O n f e r e n c e

Short paper 
Vocabulary knowledge plays an important role in language proficiency (Nation, 1993). A 
large amount of research has been focusing on exploring more effective approaches to 
deepening and widening language learners’ vocabulary knowledge.
Paraphrasing, varying the expressions to restate the same idea, is a vital skill for 
language learners to be successful in academic settings (Chen et al., 2013). Sufficient 
vocabulary knowledge contributes to a good command of such language production. 
Studies suggest that grouping synonymous phrases (or formulaic expressions) sharing 
similar meaning (Curtis, M.E., 2006; Shirazi & Yamini, 2011) but various usages not only 
reinforces vocabulary learning but helps learners raise awareness of usage differences 
(Gairns & Redman, 1986; Seal, 1991; Mashhady et al, 2011). In other words, learning 
phrasal paraphrases or synonymous formulaic expressions could activate the 
understanding of language use, which is important for fluent paraphrasing. 
While studies on grouping and learning synonymous formulaic expressions have been still 
limited, the utilization of clustering semantically equvalent words or phrases, one of 
various vocabulary-teaching techniques (Hashemi & Gowdasiaei, 2005), is quite common 
in language learning environment. For example, learners would be encouraged to vary 
their expressions to restate given ideas in their writings whereas instructors would 
suggest synonyms or synonymous formulaic expressions during the writing revision 
process. Moreover, lists of synonymous words or phrases (Hashemi & Gowdasiaei, 2005) 
can be seen in instructional materials; or even in classroom settings, teachers would 
suggest synonymous words or phrases to help learners expand their vocabulary 
knowledge. Despite the practicality of grouping formulaic expressions in meaning in 
language pedagogy, they appear scattered or unavailable throughout instructional 
materials or dictionaries. Learners are required to consult various reference tools to 
collect synonymous formulaic expressions. The time-consuming search task and 
insufficient information seem unable to be conducive to effective vocabulary expansion 
(Currie, 1998; Deckert, 1993).
To address such an issue, effective learning support is a pressing need. To our best 
knowledge, few computer-assisted aids have been developed to facilitate the acquisition 
of synonymous formulaic expressions in the research area of computer assisted language 
learning (CALL). Thus, we promote computer-assisted synonymous formulaic expressions 
learning. In this study, we adopted an automatic paraphrase suggestion system, PREFER 
(Chen et al., 2012) to equip learners with grouped synonymous formulaic expressions 
and the corresponding usages. We investigated whether and the extent of the tool 
effectiveness on learners’ acquisition of synonymous formulaic expressions. 
To evaluate the effectiveness of the paraphrasing system, PREFER on expanding EFL 
learners’ vocabulary knowledge, we designed pre
-test and post-test assessment to 
compare 49 students’ performance on identifying the synonymous formulaic expressions 
and differentiating the individual usages. We also surveyed the students’ attitudes 
towards PREFER using an evaluation questionnaire.

---

-87- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
A worksheet and a test sheet were developed for exploring participant’s improvement 
due to PREFER. Students were randomly divided into two groups.
One group (P) consulted PREFER, and the control group (C) were allowed to consult any 
existing reference tools they have been familiar with, such as Longman English Dictionary 
Online, Yahoo Dictionary, Dr. eye or Google Translate. The students were required to find 
and familiarize themselves with as more synonymous formulaic expressions and the 
corresponding usages of the 20 given phrases on the worksheet as they could between 
the pre- and post-test. The test sheet was comprised of 15 multiple selection questions.
To examine effectiveness, the scores of the two students groups were analyzed and 
compared. As shown in the upper panel of Table 1, all students clearly achieved gains 
with the help of reference tools. However, Group P achieved better performance (an 
increase of 7.6 points), compared with Group E (an increase of 3.4 point). The results of 
ANOVA showed significant differences between Group P’s and Group C’s improvement 
(F(1, 46)=6.21, p value=0.016). In other words, PREFER provided relatively substantial 
benefit for students in facilitating formulaic expressions use. 
Table 1. Students’ performance related to the tool support and their proficiency level.
N 
pre-test 
post-test 
improvement 
tool support 
Group p 
Group c 
27 
22 
56.0 
57.0 
63.6 
60.5 
13.6% 
6.1% 
proficieny 
level 
high 
proficient 
less 
proficient 
25 
24 
62.2 
50.5 
64.6 
59.7 
3.9% 
18.1% 
Note: The proficiency cut-off is 56.5. Full marks = 90.0.
Next, the difference between studens’ performance with the help of tools was 
investigated. The bottom panel of Table 1 shows that with tool support, the less 
proficient students (an increase of 9.1 points) improved more markedly than the high 
proficient ones (an increase of 2.4 points) and reach the statistical significance level (F(1, 
46)=15.45, p value=0.000). It is evident that tools plays an effective role to help the 
less proficient students become engaged learners.
We then took a close look at the relationship between students’ proficiency and tool 
support. Students’ improved scores were submitted to ANOVA. The results did not show 
significant interaction (F(1, 46)=.36, p value=0.554). It indicated either PREFER or 
existing tools did not show significant difference in students’ improvement. However, 
tools did serve a useful purpose for learning synonymous formulaic expressions. The 
result corresponded to the above analysis.
We also explored students’ reflections on the effectiveness of tool support based on their 
questionnaire responses. Based on student reported data, approximately 76% of 
students expressed satisfaction with the synonymous formulaic expressions PREFER 
provides.
Overall, learning synonymous formulaic expressions is an effective and systematic way to 
expand and deepen vocabulary knowledge (Webb, 2007). Taking advantage of the 
computer technology, the PREFER system, bridging the gaps existing in current reference 
tools, provides a relatively comprehensive view of formulaic expressions sharing the 
same meaning but various usages to assist learners in achieving substantial vocabulary 
gains.

---

-88- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be