to various degrees of Accuracy.

Keywords: Task Complexity (Task type: here-and-now & There-and-then), Task Conditions(Gender: Male & Female), Task Difficulty (Proficiency: Lower-intermediate & intermediate), Accuracy.

Introduction

     In recent years there has been a clear shift toward teaching language by doing it and cognitively involving learners in the activities they are performing. This led to the development of Task-based language teaching. Given the fact that language acquisition is influenced by the complex interactions of a number of variables including materials, activities, and evaluative feedback, TBLT has a dramatic, positive impact on these variables. It implies that TBLT provides learners with natural sources of meaningful material, ideal situations for communicative activity, and supportive feedback allowing for much greater opportunities for language use.

     Tasks are then a major locus for learners’ involvement with the language in order to promote learning. According to Long (1985) a task is:

a piece of work undertaken for oneself or for others, freely for some reward. Thus, examples of tasks include dressing a child, filling out a form, buying a pair of shoes, making an airline reservation, borrowing a library book, taking a driving test, typing a letter, weighing a patient, sorting letters, taking a hotel reservation, writing a cheque, finding a street destination, and helping someone across a road.” (p.43)

For Ellis (2000) task is “a work plan; that is, it takes the form of materials for researching or teaching language” (p.195). As a work plan it typically involves some input learners are required to process and use, and some instructions relating to the expected outcomes. The work plan stipulates the non-linguistic outcome of the task, which serves as the goal of the activity for the learners.

It is widely argued that engaging in communicative language tasks helps learners develop their L2 in several ways. “Tasks provide an opportunity not only to produce the target language but also through conversational adjustments, to manipulate and modify it” (Foster, 1998, p.1). Task participants receive comprehensible input and generate comprehensible output, both of which have been claimed to be crucial to second language acquisition.  

Specifically, in an Asian EFL environment where learners are limited in their accessibility to use the target language on a daily basis, it is first of all necessary for language learners to be provided with real opportunities to be exposed to language use in the classroom. Despite its educational benefits in language learning contexts, a task in itself does not necessarily guarantee its successful implementation unless the teacher, the facilitator and controller of the task performance, understands how tasks actually work in the classroom.

It also suggests that TBLT as an instructional method is more than just giving tasks to learners and evaluating their performance. More importantly, the teacher, who wants to try implementing TBLT successfully, is required to have sufficient knowledge about the instructional framework related to its plan, procedure, and assessment.

TBLT can be viewed from different perspective, but there are two differentgeneral and famous perspectives. The first one is a psycholinguistic perspective, and the other is socio-cultural perspective. Form a psycholinguistic perspective, there are three different models proposed by Ellis (1999). The models are:

1) Interaction hypothesis.       

2) Cognitive approach.

3) Communicative efficiency.

The first model is Interaction hypothesis. It places the emphasis on the role of input but it claims that the best input for language acquisition is that which arises the opportunity to negotiate meaning in exchange where an initial communication problem has occurred. This idea is introduced by Long (1983). Here tasks play crucial role since they can provide the conditions for language development to occur.

The second model is cognitive approach. In this approach the base is distinction in which learners are taught to show their L2 knowledge. Learners are believed to be like native speakers so they construct not only communication system, but also rule-based system. Based on this proposal Skehan(1998) distinguishes three aspects of learner performance: fluency, accuracy and complexity . For him fluency can refer to the learners' capacity to communicate in real-life situations, out of classroom atmosphere; accuracy is the learners' ability to use language according to the aimed language norms, and complexity is viewed as the ability to use more elaborate and complex language.

     In this model, tasks are also ' important but what should be taken into consideration is to find out what task variables and dimensions promote and emphasize fluency, accuracy, and complexity.

     Furthermore, the third model proposed by Yule (1996)is ‘communicative effectiveness’. Yule, Powers, and Macdonal (1992) suggest that “L2 communicative effectiveness in an information transfer task will be enhanced when the speaker is led to think primarily about the listener’s needs rather than the form of the speaker’s message” (p.250). In this model the task implementation is important since it affects the skillfulness of L2 learners ' performance.

     The socio-cultural perspective puts more emphasis on the collaborative learning in which learners construct knowledge of language collaboratively. This perspective is the fruit on the Vygotsky ' s  work (1986 ) and it looks at how tasks are jointly accomplished by learners .For Vygotsky the dialogic interaction is an essential trigger for learning a language . In the interaction process when learners interact with each other, their cognitive processes are also activated. The two processes, inter- psychological and social stage, later on move to individual's internalprocess. In other words, by participating in dialogs with others learners can internalize language well. In the classroom just through doing the joint completion of tasks learners can learn the language better. In this model the important part is how learners perform the task rather than the individual task itself. Moreover, in this theory mediated learning and mental activities like attention, planning, and problem-solving are important.

A main objective in researching language tasks has been to identify a set of task characteristics based on the assumption that learner performance varies according to tasks characteristics (Skehan, 1996, 1998). Robinson (2001) propose distinctions between cognitively defined task complexity, learner perception of task difficulty, and the interactive conditions under which tasks are performed.

Task complexityis “the result of the intentional, memory, reasoning, and other information processing demands imposed by the structure of the task on the language learner” (Robinson, 2001, p.29); moreover , it is also the variation in the intrinsic cognitive processing demands of tasks , which will explain the within-learner variation in successfully completing any two tasks .The cognitive factors contributing to complexity are a consequence of the structure of the task which imposes resource demands , and to do the task effectively depends on the resource that a learner brings to the task. It has two subcategories;

a) Resource-directing.

b) Resource-depleting.

Task conditions is the interactive conditions under which tasks are performed , and it does not include neither task factors nor learner factors alone , but rather participant factors .It has two subcategories ;

a) Participant variables.

b) Participation variables.

Taskdifficulty includes learners' factors which may make a task more or less difficult that also is the learners ' perceptions of task demands, and it will explain the between-learner variation in performing one task. For instance “tasks that are unfamiliar information, involvenumerous steps for completion… are considered more difficult to perform” (Taguchi, 2007, p.113). Task difficulty is  a consequence of differentials between learners in their available intentional , memory , and reasoning resource pools (Gopher 1992 as cited in Robbinson 2001). It has two subcategories; a) Affective variable; b) Ability variable. According to Michel , Kuiken and Vedder (2008) , task complexity is the amount of cognitive processing that is needed to perform a task . Task conditions, which includes interactive factors or interactive tasks, is the kinds of tasks in which learners perform tasks in pairs , and this gives them the opportunity for negotiation of meaning , like clarification requests and comprehension checks .

Based on the above mentioned theoretical assumptions, this study has attempted to investigate the effects of task complexity, tasks difficulty and task conditions on the learners’ grammatical accuracy while they are involved in two types of tasks: 1- Here and now (the ones in which participants are required to write a narration based on the pictures), 2- There and then (in which participants are required to write about their own past memories). Accordingly, the following questions were asked:

1. Does Task Complexity (here-and-now, there-and- then) affect the grammatical accuracy of the EFL learners?                            

2. Does Task Conditions (gender) affect the grammatical accuracy of the EFL learners?

3. Dose Task Difficulty (proficiency) affect the grammatical accuracy of the EFL learners?

4. Does the interaction among Task Complexity, Task Difficulty, and Task Conditions have any effect on the grammatical accuracy of the EFL learners?

Method

Participants

This study was conducted with 60 EFL learners (boys and girls) who are studying English at Goldis language institute. They were 18-24 years old at both lower-intermediate and intermediate levels. For homogeneity of the students prior to research two proficiency tests PET and KET were given. These 60 students were divided into 4 groups randomly. The precise classification is given in Table below:

Table  1

The Precise Classification of The Participants

Each group worked on 2 special task types and their performances were carefully scored . The learners were not aware of the research purpose. Their performances were scored according to some established criteria.

Instrumentation

Two tests of general English with different questions were used.For lower-intermediate group KET was utilized, on the other hand, PET was administrated for the intermediate groups. The tasks were based on two types of writings in the first group participants were asked to do the here-and-now task, i.e. write about a picture story "A dream holiday", (See Appendix A). In the second group they were asked to do there-and- then task which was to write about their last birthday. In both task type they had to use simple past tense.

Tasks

This research has used two task types proposed by Robbinson( 2001 ) : here-and-now; and there-and-then tasks. In here-and-now task the participants were given the picture " A dream holiday " to write about with using simple past and in there-and-then ,they were required to write about their last birthday .

Procedure

The subjects involved in this study were randomly divided into 4 groups i.e.2 male and 2   female groups. Under the here-and-now task, participants in all groups – lower-intermediate ( male and female) , intermediate  (male and female)-were given the picture which was about "A dream holiday'' which is", a story of a boy who was planning to go the seaside with groups of friends at school and he wanted to relax since in the school there were bunch of students fighting and making lots of noise and he wanted to avoid the crowd and those students .When he got into the beach he saw all those students who were quarreling  are again at the beach and this made him angry.

The participants were given the picture and one hour to write about the given picture in the classroom. The researcher based his study on the findings of other studies (e.g.Robbinson,2001) in which the here-and-now task was given with the same amount of time .The participants were asked just to use simple past , not other tenses (See Appendix A) .In there-and-then task, the participants were supposed to write about their last birthday , about the activities and their feelings. In this task also the focus was on the use of simple past, not other tenses. Fortunately, all of the participants in both here-and-now and there-and-then tasks cooperated well and the researcher' instructions and explanations were followed, thoroughly, because the participants followed all the procedures .

Design

This study had a qusai-exprimental design with certain dependent and independent variables. The independent variables were Task Complexity (Here-and-now & There-and-then), Task Difficulty (Lower-intermediate & Intermediate), and Task Condition (Male & Female). The dependent variable was the participants Accuracy, correct use of Simple past." Accuracy is the ability to produce well-form and error-free clauses or sentences". (Yuan & Ellis, 2003; p.282).

Measures

In the present study the data were coded for T-units, defined as “A finite clause together with any subordinate clauses dependent on it.” (Bygate, 2001, p.35).So the Accuracy was measured based on the following rule:

The number of each ungrammatical error

            Accuracy = The number of every T-Unit

Accuracy was reflected by calculating the incidence of errors per T-unit, the higher the number, the less accurate the language.

Results

In order to answer research questions the data were submitted to three statistical analyses: a) Paired t-test b) Independent t-test c) Multivariate analyses of variance (two-way ANOVA).

The first research question in this study addressed the effect oftask complexity on the accuracy of L2 learners in the written discourse as a whole without considering any other variables. To answer the first research question , the researcher conducted paired t-test .The accuracy of the learners ' in the written discourse of the present study was measured through dividing the number of each ungrammatical error by the number of every
T-unit .Here accuracy was measured as a single unit and just the mentioned factors in above were considered important . In a further research focusing only on accuracy measurement other elements of accuracy can be considered, too. As Table 4.2. shows, the difference between theparticipants 'accuracy in performing two task types (Here-and-now, There-and-then) was significant ( t (58) = 2.54 , p=.03) .

Table2

Descriptive data and paired t-test for Accuracy measurement in Task Complexity

It means that in performing the two different task types in different groups there is a significant variation and effect on the participants' accuracy measurements .As the descriptive data in Table 4.2. shows in performing the two task tapes Here-and-now,as well as There-and-then ,the mean score for Here-and-now was (2067), and for There-and-then was (1820).The participants were more accurate in performing There-and-then task type ,since its mean score is less.As in this study accuracy was measured based on the number of errors in every T-unit ,the less mean score is , the more accurate the writing will be .This can be because There-and-then task is a kind of personal task ,and in personal tasks learners feel free of any language barriers  since they are cognitively less demanding. Moreover ,the topic is related to their own life, and is something known to them related to their feelings, as a result they perform well on these task types .The findings of the study are in line with findings of the previous studies (e.g.,Robbinson,2001), in which participants were asked to perform the same tasks under the same conditions. In that study the participants also were more accurate in There-and-then task, maybe the topic was more familiar and known to them. But the results are not in line with Kuiken and Vedder 's (2007)study in which the participants ' accuracy in two task types were the same ,and the difference between them was not significant .

The concern of the question 2 in the present study was to figure out if there was any significant variation in the condition measures of male and female subjects in the task performance.To answer the second research question in this study the paired t-test was used .The results of paired t-test as presented in Table 4.3. indicate no main effect for Task Conditions (Sex variable) in the case of accuracy measurement (t (58)=1.41 ,p=.159).

Table 3

Descriptive date and paired t-test for Accuracy measurement in Task Conditions.

As it can be inferred by this statistical result this variable (Gender) is apparently not associated with any significant effect on accuracy . The findings of this study related to gender variation in task performance are not in line with the previous studies: Argamon , Koppel , Fine, and Shimon  (1998 ), for instance , did the study about learners' performances,both male and female,in fiction and non-fiction documents,and to find out  which group is more accurate.  In this study results show that there is a significant variation in their writings. In another study about sex variation in task performance, Shehadeh(1994) have collected data from 35 participants ,who were all adults (16 male,19 femal).The participants were required to do three communicative tasks – a decision making task , a picture dictation task, an opinion-exchange tasks. The findings of this study have shown that male tend to be more active and gain lots of  benefits in group activities than women.

     Question 3 concerns the impact of task difficulty (proficiency) on L2 accuracy. As the data in Table 4.4 reveal , there has been a significant variation between these two different proficiency levels ( t (58)=2.26 ,p=.026 ).

Table 4

Descriptive data and paired t-test for Accuracy measurement in Task Difficulty .

It means that the proficiency variable has a main effect on the accuracy measurement. Since Accuracy, in this study,is the number of ungrammatical errors in each T-units ,so the less mean score is,the more accurate that writing will be.The mean score in intermediate group was (2247), whereas in lower- intermediate group was (1640). The results reveal that the learners, regardless  of their gender and performed task type,were more accurate in lower-intermediate level.The results of the present study may seem strange to most people,because it is generally accepted that the higher the level of education is,the more accurate those learners would act.However, in the present study the results indicate that the lower-intermediate groups were more accurate,this may be because the learners in lower-intermediate paid more attention in performing the tasks.In lower-intermediate levels learners are not sure of their language abilities as a result they pay more attention to what they are doing,then they have less ungrammatical errors.Moreover,in the current study the number of ungrammatical errors were considered important not the content .It can be said that the intermediate groups might have done well in the content if it was considered.Theu did not do well in writing accurately , may be they paid less attention to what they were doing .This result can also be in line with the notion of Stimulus Generalization in moderate version  of Contrastive Analysis stated by Oller and Ziahosseiny (1970).In which they studied the two groups of students' spelling errors.One of the groups had the same Roman writing system like English, where as the other group had the non-Roman writing system .Their findings revealed that the learners in non-Roman writing system, which could also be considered odd, out performed the ones in the Roman writing system.They concluded that when there are similarities between two things, learners pay less attention to what they are doing so they have more errors. Then it can be said that attention is a very important factor in doing any task.

In order to answer the question 4 in the current study, the data were submitted to more statistical analyses : multivariate analyses of variance (Two- way ANOVA) .As the data in Table 4.5 reveal , there was significant interaction among the three independent variables (F(58)=5.462 ,p= .021).

This means that when these independent variables were considered two by two, there was no significant variation (interaction) between them; however,when the three of these independent variables came together ,the interaction among them was significant (F (58) =5.462 , P=.021). The findings of this study are in line with the findings of the study conducted by Robbinson (2001) in which he studied the interaction among all three variables.

Table.5

Test of between-Subject Effect Dependent Variable :SCORE.

Discussion

This research posed questions to find out if there is any evidence that learners benefit from Task types (Robbinson,2001). In summary, using a series of measures ,the researcher found some evidence that participants perform differently based on the Task types that they do, and also they do differently based on their levels,but the gender variation did not have any effect on their performance.The findings of the study are supported by Skehan 's and Foster 's  (1999) notion that personal tasks are cognitively less demanding ,and participants perform very well while performing them .Since in the recent study the mean score in there-and-then task type,in which participants were required to write about their own last birthday, is less than here-and-now, as a result in there-and-then task type ,participants were more accurate. The results of the study; however,did not show any significant variation between male and female participants 'performance .The accuracy measurement was not significantly different for male and female participants when they were performing two various task types in two different proficiency levels.The interesting result of the study was that participants in lower-intermediate groups perform better than the participants in intermediate groups, which means that lower-intermediate groups were more accurate. This is somehow against the generally accepted idea that learners perform well in their higher levels of education, and are more accurate. This finding can be due to the fact that participants in lower-intermediate groups pay more attention to their performances since they are not sure of their language abilities. Moreover, in this study accuracy was measured by the correct use of simple past ,not the content .The intermediate groups might have performed well if the content had been studied.

     The present study supports the findings of the previous researches regarding Task type (e.g., Robbinson, 2001; Skehan and Foster, 1999; Vender, 1996) . The most important offering of the current study is that it gives EFL learners and EFL educators a clear explanation of how Task Complexity (Task type : Here-and-now, There-and- then), Task Difficulty (Proficiency:Lower-intermediate, Intermediate),and Task Conditions (Gender : Male and Female) affect the EFL learners ' grammatical accuracy in the written discourse .In terms of language teachers the implication is that they should try hard to involve their learners in performing tasks in the classroom which promote a real-life communication ,and is very beneficiary for learners .Since the aim of language learning , in general , is being able to communicate with the language in real world situations , independently of books ,cassettes, Cds, and all other language teaching aids, task –based language teaching by using tasks of any kind promote the authentic situations for learners .When learners are involved in performing tasks they produce fluent ,accurate ,and complex language , as it was suggested by Ellis (2003) "Involving  learners in performing tasks make them forget the language forms and try to convey meaning and get just their message across " (p.25).

     The findings also suggest that teachers should not underestimate the learners' abilities in lower levels of their language learning programs because teachers are dealing with cognitively aware people who can understand easily what they are expected to do if the task rubrics are stated clearly; moreover, they should also consciously participate their learners in what they are performing, whether oral or written, since attention is a very useful factor in EFL classrooms.

     The results of the study can also have good implications and insights for the material designers .It is high time that Iranian material designers should have stopped focusing only on language forms, and providing learners with a linear syllabus, in which syllabus items are only dealt with once .The books should be replaced by spiral syllabus, and in these spiral syllabus the tasks of any kind can be used and the results would be more efficient.

References

Bygate , M. (1999) .Quality of language and purpose of task : Patterns of learners ' language on two oral communication tasks. Language Teaching Research , 3 ( 3 ) , 185-214.

Bygate , M. ,Skehan , P. , & Swain , M . (2001) .Researchingpedagogical tasks second language learning , teaching and testing .Harlow : Longman .

Ellis, R.(1999). Input-based approaches to he teaching of grammar. AnnualReview of Applied Linguistics, 19, pp. 64-80.

Ellis , R. (2000). Task-based research and language pedagogy .LanguageTeaching Research , 4 ( 3 ) ,p.p. 193-220 .

Ellis , R. (2003).Task-based language learning and teaching .Oxford : Oxford University Press .

Ellis , R. (2008).The methodology of Task-based Teaching .AsianEFL Journal. Retrieved April 12 ,2008 . http:// www.asian –efl-journal.com

Foster ,P. (1998). Aclassroom perspective on the negation of meaning . Applied linguistics ,19.pp.1-23.

Foster ,P. &Skehan ,P. (1999) . Criteria for Classifying and SequencingPedagogic Tasks .Investigating Tasks in Formal Language Settings,Multilingual Matters ,51 ,p.p.311-410.

Gilbert , R. (2007) . Effects of manipulating task complexity on self-repairs during L ( 2 ) oral production .Applied Linguistics, 45 ,p.p. 261-284 .

Heaton , J.B. (1975) .Beginning Composition through pictures .London : Longman , Fuman University Library .

Kuiken , F. &Vedder , I. (2007) .Task complexity and measures of linguistics performance in L ( 2 ) writing . Applied Linguistics , 45 , p.p. 241-259 .

Keshavarz , M.H. (2000).Tehran :Contrastive Analysis & Error Analysis. Rahnama Publications .

Kuiken , F ,&Vedder , I. (2008). Cognitive task complexity and writtenoutput in Italian French as a foreign language .Journal of second language writing ,17 ,p.p. 48-60 .

Long , M. (1983). Native speaker non-native speaker conversation andthe negotion of comprehensible input.Applied Linguistics , 4.pp.126-41 .

Long ,M. (1985). A role for instruction in second language acqusition : task-based language teaching.(In K.Hyltenstam and M.Pienemann leads) :Modelling and assessing second language acquisition.Clevedon : Multilingual Matters .              

Nunan , D. (1989). Designing tasks forthe communicative classroom . Cambridge : Cambridge University Press .

Nunan , D. (1989). Task-based language teaching .Cambridge : Cambridge University Press .

Prabhu , N.S. ( 1987 ) .Second language pedagogy .Oxford : OxfordUniversity Press .

Robbinson , P. (2001) .Task complexity , Task difficulty , and Task production :Exploring Interactions in a Componential Framework .Applied Linguistics , 22 ( 1 ) ,p.p.27-57 .

Robbinson , P. (2006) .Criteria for classifying and Sequencing Pedagogic Tasks .Investigating Tasks in Formal Language Settings , Multilingual Matters , 31 ,p.p.118-131.

Skehan ,P. (1996 a) .A Framework for the implementation of task-basedinstruction .Applied Linguistics ,17 ( 1 ), 38-61 .

Skehan , P. (1996) . A framework for the implementation of task-based instruction . Applied Linguistics , 17, 38-62 .

Skehan , P. (1998) . A cognitive approach to language learning .Oxford : Oxford University Press .

Taguchi , N. (2007) .Task difficulty in oral speech act production . Applied Linguistics ,28.1, 113-135.

Vygotsky ,L.S. (1986). Thoughts and Language .Cambridge : The MIT Press .

Yule ,G. (1996). Referential Communication tasks .Mahwah , NJ : Lawrence  Erlbaum .

Yule , G. , Powers ,M. ,& McDonald ,D.(1992) . The narrative effects of some task-based learning procudures on LZ communicative effectivness . Language Learning , 42, 249-277 .

Biodata

SaeidehAhangariis an assistant professor of English language teaching at Islamic Azad university, Tabriz Branch. She is holding Ph.D. in TEFL. Her main interests are task-based language teaching, language testing and systemic functional linguistics. She has published many papers and participated in a number of international conferences.

MahlegaAkbari is a Ph.D. candidate of TEFL in Tabriz Branch of Islamic Azad University. She has been teaching English in language institutes for many years. At the present she is teaching English in Payam- E- Nour University and JameElmiKarbordiUniversity.She is interested in Task-based language teaching.