Available online at www.jlls.

org
JOURNAL OF LANGUAGE
AND LINGUISTIC STUDIES
ISSN: 1305-578X
Journal of Language and Linguistic Studies, 18(1), 361-371; 2022

Syntactic complexity in a learner written corpus and L2 speaking quality:

Suggestions for distinguishing L2 speaking proficiency

Shinjae Park a 1
a
Incheon National University, South Korea

APA Citation:

Park, S. (2022). Syntactic complexity in a learner written corpus and L2 speaking quality: Suggestions for distinguishing L2 speaking
proficiency. Journal of Language and Linguistic Studies, 18(1), 361-371. Doi: 10.52462/jlls.187
Submission Date:20/07/2021
Acceptance Date:04/10/2021

Abstract
Based on the quantitative analysis of L2 English texts from Korean undergraduates, the present paper
demonstrates the possibility of distinguishing L2 speaking proficiency with differing measures of syntactic
complexity in English writing. To this end, 14 measures of complexity were gauged using an L2 Syntactic
Analyzer (Lu, 2010) in 89 EFL essays from a learner corpus to find out the best indicators of L2 speaking
proficiency. As a result of the normality assumption test, the logistic regression analysis was applied and the
model was significant in classification according to L2 proficiency and the classification was correct in 62.9% of
the studied cases. Furthermore, it showed that the best predictor of L2 speaking proficiency was the complex
nominal per T-unit (CNT). Specifically, upon the acquisition of CNT, EFL learners are 6.4 times more likely to
achieve higher proficiency in speaking English. Taken together, this paper suggests that EFL students’ quality of
writing can be effectively used to distinguish their speaking ability.
Keywords: learner written corpus; syntactic complexity; EFL; logistic regression; assessment

1. Introduction

Widely acknowledged as a construct to effectively evaluate L2 writing development (Ai & Lu,
2013: 249–64; Lu, 2011:36–62; Martínez, 2018), linguistic complexity has been extensively used in
previous research to quantitatively analyze meaningful linguistic characteristics of L2 writing (Ai &
Lu, 2013: 249–64; Kim, 2014; Lu, 2011: 36–62; Martínez, 2018). Linguistic complexity embraces
syntactic and lexical complexity, and both constructs are used to evaluate L1 and L2 students’
linguistic performance, proficiency, and development (Kim, 2014; Kuiken, Vedder & Gilabert, 2010;
Szmrecsanyi & Kortmann, 2012). With regard to syntactic complexity, which is the focus of this
study, measures of syntactic complexity were used to assess L2 writing to investigate the functions of
task complexity, writing ability, and teaching methods in L2 teaching of students of different ages,
proficiency levels, and developmental periods (Ahmadian, 2012; Lu, 2011; Wu & Ortega, 2013).
In this study, syntactic complexity is defined as “the range and the sophistication of grammatical
resources exhibited in language production” (Ortega, 2015: 82). Following Bulté and Housen (2014),

1
Corresponding author.
E-mail address: tlswo@naver.com.

© 2022 Cognizance Research Associates - Published by JLLS.

362 Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371

this study hypothesizes “the more the components a feature or system comprises and the denser the
relationships between its components, the more complex is the feature or system” (2014: 46). As
argued by Ortega (2015: 82), “syntactic complexity indices the expansion of the capacity to use the
additional language in ever more mature and skilful ways, tapping the full range of linguistic resources
offered by the given grammar in order to fulfil various communicative goals successfully.” Until
recently, in the realm of syntactic complexity, L2 writing researchers are in agreement that the stronger
L2 writers are, the longer and more complex syntactic structures they produce (Lu, 2011; Lu & Ai,
2015; Ortega, 2003; Wolfe-Quintero et al., 1998).
Recent studies (Biber, Gray & Poonpon, 2011; Biber, Gray & Staples, 2016; Hwang, Jung, & Kim,
2020) comparing speaking and writing in terms of syntactic complexity make the assertion that the two
modes are totally separate, but, questions may be raised about whether there is any commonality
between the two modes, and whether there is any method to support assessment of L2 performance in
speaking and writing by analyzing them in a convergent manner. To this ends, this study examines that
EFL students’ syntactic complexity in their L2 writing can be effectively used to predict their speech
proficiency. Therefore, the aim of this study is to suggest an alternative assessment method to
distinguish L2 speaking quality through writing ability by measuring syntactic complexity indices in
L2 writing and making use of holistic speaking proficiency of Korean L2 learners. Accordingly, when
no oral data are available, researchers can use EFL students’ writings to predict their speaking quality.
Finding evidence in support of this prediction would suggest an alternative equation, based on
quantitative measures gauging linguistic complexity by using L2 Syntactic Complexity Analyzer
(L2SCA: Lu, 2010; 2011) and corresponding holistic ratings of speaking quality, to help educators
adequately determine learners’ proficiency in L2 speaking.

2. Literature Review

Among other linguistic parameters studied in L2 research into written output, syntactic complexity
has been extensively investigated. However, using syntactic complexity as a means to measure writing
proficiency and assess of language complexity (e.g., Lu, 2010, 2011; Martínez, 2018; Taguchi,
Crawford, & Wetzel, 2013: 420–30; Yang, Lu, & Weigle, 2015: 53–67), most studies have relied on
quantifiable complexity indices such as length of production unit, sentence complexity, and frequency
certain sentence structures. Of these, T-unit (Hunt, 1965) is an important concept defined as the
shortest grammatical chunk of a sentence as a unit of analysis. In other words, it is defined as the
"shortest grammatically allowable sentences into which (writing can be split) or minimally terminable
unit." More technically, a T-unit is a dominant clause and its dependent clauses: as Hunt said, it is "one
main clause with all subordinate clauses attached to it" (Hunt 1965: 20). A T-unit is often used in the
analysis of written and spoken discourse, such as in studies on errors in second language writing. In a
review of 39 papers on L2 writing that investigated various indices of complexity, fluency, and
accuracy, Wolfe-Quintero et al. (1998) found that T-unit length and clause length can be used as a
measure of writing fluency. Other previously established indicators for syntactic complexity include
the mean length of sentence (MLS) and mean number of T-units per sentence (Ortega, 2003).
In addition, traditional studies on writing have mainly focused on the use of clausal subordination.
For instance, Flahive and Snow (1980) reported a strong relationship between clausal subordination
(evaluated by the number of dependent clauses in each essay) and writing quality. Likewise, in a study
on the relationship between a holistic evaluation of ESL writing proficiency and measures of
complexity, Homburg (1984) found a strong link between finite clausal subordination and quality of
writing; similarly, there was significant relationship between complexity as measured by length
measures (namely, MLS and MLC) and holistic grading of composition quality.

© 2022 Cognizance Research Associates - Published by JLLS.

. Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371 363

Furthermore, there is substantial evidence documenting that more proficient L2 writers produce
longer texts (e.g., Crossley & McNamara, 2012; Espada-Gustilo, 2011; Ferris, 1994; Lu & Ai, 2015;
Martínez, 2018; McNamara et al., 2010). Indeed, longer essays tend to be rated higher than their
shorter counterparts. For example, Ferris (1994) demonstrated that writing proficiency can be
reasonably well predicted by the number of words and the length of production. Similar findings were
also reported by Espada-Gustilo (2011) and Martínez (2018). For instance, based on the holistic
ratings of the essays from secondary school-aged writers having various levels of proficiency,
Martínez (2018) revealed a substantial relationship between syntactic complexity and writing
proficiency. Specifically, she reported that the use of longer units was a strong indicator of high-
quality output; in contrast, the common usage of simple utterences was found to be associated with
lower writing quality. In line with these results, proficient L2 writers were found to typically create
longer texts with more versatile diverse vocabulary and a higher number of words per sentence (Kim,
2014; McNamara et al., 2010).
However, there is some evidence that more proficient L2 writers do not necessarily use more
clauses. In this respect, as argued by Rimmer (2006), important features of syntactic complexity are
also phrasal features such as noun post-modifiers. Indeed, in a study that used Syntactic Analyzer to
analyze college-level Chinese learners’ essays written in the students’ L2, Lu (2011: 36–62) found that
proficiency was predicted by some of these measures, including mean length of a clause, complex
nominals per clause, complex nominals per T-unit, MLT, and MLS. Furthermore, in a study that
compared 600 English compositions by Chinese writers with the Louvain Corpus of Native English
Essays, Ai and Lu (2013) found that the two groups differed in terms of the MLS, MLC, MLT, and
complex nominals. More specifically, in an analysis of argumentative written samples by EFL
learners, Taguchi et al. (2013) found that writing quality was largely positively influenced by noun
phrase modification (i.e., preceding attributive adjective and prepositional phrase as post-modifiers of
nouns).
In a more recent study, Lan and Sun (2019) made a comparison the argumentative essays of
Chinese English learners to academic journal articles and discovered that frequent noun modifiers
appeared in students' writings much less than in academic journal articles, and the higher-proficiency
students tended to use more modifiers in writing. Similarly, Khushik and Huhta (2020) analyzed
English argumentative essays by teenagers from Pakistan and Finland, and the length of production
units, phrasal density, and subordination differed according to proficiency level. Casal and Lee (2019)
also examined the association among syntactic complexity and writing quality in assessed research
papers produced by 280 ESL undergraduates, and revealed that mean length of T-units and phrasal
measures differed across levels.
Of note, however, Halliday (2006) argued that the complexity of speech should be focused on
different aspects from writing because the underlying linguistic characteristics of them are different. In
this regard, through the attempts of several influential studies, while major grammatical complexities
of speech were found to be clausal subordination features, those of academic writing were found to be
phrasal embedding (Biber et al., 2011; Biber et al., 2016; Kyle & Crossely, 2018). Moreover,
considering that conversation is acquired first, clausal subordination features were not reported to be
associated with a high degree of production complexity (Biber et al., 2011). Instead, it is argued that,
since a large amount of complex phrasal embedding was produced and limited to more specialized
circumstances of formal forms of writing, these structures represent a extensively higher degree of
complexity (Ibid.). Recently, Hwang et al. (2020) examined syntactic complexity in spoken and
written data in English produced by 122 Korean children. The results revealed that the written data
involved more subordination (DCT), longer structures (MLT), and more verb phrases (VPT) than the
spoken data, while the spoken data included more coordination (CPT) than the written data.

© 2022 Cognizance Research Associates - Published by JLLS.

364 Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371

On the other hand, studies are also found that may imply that the underlying mechanisms are
shared. Cleland and Pickering (2006) discovered that a group of UK students tended to repeat syntactic
form across the two modes to the same extent that they did within either mode. Kormos (2014) found
that there was no difference in the percentage of subordinating clauses between the mods of Hungarian
English learners, while the writing contained more modifiers per noun phrase. Therefore, if the sharing
between the syntax complexity represented in one mode and the specific characteristics of the other
mode is found, and further studies that have not been conducted before are attempted, linguistics may
be expanded to new and practical areas.
As is clear from the review presented above, most previous studies have investigated syntactic
complexity indices as predictors of language development or proficiency in either the writing mode or
the speaking mode. However, since speaking materials are acknowledged as unwieldy and messy
(Rajadurai, 2007), there have been a number of studies on data of L2 writers. Therefore, few attempts
have been made to characterize syntactic complexity in both speaking and writing, especially studies
that have explored the complexity in cross-mode methods such as applying syntactic characteristics of
one mode to predict proficiency of other modes.
To fill these gaps, I examine whether 14 complexity measures in EFL writing samples can be used
to effectively predict English speaking proficiency. From the applied perspective, considering that this
study is the first to empirically test the convergence across modes (writing and speaking), these results
can provide valuable insights for practitioners in the EFL field. This study addresses the following
research questions:
1. Which syntactic features significantly predict L2 speaking proficiency levels?
2. Is it possible to derive a significant logistic regression equation to distinguish L2 speaking
proficiency?
3. If so, to what extent can the derived equation predict the classification of the observed data?

3. Experimental/Materials and Methods

3.1. Corpus selection

For the analysis, I used the Incheon National University Multi-Language Learner Corpus (INU-
MULC, see Park, 2020a; Park & Yoon, 2021) to select 89 essays written by college-level English
learners. The MULC is a corpus of data written and spoken by undergraduate students at a university
in Korea and is currently being compiled. Thus far, the MULC contains 109 conversations, 224
monologues, and 139 essays. To write the essays, the participants were asked to select one among five
topics, and the essays were later electronically stored in the MULC database. In the corpus, each
participant’s speaking proficiency was registered based on the rubrics of Common European
Framework of Reference for Languages (CEFR). The ratings were given by three native English
speakers who were specifically trained for that purpose: They are currently lecturers at the university
that has been compiling this corpus, and continuously have helped for this project. They rated 5% of
the speaking corpus in the pilot test, and after high consistency between their assessments was
achieved, they actually began to conduct the assessment. In this study, following the criteria articulated
by Crossley and McNamara (2013), only the texts that exceeded the 100-word cut-off in length were
selected for further analysis. Essays shorter than 100 words were deemed unreliable as they do not
contain a sufficient number of features owing to their short length.
Although the selection of essays from the corpus was random, I tried to ensure that the number of
corresponding items remained as consistent as possible. The writers of the essays included in the
sample were 54 females and 35 males (see Table 1). Based on proficiency levels, the essays were

© 2022 Cognizance Research Associates - Published by JLLS.

. Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371 365

grouped as follows: (1) low proficiency group (LP; N = 47; 52.8%) and (2) high proficiency group
(HP; N = 42; 47.2%). In the sample of selected essays, 25 participants (28.1%) wrote essays on
marriage, 22 (24.7%) on review of a book read recently, 17 (19.1%) on mobile phones, 16 (18%) on
uniforms, and 9 (10.1%) on clubs. The topics of the essays were as follows:
1) Should everyone get married?
2) Is it essential to wear uniforms in middle and high schools?
3) Should elementary, middle, and high school students be allowed to carry mobile phones in
class?
4) Should any college student join a club?
5) Write a brief review of the book you have read recently
Table 1. Participants’ characteristics

Gender Topic Proficiency

Mobile
Male Female Marriage Literature Uniform Club Low High
Phone
35 54 25 22 17 16 9 47 42
39.3% 60.7% 28.1% 24.7% 19.1% 18.0% 10.1% 52.8% 47.2%

Total: 89 (100%)

3.2. Computational tools and variables

To measure linguistic indices used for in this study, L2SCA was used. Further detail on each index
can be found in Lu (2010; 2011). The five major categories were as follows: (1) length of production
units, (2) overall sentence complexity, (3) number of subordination clauses, (4) number of
coordination clauses, and (5) phrasal sophistication. The analysis focused on 14 indices in the
following five categories: (1) length of production (N = 3), (2) sentence complexity (N = 1), (3)
number of subordination clauses (N = 4), (4) number of coordination clauses (N = 3), and (5) phrasal
sophistication (N = 3). The specific linguistic features and indices gauged by the computational tool
(L2SCA) are summarized in Table 2.
Table 2. The 14 syntactic automated complexity measures (Lu, 2010, 2011)

Length of production unit Mean length of clause (MLC: words/clause)

Mean length of sentence (MLS: words/sentence)
Mean length of T-unit (MLT: words/T-unit)
Sentence complexity Clauses per sentence (CS: clauses/sentence)
Subordination Clauses per T-unit (CT: clauses/T-unit)
Complex T-unit per T-unit (CTT: complex T-units/T-unit)
Dependent clauses per Clause (DCC: dependent clauses/clause)
Dependent clauses per T-unit (DCT: dependent clauses/T-unit)
Coordination Coordinate phrases per clause (CPC: coordinate phrases/clause)
Coordinate phrases per T-unit (CPT: coordinate phrases/T-unit)
T-units per sentence (TS: T-units/sentence)
Particular structures Complex nominals per clause (CNC: complex nominals/clause)

© 2022 Cognizance Research Associates - Published by JLLS.

366 Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371

Complex nominals per T-unit (CNT: complex nominals/T-unit)

Verb phrases per T-unit (VPT: verb phrases/T-unit)

4. Results

The statistical methods to determine best fitting model to describe the relationship between the
outcome (i.e., proficiency in English speech) and a set of independent variables (i.e., complexity
measures) include discriminant analysis and logistic regression analysis. To determine whether the
normality assumption is satisfied, the Box's M test was conducted (Table 3). The significance value
(p=.000) indicates that the data used in this study differ across factors in terms of the multivariate
normal distribution (Table 3).
Table 3. Box’s M test of equality of covariance matrices

Box’s M 181,949
Approx. 4.557
df1 36
F
df2 24,795.25
Sig. .000

With non-normality as in this case, the logistic regression method is applied to analyse categorical
outcome variables (Gong & Sun, 2009: 1366-71; Park, 2020b; Sio & Ismail, 2019: 29). Dependent
variables of L2 speaking proficiency were divided into LP and HP for the binary logistic regression.
Next, stepwise regression analysis was conducted to find indices that predict L2 speaking quality, and
as a result, the significant independent variable of the model was revealed to be complex nominals per
T-unit (CNT, p < .05) (Table 4).

Table 4. Summary of Stepwise Multiple Regression Analysis

Unstandardized Coefficients Standardized Coefficients

Model B S.E Beta t Sig.
(Constant) .947 .157 6.036 .000
CNT .393 .111 .354 3.528 .001

The results of the overall test of the model using the likelihood ratio are presented in Table 5. The
model was statistically significant, χ2(1) = 12.033 (p < .05); therefore, the model not only identified
the sub-scales that influenced L2 proficiency but also adequately fit the data.
Table 5. Omnibus test of model coefficients

Chi-square df Sig.
Step 12.033 1 .001
Step 1 Block 12.033 1 .001
Model 12.033 1 .001

The two methods used in this study to calculate the explained variation are presented in Table 6.
According to Nagelkerke R2 values (Nagelkerke R2 = .169), the explained variation in the dependent
variables in this model was 16.9%.

© 2022 Cognizance Research Associates - Published by JLLS.

. Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371 367

Table 6. Summary of Stepwise Multiple Regression Analysis

Step -2 Log likelihood Cox and Snell R-square Nagelkerke R-square

1 111.066 .126 .169

The assessment results of the effectiveness of the predicted classification compared with the actual
classification, i.e., the observed number of students in each of the two proficiency groups and the
predicted number according to logistic regression (p < .05) are reported in Table 7. As is seen in Table
7, about two-thirds (72.3%) of all cases in the LP group and more than half (52.4%) of all cases in the
HP group were correctly classified. The overall classification accuracy amounted to 62.9%.
Table 7. Category prediction

Observed Predicted
L2 Proficiency Correct
Low High %
Low 34 13 72.3
Step 1 High 20 22 52.4
Overall Percentage 62.9

Table 8 summarizes the results of the logistic regression conducted to assess the influence of
predictors on L2 proficiency in the data. First, with regard to the odds ratio (i.e., Exp(B)), CNT, with
an odds ratio of 6.386, is the strongest predictor of L2 proficiency. This finding proposes that, upon
acquisition of CNT, EFL learners were more than six times more likely to achieve HP. Second, I
computed the coefficients and the Wals values of CNT, which were used to evaluate the statistical
significance of each independent variable (Wals: 9.598, p < .05). Finally, based on the results, I
derived the following estimated logistic regression equation (See Eq.1) for English proficiency levels.
Eq.1: log (speaking proficiency level) = -2.578 + 1.854(CNT)
Employing this equation, the boundary of 0.5 was used to split the participants into two groups (LP
and HP). Specifically, when the values of learners’ CNT were entered into the derived logistic
regression equation, the participants whose values were equal to or below 0.5 or less were assigned to
the LP group, while those scoring higher than 0.5 were added to the HP group.
Table 8. Variables in the equation

B S.E, Wals df Sig. Exp(B) Low High

CNT .1.854 .598 9.598 1 .002 6.386 1.976 20.638
Step 1
Constant –2.578 .821 9.860 1 .002 .076

5. Discussion and Conclusions

In this study, I used a corpus-based approach and adjusted design to investigate Korean L2
learners’ writing characteristics. To this end, I systematically examined syntactic complexity features

© 2022 Cognizance Research Associates - Published by JLLS.

368 Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371

of EFL learners’ essays in the following major areas: (1) length of production units, (2) overall
sentence complexity, (3) number of subordination clauses, (4) number of coordination clauses, and (5)
phrasal sophistication. Based on the results, I identified the linguistic features in EFL writing that best
predict L2 learners’ speaking proficiency levels.
In previous research, syntactic complexity features in L2 writing have been widely used to improve
writing quality and evaluate writers’ proficiency levels. Unlike the aforementioned research, this study
contributes to expanding the research scope, proposing a complementary method, i. e., a crossover
research method between modes, to distinguish L2 speaking levels by measuring syntactic complexity
in writing. The results of the stepwise multiple regression analysis performed in this study suggest that
the best determinant is complex nominals per T-unit (CNT) (i. e., category: phrasal sophistication).
This study is meaningful in that it revealed that this phrasal complexity, which is closely related to the
writing, is also important for predicting speaking proficiency. These results are in line with several
previous studies that reported a positive relationship between the use of complex nominals and L2
proficiency (Biber et al., 2011; Biber et al., 2016, Taguchi et al., 2013; Kyle & Crossley, 2018).
Furthermore, this study is important in that it demonstrates that the phrasal complexity, such as
complex nominal, is an important measure in speaking as well as writing through the results that the
complex nominal per T-unit in writing is the best predictor of L2 speaking. It is also suggested that
these issues need to be addressed more in future studies. In addition, the logistic regression model
explained 16.9% of the total variance in classification and was statistically significant, χ2(1) = 12.033,
p < .05. The accuracy rate of the classification amounted to 62.9%. It also suggests that, upon
acquisition of CNT, learners are 6.4 times more likely to achieve higher proficiency in speaking
English. In other words, the model empirically demonstrates that educators can effectively classify L2
learners according to the proficiency of L2 speaking when only measuring CNT among multiple
indices in the writings of the learners in the EFL context.

6. Implications

The results of this study provide interesting implications for L2 researchers and educators. First, the
present results suggest that measuring complexity in written data can reliably predict students’
proficiency in English speaking quality. This finding is important because, unlike speech data that
requires a complex process of collection and transcription, written data are relatively easy to obtain
and analyze (Park, 2020a; Park & Yoon, 2021). In other words, it is valuable in contexts when no
official English test scores are available, or when, owing to limited time, place, or insufficient budget,
it is impossible to conduct speaking tasks. Second, EFL instructors ensure that, through explicit
vocabulary instruction, Korean L2 learners acquire a rich and diverse vocabulary. According to the
present results, the learners who improved their CNT abilities from phrasal sophistication were over
6.4 times more likely to achieve a higher proficiency in speaking English. This finding suggests that,
to enhance L2 learner’s nominal and phrasal sophistication knowledge, relevant pedagogical
interventions would be needed. Additionally, EFL instructors should prioritize the development of
phrasal sophistication in their teaching. Finally, making use of the equation proposed in this study,
educators can measure only one dimension to effectively classify EFL learners according to levels of
proficiency in L2 speaking. The equation is valid since I attained more wide-ranging results than could
have been afforded by analyzing a fewer number of syntactic features.
Given the scope of the present study and the accessibility of the corpus used, there are several
limitations. First, in terms of the scope of the study, further research is needed to explore whether the
results reported herein show the same pattern as those of non-Korean EFL learners or learners with
different academic backgrounds. Likewise, it is worth investigating whether the proposed approach
would work with other topics, writing types, context-related factors, or test formats, i.e., speaking

© 2022 Cognizance Research Associates - Published by JLLS.

. Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371 369

tests. Second, the MULC was designed to become publicly accessible inside and outside school
environments; however, it is currently in the stage of POS tagging and continues to expand, thereby
delaying its release. In addition, a more user-friendly data arrangement is still required for the corpus’
widespread use. Nevertheless, despite the limitations outlined above, this study provides valuable
insights into the effects of syntactic complexity measures on L2 speaking proficiency. By conducting a
thorough analysis of syntactic features in EFL essays and proposing a complementary method to
evaluate L2 speaking proficiency, this study makes a meaningful contribution to the field of L2
speaking and writing research.

Funding

This work has been supported by the National Research Foundation of Korea (NRF) grant funded
by the Korea government (No. 2020S1A5B5A17090102).

References
Ahmadian, M. J. (2012). The effects of careful guided online planning on complexity, accuracy and
fluency in intermediate EFL learners’ oral production: The case of English articles. Language
Teaching Research, 16(1), 129-149.
Ai, H., & Lu, X. (2013). A corpus-based comparison of syntactic complexity in NNS and NS
university students’ writing. Automatic treatment and analysis of learner corpus data, 249-264.
Biber, D., Gray, B., & Poonpon, K. (2011). Should we use characteristics of conversation to measure
grammatical complexity in L2 writing development?. TESOL Quarterly, 45(1), 5-35.
Biber, D., Gray, B., & Staples, S. (2016). Predicting patterns of grammatical complexity across
language exam task types and proficiency levels. Applied Linguistics, 37(5), 639-668.
Bulté, B., & Housen, A. (2014). Conceptualizing and measuring short-term changes in L2 writing
complexity. Journal of second language writing, 26, 42-65.
Casal, J. E., & Lee, J. J. (2019). Syntactic complexity and writing quality in assessed first-year L2
writing. Journal of Second Language Writing, 44, 51-62.
Cleland, A. A., & Pickering, M. J. (2006). Do writing and speaking employ the same syntactic
representations?. Journal of Memory and Language, 54(2), 185-198.
Crossley, S. A., & McNamara, D. S. (2012). Predicting second language writing proficiency: The roles
of cohesion and linguistic sophistication. Journal of Research in Reading, 35(2), 115-135.
Crossley, S., & McNamara, D. (2013). Applications of text analysis tools for spoken response grading.
Language Learning & Technology, 17(2), 171-192.
Espada-Gustilo, L. (2011). Linguistic features that impact essay scores: A corpus linguistic analysis of
ESL writing in three proficiency levels. 3L: Language, Linguistics, Literature, 17(1), 55-64.
Ferris, D. R. (1994). Lexical and syntactic features of ESL writing by students at different levels of L2
proficiency. TESOL Quarterly, 28(2), 414-420.
Flahive, D. E., & Snow, B. G. (1980). Measures of syntactic complexity in evaluating ESL
compositions. Research in language testing, 171-176.

© 2022 Cognizance Research Associates - Published by JLLS.

370 Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371

Gong, J., & Sun, S. (2009, June). A new approach of stock price prediction based on logistic
regression model. In 2009 International Conference on New Trends in Information and Service
Science (pp. 1366-1371). IEEE.
Halliday, A. K., & Webster, J. J. (2006). Language of Science (Vol. Volume 5). London: Bloomsbury
Academic.
Homburg, T. J. (1984). Holistic evaluation of ESL compositions: Can it be validated objectively?.
TESOL quarterly, 18(1), 87-107.
Hunt, K. W. (1966). Recent measures in syntactic development. Elementary English, 43(7), 732-739.
Hwang, H., Jung, H., & Kim, H. (2020). Effects of Written Versus Spoken Production Modalities on
Syntactic Complexity Measures in Beginning‐Level Child EFL Learners. The Modern Language
Journal, 104(1), 267-283.
Khushik, G. A., & Huhta, A. (2020). Investigating Syntactic Complexity in EFL Learners' Writing
across Common European Framework of Reference Levels A1, A2, and B1. Applied Linguistics,
41(4), 506-532.
Kim, J. Y. (2014). Predicting L2 Writing Proficiency Using Linguistic Complexity Measures: A
Corpus-Based Study. English Teaching, 69(4), 27-51.
Kormos, J. (2014). Differences across modalities of performance. In H. Byrnes and R. M. Manchón
(Eds), Task-based language learning: Insights from and for L2 writing (pp. 193-216). Amsterdam:
John Benjamins.
Kuiken, F., Vedder, I., & Gilabert, R. (2010). Communicative adequacy and linguistic complexity in
L2 writing. Communicative proficiency and linguistic development: Intersections between SLA and
language testing research, 1, 81-100.
Kyle, K., & Crossley, S. A. (2018). Measuring syntactic complexity in L2 writing using fine‐grained
clausal and phrasal indices. The Modern Language Journal, 102(2), 333-349.
Lan, G., & Sun, Y. (2019). A corpus-based investigation of noun phrase complexity in the L2 writings
of a first-year composition course. Journal of English for Academic Purposes, 38, 14-24.
Lu, X. (2010). Automatic analysis of syntactic complexity in second language writing. International
journal of corpus linguistics, 15(4), 474-496.
Lu, X. (2011). A corpus‐based evaluation of syntactic complexity measures as indices of college‐level
ESL writers’ language development. TESOL Quarterly, 45, 36-62.
Lu, X., & Ai, H. (2015). Syntactic complexity in college-level English writing: Differences among
writers with diverse L1 backgrounds. Journal of Second Language Writing, 29, 16-27.
Martínez, A. C. L. (2018). Analysis of syntactic complexity in secondary education EFL writers at
different proficiency levels. Assessing Writing, 35, 1-11.
McNamara, D. S., Crossley, S. A., & McCarthy, P. M. (2010). Linguistic features of writing quality.
Written Communication, 27, 57-86.
Ortega, L. (2003). Syntactic complexity measures and their relationship to L2 proficiency: A research
synthesis of college‐level L2 writing. Applied Linguistics, 24, 492-518.
Ortega, L. (2015). Syntactic complexity in L2 writing: Progress and expansion. Journal of Second
Language Writing, 29, 82-94.

© 2022 Cognizance Research Associates - Published by JLLS.

. Park / Journal of Language and Linguistic Studies, 18(1) (2022) 361–371 371

Park, S. (2020a). A Corpus Study of Modal Verbs in Korean Learners' Speech. Journal of Linguistic
Studies, 25(2), 121-137.
Park, S. (2020b). Determining L2 Proficiency in the Production and Perception of Consonant Clusters.
3L: Language, Linguistics, Literature®, 26(3), 41-52.
Park, S. & Yoon, S. (2021). Syntactic Complexity of EFL Learners’ Casual Conversation, Monologue,
and Writing. The Journal of Studies in Language, 37, 75-89
Rajadurai, J. (2007). Capturing L2 Spoken English: Methodological Challenges. Linguistics Journal,
2(3), 41-58.
Rimmer, W. (2006). Measuring grammatical complexity: The Gordian knot. Language Testing, 23,
497-519.
Sio, J. & Ismail, R. (2019). Binary logistic regression analysis of instructional leadership factors
affecting English language literacy in primary schools. 3L: Language, Linguistics, Literature®,
25(2), 25-37.
Szmrecsanyi, B., & Kortmann, B. (2012). Introduction: Linguistic complexity. In Linguistic
Complexity (pp. 6-34). de Gruyter.
Taguchi, N., Crawford, W., & Wetzel, D. Z. (2013). What linguistic features are indicative of writing
quality? A case of argumentative essays in a college composition program. TESOL Quarterly, 47,
420-430.
Wolfe-Quintero, K., Inagaki, S., & Kim, H. Y. (1998). Second Language Development in Writing:
Measures of Fluency, Accuracy, and Complexity (No. 17). University of Hawaii Press, Hawaii.
Wu, S. & Ortega, L. (2013). Measuring global oral proficiency in SLA research: A new elicited
imitation test of L2 Chinese. Foreign Language Annals, 46, 680-704.
Yang, W., Lu, X., & Weigle, S. C. (2015). Different topics, different discourse: Relationships among
writing topic, measures of syntactic complexity, and judgments of writing quality. Journal of
Second Language Writing, 28, 53-67.

AUTHOR BIODATA

Dr. Shinjae Park is a research professor at Incheon National University, South Korea. Her research interests are
corpus studies, phonetics, phonology, and language teaching and assessment.

© 2022 Cognizance Research Associates - Published by JLLS.