Volume 5 (2014) Issue 2 - Article Ruiz-Funes
JLLT Volume 5 (2014) Issue 2.pdf

Journal of Linguistics and Language Teaching

Volume 5 (2014) Issue 2 (PDF)

pp. 181-206

An Exploration of Task Complexity in Foreign Language Writing at the Intermediate Level and Measures of Linguistic Production


Marcela Ruiz-Funes (Statesboro (GA), USA)


Abstract (English)

This study expands on the emerging course of research on cognitive task complexity in second / foreign language (L2/FL) writing (Kormos 2011, Kuiken & Vedder 2007, 2008, Ong & Zhang 2011; Byrnes & Manchón 2014, Ruiz-Funes 2014). Specifically, it sheds light on the interplay between task complexity and linguistic production within the domain of L2 / FL writing focusing on FL learners of Spanish at an intermediate level of language proficiency based on the ACTFL Proficiency Guidelines (ACTFL 2012). The theoretical framework of this investigation is Skehan & Foster’s Trade Off / Limited Attentional Capacity Model and Robinson’s Triadic Componential Framework / Cognition Hypothesis. Twenty-four undergraduate learners of Spanish wrote two essays of different levels of cognitive complexity in relation to familiarity of topic, genre, and reasoning demands that were analyzed for syntactic complexity, accuracy, and fluency (CAF). Data analysis was conducted in two ways: (i) by task type only and (ii) by task type and level of performance on assigned tasks. Findings indicate a tendency towards support for the Trade-Off / Limited Attentional Capacity model with no differing effect based on the level of performance of the students on the assigned tasks.

Keywords: Task complexity; syntactic complexity; accuracy; fluency; proficiency level; performance level

Abstract (Español)

Este estudio amplía la investigación emergente sobre la complejidad de tareas cognitivas en la escritura en un segundo idioma (Kormos 2011, Kuiken y Vedder 2007, 2008, Ong y Zhang 2011, Byrnes & Manchón 2014, Ruiz-Funes 2014). Específicamnete, este estudio explora la relación entre la complejidad de tarea y la complejidad lingüística dentro del dominio de la escritura en un segundo idioma centrándose en estudiantes de español como lengua extranjera a un nivel intermedio basado en las directrices de competencia de ACTFL (ACTFL 2012). El marco teórico de esta investigación es el modelo de Skehan y Foster de Intercambio / Capacidad de Atención Limitada y la Hipótesis Cognitiva de Robinson. Veinticuatro estudiantes universitarios de español escribieron dos ensayos de diferentes niveles de complejidad cognitiva en relación a la familiaridad del tema, el género y las demandas de razonamiento los cuales se analizaron según la complejidad sintáctica, la precisión y la fluidez (CAF). El análisis de datos se llevó a cabo de dos maneras : ( i ) por tipo de tarea solamente y ( ii ) por tipo de tarea y el nivel de rendimiento en las tareas asignadas. Los resultados indican una tendencia hacia el apoyo al modelo de Intercambio / Capacidad de Atención Limitada sin efecto diferente en función del nivel de rendimiento de los estudiantes en las tareas asignadas.

Palabras claves: Complejidad de tarea; complejidad sintáctica; precisión; fluidez; nivel de proficiencia; nivel de rendimiento


1 Introduction

Task-based language teaching (TBLT) has had a significant impact in the field of second language acquisition (SLA) and has widely influenced the exploration of tasks and their effect on language learning and development. Particularly, research in this area has focused on

understanding how task characteristics can affect the second language acquisition (SLA) processes involved in learning while attempting to meet the challenges certain tasks set”. (Robinson 2011: 4).

The vast majority of the investigation in this field has dealt primarily with task complexity within the oral production domain (Samuda & Bygate, 2008, Van den Branden, Bygate & Norris 2009, Robinson 2011). As a consequence, and as pointed out in earlier studies (Byrnes & Manchón 2014, Kormos 2011, Kuiken & Vedder 2008, Ong & Zhang 2010; Ruiz-Funes 2014), research on the effect of tasks in writing, and L2 / FL performance and development continues to be scant.

This investigation aims to help expand the literature base by advancing on the emerging course of research on cognitive task complexity in L2 / FL writing. It brings further understanding to the relationship between writing tasks, linguistic production, and language proficiency by providing additional empirical evidence to that offered by Kormos (2011), Kuiken & Vedder (2007, 2008, 2011, 2012), Ong & Zhang (2011), and Ruiz-Funes (2014), among others. Specifically, it focuses on FL learners of Spanish with an intermediate level of language proficiency based on the ACTFL Proficiency Guidelines (ACTFL 2012) and follows a similar experimental research design from a previous study (Ruiz-Funes, 2014) conducted in terms of types of analysis conducted and linguistic measures computed. Twenty-four undergraduate learners of Spanish wrote two tasks of different levels of cognitive complexity in relation to familiarity of topic, genre, and reasoning demands which were analyzed for syntactic complexity, accuracy, and fluency (CAF). The data analysis was conducted in two ways: (i) by task type only and (ii) by task type and level of performance on assigned tasks. Task type refers to task complexity as indicated in Task 1 and Task 2; level of performance refers to the overall quality of the essays the students wrote, based on the assessment made by two raters.

The theoretical framework of this investigation is Skehan and Foster’s Trade Off / Limited Attentional Capacity Model (Skehan 1998a, 2001, 2003, Skehan & Foster 1999, 2001) and Robinson’s Triadic Componential Framework or Cognition Hypothesis (Robinson 2001a, 2001b, 2003, 2005, 2007). Even though these models were conceptualized for oral tasks, they provide a useful structure for exploring task characteristics and their effect on linguistic production within the field of L2 / FL writing (Kormos 2011). They

have attempted to answer the question of how attentional resources can be used, coordinated, and directed to different aspects of language production during task completion. (Kormos 2011: 50)

Since these models bring to the forefront the role of attentional resources in the task performance process and the effect that task characteristics have on such resources, the theoretical background of this study also includes the construct of working memory capacity and its function in writing processes (Baddeley 1986, Kellogg 1996, Alamargot and Chanquoy 2001).

In the present paper, first, the theoretical background of the study is presented, followed by a review of the most salient work on task complexity in L2 / FL writing. Second, the study and its findings are examined. A discussion of the results and implications for future research conclude the paper.


2 Theoretical Background

As Hyland states

[writing] tasks are fundamental in learning to write and represent a central aspect of the teacher’s planning and delivery of a writing course. (Hyland (2003:112)

Tasks are important because they affect students’ learning experiences, setting in motion a series of cognitive mental processes that writers orchestrate during all the steps of the writing process (Dvorak 1986, Flower & Hayes 1981, Flower et al. 1990, Kellogg 1996). Moreover, the notion of task, has been defined from various perspectives. From a psycholinguistics perspective, a task engages students in certain types of mental processing that leads them to effective language use and acquisition (Ellis 2009). Two of the most widely influential models that address how the cognitive demands of tasks affect L2 learning and that frame the present study are those proposed by Peter Skehan (1998, 2001, 2003, 2009) and Peter Robinson (2001a, 2001b, 2003, 2005, 2007) (Table 1). Their frameworks have been used to classify and manipulate tasks during L2 instruction (Robinson 2011) but, more importantly, to investigate the effect of task characteristics on language acquisition and development.

Skehan’s (1998a, 1998b, 2001, 2003; Skehan & Foster, 1999, 2001) Trade Off or Limited Capacity Hypothesis model, based on working memory theories (Carter 1998 and Gathercole and Baddeley 1993), claims that tasks that are cognitively more demanding “consume more attentional resources…with the result that less attention is available to focus on form” (Skehan 1998: 97). As such, it suggests that tasks be sequenced from less to more demanding ones to allow for opportunities for the learner to attend to language form and thus to increase the opportunities for task-based language learning in a balanced way where accuracy, fluency, and complexity of production are promoted. It further states that task variables can be grouped under two general categories: task features (e.g. how information is structured, what elements can be manipulated) and task implementation (including planning time, and task repetition). Both interact in complex ways that affect the output of the learner in terms of fluency, accuracy, and syntactic complexity. Specifically, Skehan (1998a, 2001, 2003) argues that learners possess a limited processing capacity, that only one of these may be in focus at a given time during a given task, and that the manipulation of task variables affects the fluency, accuracy, and complexity of production. Skehan hypothesizes that, as task complexity increases, learners will attend to only one aspect of performance – syntactic complexity, accuracy, or fluency – at the expense of the others.

On the other hand, Robinson’s model, known as the Triadic Componential Framework or Cognition Hypothesis (Robinson 2001a, 2001b, 2003, 2005, 2007), is based on information processing theories (Long 1996, Schmidt 2001, and Wickens 1992). It suggests a multiple-resources view of processing where learners can attend to multiple aspects of language and language processing simultaneously. As such, Robinson proposes that more cognitively complex tasks lead to more attention to input / output, spur rule and instance learning, and, as a result, facilitate the incorporation of input and better performance. This framework distinguishes two dimensions for task complexity: the resource-directing dimension that makes conceptual demands such as reference to past or present events, few or many elements, and more or fewer reasoning demands; and by contrast, the resource-dispersing dimension that makes procedural demands on the learner and includes planning time, prior knowledge provided in the task, and the number of tasks to complete. Task complexity along the resource-directing dimension results in increased accuracy and complexity as learners have to put their attentional resources to the demands of the task. Yet, fluency decreases as students have to process language. On the contrary, task complexity along the resource-dispersing dimension will result in decreased fluency, accuracy, and complexity levels in language production as it will limit the attentional and working memory of learners (Robinson 2003, 2005). The Cognition Hypothesis has been questioned, and support for its claims has varied (Kuiken and Vedder 2007a, 2007b; Michel, Kuiken & Vedder 2007, Robinson 2007). Kormos and Trebits (2011) suggest that this mixed support may be due to the fact that in some tasks, the two dimensions of resource-directing and resource-dispersing may be simultaneously present.

Table 1: Cognitive complexity frameworks


3 Literature Review on Task Complexity in L2 / FL Writing

The research on task complexity in L2 / FL writing suggests a relationship between task complexity and linguistic production. Yet, findings are still not consistent enough to formulate generalizations. This may, in large, be due to the emerging nature of this type of research within the domain of L2 / FL writing.

Varying patterns of connection between task complexity and syntactic complexity, accuracy, lexical variation, and fluency have been noted. For example, Kuiken and Vedder’s (2007, 2008, 2011) studies with college-level Dutch learners of Italian and Dutch learners of French, who completed two letters of different degrees of cognitive complexity in terms of number of requirements and types of decisions, showed a decrease in the number of errors and an increase in lexical variation on the more complex task. No effect was found between task complexity and syntactic complexity or the proficiency level of the students. In a more recent publication, Kuiken and Vedder (2012) explored more in depth the effect of L2 proficiency on the relationship between task complexity and linguistic production based on data from the three previous studies they conducted (Kuiken & Vedder 20007, 2008, 2011). Proficiency levels were determined by scores the participants received on a cloze test. Findings indicated no effect of task complexity on linguistic performance based on proficiency levels as defined in their studies.

Kormos (2011) explored task complexity in narration produced by upper-intermediate students of English as a foreign language (EFL) in a bilingual high school in Hungary and L1 students of English in the UK. The two narrative tasks provided different levels of cognitive complexity in terms of more / less demand for plot conceptualization. Findings indicated no significant effect in relation to task complexity and linguistic performance except that more abstract words were elicited by the more complex task.

Ruiz-Funes (2014) examined task complexity in essay writing composed by college learners of Spanish with an advanced level of language proficiency based on the ACTFL Proficiency Guidelines (ACTFL 2012). One task was an analytical essay; the second task, more complex, was an argumentative one that called for evaluation of a given argument with appropriate support. This investigation established some connections with claims in both Skehan and Foster’s (1999, 2001) and Robinson’s (2001a) processing models. Results suggested that considering task complexity, familiarity of topic, genre, and task type, along with reasoning demands, had an effect on L2 / FL written production: task complexity had a positive effect on syntactic complexity or accuracy and fluency, but not on all aspects simultaneously except when the level of performance on the assigned tasks was taken into account. In that investigation, high performance students showed somewhat higher values of syntactic complexity, accuracy, and fluency on the more complex task; low performance students on the less complex one.

Other studies explored task complexity in writing manipulating factors such as planning time, draft availability, or the here and now dimension. First, Ellis and Yuan (2004) established a positive relationship between pre-task planning on the one hand and length and syntactic structures used by students on the other; second, Ong and Zhang (2010) found a connexion between fluency and lexical variation; and third, Ishikawa (2006) showed a positive effect between the here and now dimension and syntactic complexity, accuracy, and fluency.

In all, findings point to a potentially significant relationship between elements of task complexity and written output. However, definite conclusions cannot yet be reached due largely to the multiplicity of variables present in the current investigations. As stated earlier (Ruiz-Funes 2014), there is variation in the types of written tasks manipulated including letter, narration, and essay writing, the manner in which task complexity in writing is operationalized, the language proficiency levels of the learners as well as their levels of performance on the tasks assigned, the way language proficiency is measured, and the academic levels and settings of the writers. More research should validate what and to what extent task characteristics and task conditions account for the variation in linguistic performance (Ruiz-Funes 2014). This study seeks to bring more empirical evidence to help isolate factors that interact in the relationship between writing tasks, linguistic production, and language proficiency in L2 / FL writing.


4 The Study

4.1 Research Questions

The main purpose of this study is twofold: (i) to explore the effect of task complexity on the linguistic production of college-level FL writers of Spanish at the intermediate level of language proficiency, and (ii) to investigate to what extent such relationship, if any, is linked to the level of performance of the learners as indicated by the quality of writing they produced on the assigned tasks. The following research questions guided this study:

    1. What is the effect of task complexity on syntactic complexity measured by length of production, amount of subordination, and coordination?
    2. What is the effect of task complexity on accuracy measured by number of errors per T-unit and number of errors per 100 words?
    3. What is the effect of task complexity on fluency measured by length of text produced within a given time frame?
    4. Is there a relationship between task complexity and measures of syntactic complexity, accuracy, and fluency (CAF) based on the level of performance of the learners on the assigned tasks?

4.2 Methodology

4.2.1 Participants and Procedures

Data were gathered from 24 undergraduate learners of Spanish from a large southeastern university in the United States, who attained an intermediate-mid to intermediate-high language proficiency level on an oral examination based on the ACTFL Proficiency Guidelines (ACTFL 2012). 16 students were female and eight (8) male of ages ranging between 19 and 25 years old. 21 students were native speakers of English for whom Spanish was a foreign language: Three students were heritage language speakers of Spanish born, raised, and educated in the United States. All students were majors or minors in Hispanic Studies enrolled in two or three Spanish courses at the intermediate level. In general, the students were engaged in a rather traditional curriculum that included courses in language, culture, literature, and linguistics with an emphasis in the development of language proficiency in all four skills and with some focus on genre-based tasks.

The students completed two timed written tasks (Appendix) as in-class assignments. They were allowed 50 minutes for each task which they handwrote while being supervised by their instructor. The written products were evaluated by two trained raters, who were experienced Spanish instructors at the university level and native speakers of Spanish. They used a holistic rubric that assessed students‘ composing skills in reference to organization, content, and linguistic skills regarding grammar, vocabulary, and conventional usage of written language. A minimum inter-rater agreement of 90% was reached for each essay.

Based on the raters’ evaluation, seven students received a score between 5 and 6 (excellent / sophisticated; Group 1), 14 students scored 4- (appropriate and acceptable; Group 2), and three students received a mark of 3- (inadequate; Group 3). This grouping indicated their level of performance on the assigned tasks. Group 1 was considered the high-performance group (HP), Group 2 the mid-performance group (MP), and Group 3 the low-performance group (LP) (see Data Analysis section below).

4.2.2 Writing Tasks

In each task assigned, the participants were asked to write an essay of 250 words in length on a 50-minute class session of an intermediate writing course. No dictionary, textbook, or any other resource was allowed. Task 1 was assigned on day one; Task 2 was assigned on the second day of the same week. Both tasks were framed and contextualized around the theme of study abroad, yet reflected different levels of cognitive complexity. Task 1 required the learners to write a personal essay about themselves, their interests, and expected goals for studying abroad. For Task 2, students had to write an expository essay stating formally the benefits and challenges of studying abroad for college students.

The personal essay (Task 1) was considered to be less complex than the formal expository essay (Task 2) since in the former, the writer was asked to write about his or her personal experiences without having to prove a point. The author was to introduce the subject and theme, based on personal life experiences, feelings, emotions, and opinions, using both description and narration. In the formal, expository essay (Task 2), on the other hand, the writer had to state a thesis and support his or her ideas with facts and objective evidence. For this task, the student was expected to present the information in order to inform and / or explain.

Based on Skehan and Foster’s (1999, 2001) model, Task 2 is assumed to be more cognitively complex in terms of topic, discourse genre, and task type, and more demanding of cognitive processing in terms of information organization and information type (Cognitive Complexity). Using Robinson’s (2001a) model, this task involved the orchestration of more reasoning demands (Task Complexity / Resource-Directing). The task characteristics manipulated on both tasks appear in Table 2 below.

Limited Attentional Capacity Model

(Skehan & Foster, 1999, 2001)

Cognitive Complexity

· Cognitive familiarity

· Familiarity of topic and its predictability

· Familiarity of discourse genre

· Cognitive processing

· Information organization

Triadic Componential Framework (Robinson 2001a)

Task Complexity

· Resource-directing

+/- No reasoning demands

Table 2: Tasks characteristics manipulated

In addition, based on a brief questionnaire on the participants’ demographic information and their opinion about which task was more demanding, the majority (88%) responded that Task 2 (formal, expository essay) was more difficult. Some of the reasons the students stated are: Among the reasons, they indicated, having little control of more technical, formal lexicon, lack of experience studying abroad, and the need to think more to generate and organize ideas for the completion of Task 2. The following excerpts taken from the questionnaire illustrate this :

Student A: Essay 2 (Task 2) required more vocabulary that I do not use every day.

Student B: I had to use more terms and some of which I did not know how to spell. Writing about myself (Task 1) was much easier.

Student C: Essay 2 was more challenging because I had to come up with ideas and points before I was able to write about them.

Student D: Essay 2 was more difficult because it wasn’t about myself, I had to list things.

Student E: Essay 2 because I never really studied abroad and thought it was harder for me to think some ideas to describe.

Student F: The second essay was harder to write because there were more technical terms that I wanted to use but didn’t know how to say in Spanish. Also, I couldn’t think of a lot of benefits and challenges of studying abroad”.

Student G: The second essay was more difficult to write because I have not spent a lot of time thinking about the advantages and disadvantages of studying abroad. On the other hand I have thought about myself, my goals, and interests a lot in the past.

Student H: It was more difficult to write about the challenges and benefits because I have not studied abroad yet and haven’t been able to experience the challenges / benefits.

A few students (12%), however, considered Task I (personal essay) to be more difficult. Somewhat contradictory reasons were posed such as that Task 1 was more open-ended which, in turn, hindered the generation of ideas, on the one hand, and that Task 1 constrained creativity which led to redundancy of information, on the other:

Student I: Essay 1 (Task 1) was more difficult to write because it was more open ended so I couldn’t decide what to say.

Student J: Essay 1 (Task 1) was more difficult because it left less room for me to get creative with what I wanted to talk about.

The students’ perception of task difficulty touches upon the relationship explored in this study, i.e., the effect of task complexity on cognitive processing and language production and as such, it is presented only as an additional reference that differentiates the two tasks.


5 Linguistic Production: Measuring Syntactic Complexity, Accuracy, and Fluency (CAF)

5.1 Measures of Syntactic Complexity and the Use of the T-Unit Analysis

In this study, syntactic complexity was measured, following a multidimensional approach with metrics that are distinct and complementary as suggested by Norris & Ortega (2009). The measurers included

    • length of production - Mean length of T-unit (MLTU): total number of words divided by the total number of T-units (MLTU W/TU);
    • amount of coordination - Mean number of T-units per sentence (TUS): the total number of T-units divided by the total number of sentences: (TU/S); and
    • amount of subordination - Mean number of dependent clauses per T-unit (CTU): the total number of dependent clauses divided by the total number of T-units: (C/TU)

The central unit used was the T-unit coined by Hunt (1965) as the minimal terminable unit of language consisting of a main clause plus all subordinate clauses and nonclausal structures that are attached to or embedded in it. The T-unit has been adopted extensively in L2 / FL research to measure writing development, and it is considered to be the most appropriate way to code and record changes regarding fluency, accuracy, and complexity (Arnold et al. 2009, Polio 1997, Spelman Miller 2006).

5.2 Accuracy

Errors were tallied based on an adapted version of Lalande’s (1984) Essay Correction Code-ECCO. An error was operationalized as any digression in syntactical, morphological, and lexical norms, but not in punctuation or capitalization (Ellis & Yuan 2004). A number of measures were computed: total number of errors, total number of errors per T-unit (Etot/T=total number of errors divided by the total number of T-units), and number of errors per 100 words (NER) (Mehnert 1998).

5.3 Fluency

As indicated in earlier studies (Larsen-Freeman 1978, Wolfe-Quintero et al. 1998), one marker of fluency is length of text produced in timed writing. Since the completion of both essays in this study were done under timed conditions (50 minutes for each task), the number of words produced became a rate measure for fluency calculated by the mean number of words written per minute out of the 50 minutes spent on the task.


6 Data Analysis

For data analysis, the essays were first typed in MS-Word documents and coded by the researcher and a trained graduate assistant according to the various linguistic measures for syntactic complexity, accuracy, and fluency as listed above. A minimum inter-rater reliability coefficient of .94 was achieved for each measure. The following values were calculated per student individually on each task separately:

(a) the total number of words, the total number of sentences

(b) the total number of T-units

(c) the total number of clauses, the total number of dependent clauses

(d) the ratio of dependent clauses per T-unit

(e) the total number of errors and

(f) the total number of errors per 100 words.

Following a similar data analysis procedure used in a previous study (Ruiz-Funes 2014), the values of syntactic complexity, accuracy, and fluency were obtained in two ways:

(i) per task type and

(ii) per task type and level of performance achieved on the writing tasks.

The category task type refers to task complexity as indicated in Task 1 and Task 2. The category level of performance refers to the quality of the essays the students wrote based on the raters’ assessment. It includes three subgroups:

    • High performance group (HP; n=7), students who scored 5/6- (excellent / sophisticated)
    • Mid-performance group (MP; n=14), students with a score of 4- (appropriate and acceptable) and
    • Low-performance group (LP; n=3), for those with a mark of 3- (inadequate).

Due to the relatively small number of participants, which constitutes one of the weaknesses of this study, only descriptive statistics were used for data analysis. The values calculated include: group mean, mean standard deviation, mean standard error of measurement, and the 95% confidence intervals (CIs) above and below the means for each type of analysis to make the relevant comparisons.


7 Results

7.1 Analysis per Task-Type

The descriptive statistics for the measures of syntactic complexity showed higher values on Task 2 (the more complex task) than on Task 1 for MLTU (overall length of production) and CTU (subordination), and lower on Task 2 for coordination (TUS), indicating a higher level of syntactic complexity on Task 2 than on Task 1. The group mean standard deviation and standard error values for all three measures, on the other hand, were almost identical in both tasks which shows a similar average variability for individuals on each task. This result suggests that individual students produced T-units and clauses of similar varying lengths on Task 2 and on Task 1 (Table 3).

The analysis of variance to compare group mean values for MLTU was calculated, using the 95% confidence intervals (CIs p < .05). The CIs obtained indicated no statistically significant difference of the means as the 95% CI values overlapped (Figure 2).

Table 3: Descriptive statistics for syntactic complexity measures by task type (n = 24)

Figure 1: Differences in mean lenth of T-unit (MLTU) by task type

In regards to accuracy, the group values for the number of errors per T-unit (Etot / T) and the number of errors per 100 words (NER) showed higher levels of errors on Task 2 than on Task 1, with different levels of individual variability within each task, as indicated by the index of standard deviation for NER (Table 4). The CIs obtained indicated no statistically significant difference of the means for any of these measures as the 95% CI values overlapped (Figures 2 and 3).

Table 4: Descriptive statistics for measures of accuracy by task type; n = 24

Figure 2: Differences in mean number of errors per T-unit (Etot/T)

Figure 3: Differences in mean number of errors per 100 words (NER)

For fluency, the values indicated a higher mean number of words per minute out of 50 minutes on Task 1 (the less complex task) than on Task 2 (Table 5). Yet, no significant difference was found as shown by the 95% CI values (Figure 4).

Table 5: Descriptive statistics for fluency by task type; n = 24

Figure 4: Differences in mean number of words per minute in 50 min on task

7.2 Analysis per Task-Type and Level of Performance

The results in the category analysis per task type and level of performance showed a similar pattern to the one noted in the analysis per task type only. The values for syntactic complexity, particularly for length of production measured by the mean length of the T-unit, were higher and with more variability for all three groups on Task 2 (the more complex task). For coordination, results indicated slightly higher mean values on Task 2 for the LP performance group, whereas the other two groups showed slightly higher values of coordination on Task 1. The variability for all the groups on each task was almost identical (Table 6). However, no statistical difference was found between any of these measures, which may be due largely to the small sample in each subgroup.

Table 6: Descriptive statistics for Syntactic Complexity by Task Type and Level of Performance

(n=3 for LP; n=14 for MP; n=7 for HP)

All measures of accuracy indicated higher values of errors with more variability on Task 2 for all the three groups except for the LP group that showed almost identical values on the measure of number of errors per 100 words (NER) on both tasks (Table 7). In fluency, slightly higher values were obtained for all the three groups on Task 1 with almost identical variability except for the HP group (Table 8).

Table 7: Descriptive statistics for accuracy by task type and level of performance

(n=3 for LP; n=14 for MP; n=7 for HP)

Table8: Descriptive statistics for fluency by task type and level of performance

(n=3 for LP; n=14 for MP; n=7 for HP)


8 Discussion

Results from both analyses - per task type and per task type and level of performance - suggest an increase on Task 2 (the more complex task) in syntactic complexity and a decrease in accuracy as well as in fluency.

Based on Skehan and Foster’s Trade Off / Limited Attentional Capacity Model (Skehan & Foster, 1999, 2001), these findings are in line with the claim that as task complexity increases, learners will attend to only one aspect of language production - syntactic complexity, accuracy, or fluency - at the expense of the others due to the learners’ limited processing capacity (Skehan 1998a, 2001, 2003). In particular, Skehan and Foster (1999, 2001) propose that there is tension between complexity and accuracy, meaning that either an increase in complexity or in accuracy occurs at a time, but not in both together. In the present study, a tendency towards this tension between complexity and accuracy along with fluency has been observed, supporting the claims of the Trade-Off / Limited Attentional Capacity model.

Moreover, the same tension between syntactic complexity and accuracy and fluency was found across the levels of performance on the assigned tasks. The three subgroups (HP, MP, and LP) showed evidence of this tension between higher values in measures of syntactic complexity with lower levels of accuracy and fluency under Task 2 (the more complex task). This also suggests support for the Trade-Off / Limited Attentional Capacity model when dealing with students at the intermediate level, regardless of their level of performance on the assigned tasks. It can be inferred, then, that none of the subgroups was ready for the challenge that a more complex task may put on attentional resources as Robinson’s Cognition Hypotheses (Robinson 2001a, 2001b, 2003, 2005, 2007) would suggest. Robinson proposes that more cognitively complex tasks lead to increased accuracy and complexity as learners have to put their attentional resources to the demands of the task. However, this principle did not hold true with learners at the intermediate level of language proficiency, irrespective of their level of performance on the assigned tasks.

The results from this study are only partially in line with those obtained in a previous research conducted on task complexity with advanced language proficiency learners of Spanish (Ruiz-Funes 2014). In such an investigation, a positive effect between task complexity and syntactic complexity or accuracy and fluency was found, but not on all aspects simultaneously except when the learners’ level of performance on the assigned tasks was taken into account. Thus, high performance students showed higher values of syntactic complexity, accuracy, and fluency on the more complex task while low performance students showed higher values on the less complex task (Ruiz-Funes 2014). Contrary to those results, in the present study, the values from both types of analysis indicate that regardless of the level of performance, the students at the intermediate level of language proficiency showed higher values of syntactic complexity and lower values of accuracy and fluency on Task 2 (the more complex task) than on Task 1 (the less complex task).

Furthermore, the findings of the present study did not support results from other studies that share a related research design in terms of variables used. For example, Kuiken and Vedder (2008, 2011, 2012) found a decrease in errors on the more complex task with no interaction effect between task complexity and syntactic complexity or with the writers' level of expertise. Kormos’ (2011) results indicated no significant effect in relation to task complexity and linguistic production, except in terms of an increase in abstract words on the more complex task. This lack of consistency in the findings may be linked to the emerging nature of the research in this field within the domain of writing and the variability in genres, the learners' academic levels, their language proficiency levels and their performance levels manipulated across studies (Ruiz-Funes 2014).


9 Conclusions

The results obtained showed a relationship between task complexity and linguistic production that needs to be further explored within a methodology framework that includes a much larger sample size. Like earlier studies, this study also shows a weakness in its relatively small sample size; hence its findings are preliminary. However, some clear trends were observed.

Findings indicate a tendency towards support for the Trade-Off / Limited Attentional Capacity model, with no differing effect based on students' level of performance on the assigned tasks. Evidence was found in support of the claim that task complexity, as determined by familiarity of topic, genre, and task type, along with reasoning demands, seemed to have an impact on CAF measures. The more complex task led to more syntactic complexity at the expense of less accuracy and less fluency.

In addition and more revealing, findings from this study suggest that the relationship between task characteristics in L2 / FL writing and their effect on attentional resources and CAF measures may be associated with students' language proficiency level (Norris & Ortega, 2009). At the intermediate level of language proficiency, regardless of the level of performance on the assigned tasks, the trade-off effect among CAF measures seems to be more prominent. In other words, for learners whose writing and language skills are not yet fully developed, tasks that are too complex may have a negative effect on some aspects of language production.

To sum up, it can be stated that this interplay between task complexity, CAF traits, and proficiency levels in writing needs to be further explored in light of the call made by Norris & Ortega (2009) about the importance to include proficiency considerations in the interpretation of empirical CAF findings. As they remarked:

Different qualities of production, alone or combined, can have lesser or greater predictive value depending on the relative proficiency levels of L2 learners under investigation. That is, qualities of L2 production cannot and should not be considered constant along the CAF developmental continuum, and different CAF traits must serve different interpretive purposes for different proficiency levels (Norris & Ortega 2009: 573).

Moreover, the lack of consistency in the findings across other studies presses for a more defined research agenda. As Kuiken and Vedder (2008) also noted, further work is needed to determine which variables should be included in a model of task complexity in L2 / FL writing, how these variables should be differentiated and prioritized, and how levels of language proficiency and levels of performance on assigned tasks are to be defined and manipulated.


References

Abu-Rabia, Salim (2003). The Influence of working memory on reading and creative writing processes in a second language. In: Educational Psychology 23 (2003), 209–222.

ACTFL Proficiency Guidelines Speaking 2012. American Council on the Teaching of Foreign Languages. (http://www.actfl.org/publications/guidelines-and-manuals/actfl-proficiency-guidelines-2012/english/speaking).

ACTFL Proficiency Guidelines Writing 2012. American Council on the Teaching of Foreign Languages. (http://www.actfl.org/publications/guidelines-and-manuals/actfl-proficiency-guidelines-2012/english/writing).

Alamargot, Denis & Liu Chanquoy (2001). Through the models of writing, Boston: Kluwer Academic Publishers.

Appel, Gabriela & James P. Lantolf (1994). Speaking as mediation: A study on L1 and L2 text recall tasks. In: Modern Language Journal 78 (1994), 437–452.

Arnold, Nike, Lara Ducate & Claudia Kost (2009). Collaborative writing in wikis: Insights from culture project in German class. In Lomicka, Lara & Gillian Lord (Eds.) (2009). The next generation: Social networking and online collaboration in foreign language learning. CALICO Monograph Series Volume 5. San Marcos, TX: Texas State University, 115-144.

Baddeley, Alan D. (1986). Working memory. Oxford: Oxford University Press.

Becker, Anne (2006). A review of writing model research based on cognitive processes. (http://www.wac.colostate.edu).

Bereiter, Carl & Marlene Scardamalia (1987). The psychology of written composition. Hillsdale, NJ: Lawrence Erlbaum.

Bygate, Martin (2001). Effects of task repetition on the structure and control of oral language. In: Bygate, Martin, Peter Skehan & Merrill Swain (Eds.) (2001) Researching pedagogical tasks: Second language learning, teaching and testing. Harlow: Pearson Education, 23-48.

Byrnes, Heidi, Hiram H. Maxim & John M. Norris (2010). Realizing advanced foreign language writing development in collegiate education: Curricular design, pedagogy, assessment. In: Modern Language Journal 94 (2010) Supplement s-1, i–iv.

Byrnes, Heidi & Rosa M. Manchón (2014). Task-based language learning: Insights from and for writing. Philadelphia/Amsterdam: John Benjamins.

Carter, Rita (1998). Mapping the mind, Berkeley, CA: University of California Press.

Cumming, Alister (1990). Metalinguistic and ideational thinking in second language composing. In: Written Communication 7 (1990), 482–511.

Dvorak, Trisha (1986). Writing in the foreign language. In: Wing, Barbara H. (Ed.) (1986). Listening, reading and writing: Analysis and application. Middlebury, VT: Northeast Conference, 145-167.

Ellis, Rod (2009). Task-based research and language pedagogy. In: Van den Branden, Kris, Martin Bygate, & John M. Norris (Eds.) (2009). Task-based language teaching: A reader. Philadelphia/Amsterdam: John Benjamins, 109-129.

Ellis, Rod & Fangyuan Yuan (2004). The effects of planning on fluency, complexity, and accuracy in second language narrative writing. In: Studies in Second Language Acquisition 26 (2004) 59–84.

Flower, Linda (1990). The role of task representation in reading-to-write. In: Flower, Linda, Victoria Stein, John Ackerman, Margaret J. Kantz, Kathleen MacCormick, & Wayne C. Peck (Eds.) (1990). Reading-to-write: Exploring a cognitive and social process. New York: Oxford University Press, 35-73.

Flower, Linda & John R. Hayes (1980). The cognition of discovery: Defining a rhetorical problem. In: College Composition and Communication 31 (1980) 21–32.

Flower, Linda & John R. Hayes (1981). A cognitive process theory of writing. In: College Composition and Communication 32 (1981) 365–387.

Flower, Linda, Victoria Stein, John Ackerman, Margaret J. Kantz, Kathleen MacCormick, & Wayne C. Peck (1990). Reading-to-write: Exploring a cognitive and social process. New York: Oxford University Press.

Gathercole, Susan E. & Alan D. Baddeley (1993). Working memory and language. Hillsdale, NJ: Lawrence Erlbaum.

Horowitz, Daniel (1986a). Process not product: Less than meets the eye. In: TESOL Quarterly 20 (1986) 141–144.

Horowitz, Daniel (1986b). What professors actually require: Academic tasks for the ESL classroom. In: TESOL Quarterly 20 (1986), 445–462.

Hunt, Kellog W. (1965). Grammatical structures written at three grade levels. Research Report 3. Urbana (Ill): NCTE.

Hyland, Ken (2003). Second language writing. New York: Cambridge University Press.

Ishikawa, Tomohito (2006). The effect of manipulating task complexity along the (+/−Here-and-Now) dimension on L2 written narrative discourse. In: García Mayo, María d. P. (Ed.) (2006). Investigating tasks in formal language learning. Clevedon, UK: Multilingual Matters, 136-156.

Kellogg, Ronald T. (1996). A model of working memory in writing. In: Levy, Michael C. & Sarah Ransdell (Eds.) (1996). The science of writing: Theories, methods, individual differences and applications. Mahwah, NJ: Lawrence Erlbaum, 57-71.

Kellogg, Ronald T. (2001). Competition for working memory among writing processes. In: American Journal of Psychology 114 (2001), 175-191.

Kormos, Judith (2011). Task complexity and linguistic and discourse features in narrative writing performance. In: Journal of Second Language Writing 20 (2011) 148–161.

Kormos, Judith & Anna Trebits (2011). Working memory capacity and narrative task performance. In: Robinson, Peter (Ed.) (2011). Second language task complexity. Philadelphia/Amsterdam: John Benjamins, 267-289.

Kuiken, Folkert & Ineke Vedder (2007). Task complexity and measures of linguistic performance in L2 writing. In: International Review of Applied Linguistics 45 (2007) 261–284.

Kuiken, Folkert & Ineke Vedder (2008). Cognitive task complexity and written output in Italian and French as a foreign language. In: Journal of Second Language Writing 17 (2008), 48–60.

Kuiken, Folkert & Ineke Vedder (2011).Task performance in L2 writing and speaking: The effect of mode. In: Robinson, Peter (Ed.) (2011). Second language task complexity: Researching the Cognition Hypothesis of language learning and performance. Philadelphia/Amsterdam: John Benjamins, 91-104.

Kuiken, Folkert & Ineke Vedder (2012). Syntactic complexity, lexical variation and accuracy as a function of task complexity and proficiency level in L2 writing and speaking. In: Housen, Alex, Folkert Kuiken & Ineke Vedder (Eds.) (2012). Dimensions of L2 performance and proficiency: Complexity, accuracy and fluency in SLA. Philadelphia/Amsterdam: John Benjamins, 143–170.

Lalande, John F. (1982). Reducing composition errors: An experiment. In: The Modern Language Journal 66 (1982), 140-149.

Larsen-Freeman, Diane (1978). An ESL index of development. In: TESOL Quarterly 12 (1978) 439-448.

Lennon, Paul (1990). Error: Some Problems of Definition, Identification, and Distinction. In: Applied Linguistics 12 (1990) 2, 180-195.

Long, Michael H. (1992). Three approaches to task-based syllabus design. In: TESOL Quarterly 26 (1992), 27– 56.

Long, Michael H. (1996). The role of the linguistic environment in second language acquisition. In: Ritchie, William C. & Tej K. Bhatia (Eds.) (1996). Handbook of second language acquisition. San Diego, CA: Academic Press, 413-468.

Manchón, Rosa M. (2011). Writing to learn the language: Issues in theory and research. In: Manchón, Rosa M. (Ed.) (2011). Learning-to-write and writing-to learn in an additional language. Philadelphia/Amsterdam: John Benjamins, 61-82.

McCutchen, Deborah (1996). A capacity theory of writing: Working memory in composition. In: Educational Psychology Review 8 (1996), 299-324.

Norris, John M. (2009). Task-based teaching and testing. In: Long, Michael H. & Catherine J. Doughty (Eds.) (2009). Handbook of language teaching. Cambridge: Blackwell, 578-594.

Norris, John M. & Lourdes Ortega. (2009). Towards an organic approach to investigating CAF in instructed SLA: The case of complexity. In: Applied Linguistics 30 (2009), 555–578.

Nunan, David (2004). Task-based language teaching. Cambridge: Cambridge University Press.

Ong, Justina & Lawrence J. Zhang (2010). Effects of task complexity on the fluency and lexical complexity in EFL students’ argumentative writing. In: Journal of Second Language Writing 19 (2010), 219–233.

Ortega, Lourdes (2003). Syntactic complexity measures and their relationship to L2 proficiency: A research synthesis of college-level L2 writing. In: Applied Linguistics 4 (2003), 492–518.

Polio, Charlene (1997). Measures of linguistic accuracy in second language writing research. In: Language Learning 47 (1997), 101–143.

Robinson, Peter (1995). Task complexity and second language narrative discourse. In: Language Learning 45 (1995), 99–140.

Robinson, Peter (2001a). Task complexity, cognitive resources, and syllabus design: A triadic framework for examining task influences on SLA. In: Robinson, Peter (Ed.) (2001). Cognition and second language instruction. Cambridge: Cambridge University Press, 287-318.

Robinson, Peter (2001b). Task complexity, task difficult, and task production: Exploring interactions in a componential framework. In: Applied Linguistics 22 (2001), 27–57.

Robinson, Peter (2003). The Cognition Hypothesis of adult, task-based language learning. In: Second Language Studies 21 (2003), 45–107.

Robinson, Peter (2005). Cognitive complexity and task sequencing: Studies in a componential framework for second language task design. In: International Review of Applied Linguistics 43 (2005), 1–32.

Robinson, Peter (2007). Re-thinking-for-speaking and L2 task demands: The Cognition Hypothesis, task classification, and sequencing. Plenary 163-191.address at the Second International Conference on Task-Based Language Teaching, University of Hawai’i.

Roca de Larios, Julio, Rosa M. Manchón & Liz Murphy (2006). Generating text in native and foreign language writing: A temporal analysis of problem-solving formulation processes. In: Modern Language Journal 90 (2006), 100–114.

Ruiz-Funes, Marcela (2014). Task complexity and linguistic performance in advanced college-level foreign language writing. In: Byrnes, Heidi & Rosa M. Manchón (Eds.) (2014). Task-based language learning: Insights from and for writing. Philadelphia/Amsterdam: John Benjamins, 163-191.

Samuda, Virginia & Martin Bygate (2008). Tasks in second language learning. Basingstoke, UK: Palgrave.

Schmidt, Richard (2001). Attention. In Robinson, Peter (Ed.) (2001). Cognition and second language learning (pp. 3–32). Cambridge: Cambridge University Press.

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

Skehan, Peter (1998b). Task-based instruction. In: Annual Review of Applied Linguistics 18 (1998), 268-286.

Skehan, Peter (2001). Tasks and language performance. In: Bygate, Martin, Peter Skehan & Merrill Swain (Eds.) (2001). Researching pedagogic tasks: Second language learning, teaching, and testing. London: Longman, 167-185.

Skehan, Peter (2003). Task-based instruction. In: Language Teaching 36 (2003), 1–14.

Skehan, Peter & Pauline Foster (1997). The influence of planning and post-task activities on accuracy and complexity in task-based learning. In: Language Teaching Research 1 (1997), 185–211.

Skehan, Peter & Pauline Foster (1999). The influence of task structure and processing conditions on narrative retellings. In: Language Learning 49 (1999), 93–120.

Skehan, Peter & Pauline Foster (2001). Cognition and tasks. In: Robinson, Peter (Ed.) (2001). Cognition and second language instruction. New York: Cambridge University Press, 183-205.

Skehan, Peter, Pauline Foster & Uta Mehnert (1998). Assessing and using tasks. In: Renandya, Willy A. & George M. Jacobs (Eds.) (1998). Learners and language learning. Singapore: SEAMEO Regional Language Centre, 227-248.

Spelman Miller, Krityan (2006). The Pausological study of written language production. In: Sullivan Kirk P. H. & Eva Lindgren (Eds.) (2006). Computer keystroke logging and writing. Amsterdam, The Netherlands: Elsevier, 11-39.

Swain, Merrill (1985). Communicative competence: Some roles for comprehensible input and comprehensible output in its development. In: Gass Susan M. & Carolyn Madden (Eds.) (1985). Input in second language acquisition. Rowley, MA: Newbury House, 235-256.

Swain, Merill (1995). Three functions of output in second language learning. In: Cook, Guy & Barbara Seidlhofer (Eds.) (1995). Principles and practice in applied linguistics. Oxford: Oxford University Press, 124-144.

Swain, Merrill (2005). The output hypothesis: Theory and research. In: Hinkel, Eli (Ed.) (2005). Handbook of research in second language teaching and learning. Mahwah, NJ: Lawrence Erlbaum, 471-481.

Swain, Merrill & Sharon Lapkin (1995). Problems in output and the cognitive processes they generate: A step toward second language learning. In: Applied Linguistics 16 (1995), 371–391.

Swales, John (1986). Utilizing the literatures in teaching the research paper. Unpublished manuscript.

Van den Branden, Kris, Martin Bygate, & John M. Norris (2009). Task-based language teaching: A reader. Philadelphia/Amsterdam: John Benjamins.

Wickens, Christopher D. (1992). Engineering psychology and human performance. New York: Harper Collins.

Widdowson, Henry G. (1998). Skills, abilities, and context of reality. In: Annual Review of Applied Linguistics 18 (1998), 323–333.

Wolfe-Quintero, Kate, Shunji Inagaki & Hae-Young Kim (1998). Second language development in writing: Measures of fluency, accuracy, and complexity. Honolulu, HI: University of Hawai’i, Second Language Teaching and Curriculum Center.

Yuan, Fangyuan & Rod Ellis (2003). The effects of pre-task planning and on-line planning on fluency, complexity and accuracy in L2 monologic oral production. In: Applied Linguistics 24 (2003), 1−27.

Yule, George (1997). Referential communication tasks. Mahwah, NJ: Lawrence Erlbaum.

Zamel, Vivian (1983). The composing processes of advanced ESL students: Six case studies. In: TESOL Quarterly 17 (1983) 165–187.


Appendix

Writing Tasks

You are planning to study abroad next year in a Spanish-speaking country of your choice and will apply through the International Student Exchange Programs (ISEP). As part of the application process, you are asked to submit two essays written in Spanish to share with the selection committee information about your life and readiness to study abroad. These essays will also help the selection committee determine your proficiency level in Spanish. You will write the essays under the supervision of a proctor at your home institution. No dictionary or any other resources will be allowed. You will have 50 minutes to write each essay. Each one should be of at least 250 words in length.

TASK 1 - Essay One: [personal essay]

It is important for the selection committee to get to know something about you. For this essay, write about yourself, your interests, and your expected goals for studying abroad. Include a personal life story that you think has prepared you for this experience and that will contribute to making your stay abroad a positive one.

TASK 2 - Essay Two: [expository essay]

Over the years, the benefits and challenges of studying abroad have been well documented by academics in the field. For this essay, present the benefits and challenges of studying abroad for college students.

Author:

Marcela Ruiz-Funes, PhD

Assistant Professor

Department of Foreign Languages

Georgia Southern University

Statesboro

Ga 30458

E-mail: mruizfunes@georgiasouthern.edu