ArticlePDF Available

Do Productive Skills Improve in Content and Language Integrated Learning Contexts? The Case of Writing

  • Universitat Internacional de Catalunya. Barcelona

Abstract and Figures

This study investigates the differential effects of two learning contexts, formal instruction (FI) and content and language integrated learning (CLIL), on the written production skills of intermediate-level Catalan Spanish adolescent learners of English as a foreign language. Written samples elicited through a composition at two data collection times over one academic year were quantitatively and qualitatively assessed for complexity, accuracy, and fluency and for task fulfilment, organization, grammar, and vocabulary, respectively. Based on the findings, the superiority of CLIL cannot be confirmed: although improvement in the case of the FI + CLIL group is shown, results were only significant in the domain of accuracy.
Content may be subject to copyright.
Applied Linguistics 2015: 1–24 ßOxford University Press 2015
Do Productive Skills Improve in Content
and Language Integrated Learning
Contexts? The Case of Writing
Universitat Pompeu Fabra
This study investigates the differential effects of two learning contexts, formal
instruction (FI) and content and language integrated learning (CLIL), on the
written production skills of intermediate-level Catalan Spanish adolescent lear-
ners of English as a foreign language. Written samples elicited through a com-
position at two data collection times over one academic year were quantitatively
and qualitatively assessed for complexity, accuracy, and fluency and for task
fulfilment, organization, grammar, and vocabulary, respectively. Based on the
findings, the superiority of CLIL cannot be confirmed: although improvement in
the case of the FI + CLIL group is shown, results were only significant in the
domain of accuracy.
Second language acquisition (SLA) research has often related differences in
linguistic outcomes achieved by language learners to differences in learning
contexts (Freed 1998;Collentine and Freed 2004), and has examined what
such contexts offer in terms of language exposure and opportunities for practi-
cing the target language (DeKeyser 2007).
In recent decades, a new learning context called Content and Language
Integrated Learning (CLIL) has spread throughout European education. The
term has gained popularity as the way to refer to an educational approach
adopted in school programmes at primary and secondary levels, whereby a
second or foreign language is used as the medium of instruction for specific
curricular subjects (see Dalton-Puffer 2007).
The CLIL approach has been
repeatedly recommended by European institutions for its multidimensional
nature, as it aims at second/foreign language education, the promotion of
plurilingualism, pluriculturalism, and learners’ international stance (see Ruiz
de Zarobe and Lasagabaster 2010;Cenoz et al. 2014;Pe
´rez-Vidal 2014).
Typically, CLIL has been adopted as part of a school strategy, being used at
specific points in the curriculum; the favourite CLIL subjects being Science,
Physical Education, Informatics, and Arts and Crafts.
Hand in hand with the growing popularity of CLIL programmes, an initial
body of research analysing progress made by learners in such European CLIL
contexts has been taking shape (to name but a handful of studies carried out in
Applied Linguistics Advance Access published October 30, 2015
by guest on November 23, 2015 from
Spain, see Mun
˜oz and Nave
`s 2007;Lorenzo et al. 2009;Lasagabaster and Ruiz
de Zarobe 2010;Pe
´rez-Vidal 2011). To a certain extent, research in Europe has
tried to find out whether findings so far concerning the beneficial impact of
CLIL contexts are comparable with those reported in the literature on the
results of Canadian immersion programmes of different sorts. In these
Canadian programmes, students’ proficiency is described as comparable with
that of second language speakers who are relatively fluent and effective
communicators, but non-target like in terms of grammatical structure,
and non-idiomatic in their lexical choices and pragmatic expressions—in
comparison with native speakers of the same age (Cummins and Swain
1986;Genesee 1987;Swain and Lapkin 1995;Lyster 2007). One methodo-
logical concern which most existing studies have raised regarding the effects of
CLIL contexts on learners’ linguistic progress is that, more often than not,
either the CLIL groups are pilot groups in schools, as CLIL programmes are
young, or students in them are selected, be it by the school itself or through a
process of natural selection, so that it is the best or most motivated students
who choose to join the programme (Moore 2007). In this respect, the current
article aims at filling a research gap in the literature and adding further evi-
dence to the already existing one, with a study in which the CLIL programme
is neither a pilot programme nor one in which learners are selected. The pro-
gramme had been thoroughly designed and planned by the school previously
to its implementation, and it involves one subject, Science, taught through the
medium of English as the only option for all learners alike, beginning at pri-
mary education level (age 10). However, when implementation began, it was
introduced at stages, so, over one year, CLIL groups and non-CLIL groups
coexisted, which made a contrastive study possible. Hence, the current study
examines the effects of such a programme, in contrast with the effects of a
conventional formal instruction (FI) programme. Further, the study measures
development in the learners’ written abilities, one of the least examined to
date in the CLIL literature, across both quantitative and qualitative
From an educational perspective, CLIL programmes have been described as
‘the second time around’ or an improvement on communicative language
teaching (CLT) after the success of the communicative approach (see Pe
Vidal 2009). In that respect, robust CLIL programmes may be claimed to
stretch learners’ use of languages, making them act as language users rather
than language novices in the classroom (Dalton-Puffer 2007) and, accordingly,
lead them to use the language in a more naturalistic manner than in FI. This is
central to the comparison established in this article, the fact that CLIL stands in
contrast to the experience which learners go through in FI classrooms. In
them, and even if CLT is adopted, communication is not clearly always
taking place through the medium of English, or, in the worse possible case,
by guest on November 23, 2015 from
it often deals with form rather than content, with few opportunities for lan-
guage practice in the target language to take place (Ellis 2001;Doughty 2006).
This is certainly the case of the setting in which our data were collected be-
cause in this specific educational system (primary and secondary education in
Spain), English as a foreign language (EFL) courses are taught following a
centralized curriculum geared towards preparing EFL learners to sit an exter-
nal examination, with an emphasis on academic, and receptive skills. Such an
emphasis has a backwash effect on teaching practices, which easily distances
them from CLT.
Taking a SLA perspective, CLIL programmes use the target language as the
vehicle for communication, and deal with curricular content; hence, they are
meaning-oriented, and naturalistic, a central feature in CLT, instead of focusing
on language as a system as occurs in conventional FI in EFL classrooms (Mun
and Nave
´s2007). According to SLA research, those naturalistic settings for lan-
guage learning generate the optimal conditions for linguistic progress to take
place. Indeed, from the perspective of language acquisition models, such as
Krashen’s Input Hypothesis (1985), or cognitive interactionist models (Long
1996;Gass 1997;Sanz 2014), meaning-oriented classrooms offer high-quality
input and opportunities for output practice and negotiation of meaning, a
requisite for linguistic development to occur. From a cognitive perspective,
such contexts offer additional opportunities for input and output practice, allow-
ing learners to transform declarative knowledge into procedural knowledge,
and, in turn, into automatized knowledge (DeKeyser 2007).
Turning to the scope and findings of CLIL research in Europe, it is important
to highlight here that what emerges from the existing studies so far points at
much of the linguistic impact of CLIL approaches being positive rather than
negative, both linguistically and content wise.
If we refer to skill develop-
ment, the perspective of the current article, the area where a difference be-
tween CLIL and mainstream learners is most noticeable is spontaneous oral
production (Admiraal et al. 2006; Zydatiss 2007; Lasagabaster 2008;Ruiz de
Zarobe 2008); the specific general language areas of improvement being mor-
phosyntax, phonetics, and lexicon. Additionally, CLIL students also seem to be
more fluent and risk-taking, self-rating their abilities significantly higher than
non-CLIL students (Sylve
´n 2006;Dalton-Puffer et al. 2009).
Written abilities, the language competence investigated in this study, have
only recently began to be measured (Jexenflicker and Dalton-Puffer 2010;
Ruiz de Zarobe 2010;Whittaker et al. 2011). In the Basque country, Ruiz de
Zarobe (2010) analysed general written competence by two groups of bilingual
students who followed two different CLIL programmes,
and another group
enrolled in a traditional EFL programme. The results show that both CLIL
groups scored better in written production in relation to the five categories
analysed through a holistic approach: content, organization, vocabulary, lan-
guage usage, and mechanics (with significant differences in content and vo-
cabulary). This suggests that there is a positive relationship between the
amount of exposure and opportunities for communicative interaction through
by guest on November 23, 2015 from
English and written foreign language proficiency. Since this advantage in-
creases with grade, it confirms the effectiveness of the CLIL approach as far
as written production outcomes.
When comparing the writing of CLIL and non-CLIL students in higher col-
leges of technology in Austria, Jexenflicker and Dalton-Puffer (2010) found
that, in the area of lexico-grammar, the CLIL students showed significant ad-
vantages throughout, as they did in vocabulary range and orthographic cor-
rectness. At the level of discourse abilities and textual organization, however,
differences were difficult to discern. The areas analysed using a rating scale
were task fulfilment, organization and structure, grammar, and vocabulary. In
short, current knowledge on the impact of CLIL on progress in writing might
be described as follows:
CLIL students have at their disposal a wider range not only of lexical
but also of morphosyntactic resources, which they deployed in
more elaborate and more complex structures. What was not to be
assumed outright given the focus on meaning (and not form) in
CLIL classrooms is the fact that CLIL students also show a higher
degree of accuracy, not only in inflectional affixation and tense use
but also in spelling. The greater pragmatic awareness of CLIL stu-
dents was shown in their better fulfilment of the communicative
intentions of writing tasks. There are, however, dimensions of writ-
ing on which CLIL experience seemed to have little or no effect.
(Dalton-Puffer 2011: 187)
Another interesting approach in research on written production has thrown
significant insights to this matter by comparing CLIL students’ L2 writing abil-
ity with their subject writing in the L1, which, perhaps surprisingly, has not
been found to necessarily surpass CLIL-L2 abilities (Vollmer et al. 2006;
Lorenzo et al. 2009;Ja
¨rvinen 2010;Llinares and Whittaker 2010).
Interesting practical as well as theoretical implications arise from this:
indeed, they suggest deficiencies in writing, both in CLIL and non-CLIL class-
rooms (Llinares and Whittaker 2006;Vollmer et al. 2006). More specifically,
Vollmer et al. (2006) suggest the need for developing learners’ general
writing competence. This is further stressed by Dalton-Puffer in the following
Might we be justified in postulating some kind of general level of
writing development that has an impact on how learners deal with
a writing task independently of whether it is in their L1 or in L2?
This is an issue that needs to be further developed. (Dalton-Puffer
2011: 187)
In a similar vein, Whittaker et al. (2011) and Llinares et al. (2012) take a
further step in trying to come to terms with written competence from a de-
velopmental as well as a pedagogical perspective. They conducted a longitu-
dinal study on written discourse development with a sample of CLIL secondary
schools with EFL in history classes. It showed development in the control of
by guest on November 23, 2015 from
textual resources, as well as some increase in nominal group complexity, over
the four years of the study. The authors suggest that CLIL settings, which focus
primarily on the learning of content, provide suitable contexts in which to
develop written discourse.
Against this background on the differential effects of CLIL and non-CLIL
contexts on learners’ written development, the study presented here seeks
to fill a research gap. It sets out to contribute further evidence to the already
existing, regarding the acquisition of written English in two differentiated edu-
cational programmes: a conventional FI programme geared towards learners’
development in EFL, and a non-selective, well-established programme includ-
ing, in addition to FI, a CLIL component with a European perspective.
Furthermore, it examines CLIL gains in writing across both quantitative and
qualitative dimensions, taking into account that, to the best of our knowledge,
no study has done so to date.
With the above overarching objective in mind, the study addresses the fol-
lowing research question: How do different learning contexts affect the written
linguistic development of young bilingual secondary education EFL learners
when contrasting two groups with a highly similar accumulated hours of ex-
posure to EFL at the onset of the study? The learning contexts they experience
are different because one group is experiencing FI only, and the other group is
experiencing FI in combination with CLIL; the CLIL hours being the extra
hours which separate them from the FI-only group.
In order to address the above question, the study presented in this section was
Hence, the study focuses on the development of written abilities
over one year, as reflected through the learners’ composing skills measured
with a pre-test–post-test design. It aims at examining written development by
analysing, on one hand, the participants’ progress in the domains of syntactic
and lexical complexity, accuracy, and fluency, and, on the other, their
achievement in the domains of task fulfilment, organization, grammar, and
vocabulary. Quantitative and qualitative analyses were respectively applied.
It must be emphasized, as already mentioned above, that the total number of
hours of exposure to English at pre-test is highly similar albeit not equal, for
both participant groups. Then, during the experimental period (between pre-
test and post-test), both groups had an equal number of FI hours; however, the
experimental group had a 50 per cent additional number of hours of exposure
to English through the CLIL programme, with the subject Science being taught
in this language. As Tables 1 and 2display, this design has allowed us to
analyse time issues in the following manner: we contrast the effects of a FI
and a FI + CLIL programme, which are different in that one includes FI only
but the other includes FI and an additional CLIL component, while keeping the
number of hours of instruction highly similar at pre-test (Mun
˜oz 2012). Thus,
the linguistic progress made by the two groups A and B is respectively
by guest on November 23, 2015 from
measured, allowing us to gauge the impact of the additional (and also quali-
tatively different) hours of exposure to English experienced by Group A (GA)
in the CLIL programme. Such a design represents a methodological alternative
to comparing FI with CLIL intact groups: they are more often than not differ-
ent from one another from the start. Indeed, as suggested above, CLIL students
have been censed to attract the best students (Moore 2007). In this respect,
regarding the learners’ level of English at the onset of the study, the CLIL stu-
dents have been often described the best students, as all students enrol on the
CLIL programme, and reportedly find themselves at an intermediate level of
English with an accumulated 1,300 h of instruction at pre-test (´az 2013).
What follows is a more accurate description of the CLIL programme on focus.
Context: the CLIL programme
The study is conducted within the Catalan/Spanish bilingual educational con-
text in Catalonia. The setting in which the subjects are immersed has been
defined as additive trilingualism (Cenoz and Valencia 1994). In Catalonia,
Catalan, the language of instruction, together with Spanish, are the majority
languages, and English is taught as the main foreign language in mainstream
education, with the recent introduction of CLIL programmes in some schools.
In this environment, the foreign target language, English, is hardly ever used
and heard outside the school setting.
Table 1: Participants (N = 50)
AoI in English Data collection T1 Data collection T2
GA: FI + CLIL Grade 7 (12/13 years) Grade 8 (13/14 years)
FI: Nursery (5/6 years)
CLIL: Grade 5 (10/11 years)
FI: Nursery (5/6 years) Grade 8 (13/14 years) Grade 9 (14/15 years)
Table 2: Design
T1 T2
GA: FI + CLIL Grade 7 (12/13 years) Grade 8 (13/14 years)
FI: 1,120 h + CLIL: 210 h FI: 1,260 h + CLIL: 280 h
= 1,330 h= 1,540 h
GB: FI Grade 8 (13/14 years) Grade 9 (14/15 years)
FI: 1,260 h FI: 1,400 h
by guest on November 23, 2015 from
Learners are also studying a second foreign language in school at later ages
(see Vila-Moreno 2008 for a description of educational multilingualism in
Catalonia). It must be borne in mind that learning foreign languages through
the CLIL approach is not only a European recommendation which member
states have diligently taken on board, as mentioned above, but also an attract-
ive opportunity for schools and teachers. This is especially so in bilingual con-
texts in which two co-official languages coexist, such as the one presented here
(the Catalan one), as CLIL was initially the approach adopted to revitalize
Catalan after the Spanish dictatorship (all subjects except for Spanish language
have since been taught through Catalan, Pe
´rez-Vidal 2002). However, CLIL is
also an attractive opportunity for schools and teachers who do not belong to
bilingual settings, as many monolingual communities such as Andalusia have
shown (see Lorenzo et al. 2009). All in all, Catalonia is a fertile ground for
conducting research on the effects of CLIL approaches to education, that is,
CLIL contexts of learning, in the case of foreign languages. As a case in point,
the school that generously offered to contribute to our research goal is a state-
run (ordinary government-supported) school, catering for education from
infant school to post-compulsory second. In the year 2001, the school con-
sidered the possibility of introducing a third language, English, as the medium
of instruction in the school curriculum, in parallel to the conventional EFL
programme. The school carefully prepared the programme over one year, in
collaboration with academic scholars specialists in CLIL (Pe
´rez-Vidal and
Escobar 2004). They designed the programme, they trained teachers, who
produced specific plans and materials, while all stakeholders were being ad-
equately informed and involved in the project (see Pe
´rez-Vidal and Escobar
2004 for an account of the planning process).
As regards the design and the
implementation of the CLIL programme, it was decided that it would start a
year later, in 2002, and in a staged manner, as already mentioned, with lear-
ners who were 8 and 10 years old, that is, at Grades 3 and 5 at the start, and
those in Grades 4 and 6 in the second year of implementation. Hence, the
school’s new Science CLIL programme progressively involved all learners as of
their third grade, year by year.
Participants were two groups of Catalan/Spanish bilingual EFL learners for
which English was their L3. GA (N= 50) was the experimental group experi-
encing the FI plus CLIL, so they are the FI + CLIL group in the design of the
study (from now on GA: FI + CLIL group). Group B (GB) is the control group
(N= 50) experiencing only FI, so they are the FI group (from now on GB: FI
group). There are 50 per cent of males and females in each group. Ages range
from 12 to 15.
Having been together in the same school since nursery (at the age of three in
Spain), both groups had started learning English at the age of five/six, hence
shared the age of onset of instruction (AoI). By the time both groups of
by guest on November 23, 2015 from
learners were tested at T1, they had an accumulated 1,300 hours of instruc-
tion. Data collection started at the end of their first year of secondary education
(Grade 7) at the age of 13. They had both then had eight years of FI. However,
GA: FI + CLIL had received three years of the extra CLIL hours, since the age of
10 (Grade 5). In order to make comparisons possible, GA: FI + CLIL was not
matched for age with GB: FI, which would have created a disadvantage in
terms of time of exposure to English at T1, but for as similar as possible total
number of hours of exposure. Consequently, this entailed that the latter group
included learners who were a year older than the former, as Table 1 displays.
Design and rationale of the study
The study has a longitudinal pre-test–post-test design as Table 2 shows. Both
groups of learners were measured respectively before (pre-test) and after (post-
test) one academic year in order to tap into gains obtained over the course of
that year. Then, as their respective accumulated hours of exposure to English
were very similar at pre-test, that is, at the first data collection time (T1),
although for GA: FI+CLIL some of the hours had been CLIL hours, the dif-
ference in gains obtained by each group over that year was calculated.
The quantity of hours of exposure received after the experimental
period being 50 per cent higher for GA: FI + CLIL, and the quality of
those hours, the CLIL hours, being different, any significant difference in
the results obtained by each group after a year’s treatment would reveal what
kind of effect the hours of CLIL treatment had on the GA: FI + CLIL learners’
linguistic knowledge.
GA: FI + CLIL learners were measured in secondary school when they were
13 (T1) and 14 (T2) years old, at the end of Grades 7 and 8, respectively. At T1,
they had altogether eight years of FI and three years of CLIL. At T2, they had
had nine years of FI and four years of CLIL. GB: FI learners were measured in
Grades 8 and 9 when they were 14 (T1) and 15 (T2) years old, respectively,
also at the end of each academic year. They had had altogether 9 years of FI at
T1, and 10 years of FI at T2. As already mentioned and needs emphasizing,
they were measured at different ages in order to have a similar number of
accumulated hours of total exposure to English although in different learning
contexts (GA: FI + CLIL, whereas GB: FI) (see Table 2).
Table 2 displays the accumulated number of hours of English at T1 and T2
for each group. In the case of GA: FI + CLIL, at T1 data collection, in addition to
1,120 hours of FI (approximately 140 per year since AoI), they had had a total
of 210 of CLIL hours (70 per year). Their total exposure to English was 1,330 h.
One year later, at T2, GA had had 1,260 h of FI and 280 h of CLIL, that is,
1,540 h in total. GB: FI, at T1 data collection, had had a total of 1,260 h of FI
(approximately 140/year since AoI), and at T2, a total of 1,400 h.
In order to assess the differential degree of gain between both groups, GA: FI
+ CLIL gains between T1 and T2 are compared with gains by GB: FI, the control
group, throughout the same period. The design allows for a between-groups
by guest on November 23, 2015 from
comparison of the effect of 50 per cent more hours of exposure and CLIL
instruction over one year: 210 h (140 FI + 70 CLIL) in GA versus 140 h (FI)
in GB.
Instruments and data collection procedures
The written test administered in the current study was part of a larger battery
of tests tapping on both productive and receptive skills of a larger study, as
mentioned above.
In order to gauge the participants’ written production, they
were administered a writing activity in class groups, in an exam-like situation.
They had to write a dialogue on the basis of a picture. It showed two policemen
talking to a mother and a boy at the entrance of their home. Learners were
shown the picture and then they were asked to answer the following two
Why did this happen?
How do you think the situation will end?
They were allowed 20 min to complete the task. The choice of a composition
was based on the number of subskills that come to play when learners write a
piece of text, and it is also a task that is practised in the classroom (Foster and
Skehan 1996). The picture and questions were chosen because it was thought
that the young boy in the picture would allow for a process of identification on
the part of our participants, something which should be inspiring for them
when writing (Foster and Skehan 1996;Tavakoli and Foster 2008). Finally,
since they were asked to write a dialogue and the answer to the two questions
(two very short narratives), different genres had to be used.
Analysis and measures
The participants’ compositions obtained from the writing test were transcribed
using the CLAN programme. As Table 3 shows, they were then analysed quan-
titatively for lexical (Guiraud’s index) and syntactic complexity (coordination
index), fluency (total words in 20 min), and accuracy features (errors per
word), following Wolfe-Quintero et al. (1998). The data were also analysed
qualitatively following a rating scale (Friedl and Auer 2007), whereby task
fulfilment, organization, grammar, and vocabulary features were measured.
Table 3: Measures used to analyse written development
complexity CI
complexity GI
Accuracy E/W Fluency WM
Qualitative measures Task fulfilment Organization Grammar Vocabulary
CI = coordination index; GI = Guiraud’s index; E/W = errors per word; WM = words per minute.
by guest on November 23, 2015 from
Results were introduced to a Stats Graphic matrix, and the formulae for each
ratio were calculated. After that, mean results for all measures per group were
drawn and compared with an analysis of variance (ANOVA) statistical analysis.
In order to address the research question, comparisons between the progress
made in written abilities over one academic year (T2-T1) by the two groups of
participants, GA and GB, are established. That is, the effect of the number of
hours and treatment in GA, 210 h (140 FI + 70 CLIL), versus that of B (140 FI)
are scrutinized. First, the analyses of the data with quantitative measures are
presented, and then writing results measured qualitatively are displayed in the
following sections.
Writing: quantitative measures
The participants’ results in the domains of syntactic and lexical complexity,
accuracy, and fluency are presented in this section.
Syntactic complexity
An ANOVA statistical analysis with the significance level set at <0.05 was
performed. The results appear in Figure 1. It must be noted that when using
the Coordination Index, a lower ratio, reflecting higher amounts of subordin-
ation, is expected in the post-test than in the pre-test. GA (FI + CLIL) obtained
a coordination index of 0.40 at T1, versus a coordination index of 0.39 at T2.
This results in a 0.01 improvement. In contrast, GB (FI) obtained an average of
0.47 at T1 versus an average of 0.49 at T2. This results in a loss of 0.02.
Therefore, there was a minor increase in the level of ability in GA (FI +
CLIL) versus a decrease in GB’s (FI) level of ability once the marks were
averaged. The comparison between GA’s progress (T2-T1) and GB’s progress
(T2-T1) shows that greater progress is made by GA than by GB. However, this
difference turned out not to be statistically significant [F(1,196) = 0.25,
p= .6201].
The previous results are further represented in Figure 2. It shows how GA
(FI + CLIL) makes a small progress since there is a 0.01 decrease in the coord-
ination index and therefore an increase in the use of subordination. In con-
trast, such a progress cannot be seen in GB (FI), as the decrease in the
coordination index reflects the fact that it produces more coordinate clauses
at T2 than at T1. Hence, in a FI (GB) context our subjects make use of higher
levels of coordination than in a FI + CLIL (GA) context. However, these results
are not significant. These gains actually occur at the expense of subordination.
One point deserves further attention here; Figures 1 and 2, in addition,
clearly show how GA (FI + CLIL) subjects start with a lower coordination
index, hence with higher levels of subordination. This fact must be emphasized
by guest on November 23, 2015 from
as it places GA (FI + CLIL) at a different onset level than GB (FI), something
worth taking into account for the discussion of results.
Lexical complexity
Guiraud’s index was used to measure changes in the vocabulary used by our
two groups of subjects after a FI + CLIL (GA) treatment and a purely FI (GB)
In this case, the ANOVA revealed that the differential effect between GA’s
(FI + CLIL) progress (T2-T1) and GB’s (FI) progress (T2-T1) as regards lexical
complexity in the compositions analysed showed greater progress made by GB
(FI). However, the difference was not statistically significant
Figure 1: Average performance in the syntactic complexity measure (coordin-
ation index) at T1 and T2
Figure 2: Progress in one year in GA (FI + CLIL) and GB (FI) syntactic
complexity (coordination index) measure
by guest on November 23, 2015 from
[F(1,196) = 0.69, p= .406]. Guiraud’s index was obtained by dividing the total
number of lexical types by the square root of the total number of lexical
tokens. Therefore, a higher value in the index indicated that the participants’
vocabulary was richer. Figure 3 shows the average performance in this area of
lexical richness and their graphic representation. On one hand, GA (FI + CLIL)
obtained a Guiraud’s index of 6.49 at T1 and 6.7 at T2. This is an improvement
of 0.21. On the other hand, GB’s (FI) Guiraud’s index was 6.3 at T1 versus 6.7
at T2. This is a 0.41 gain. Hence, the rate of progress was higher in GB (FI).
Such a difference in progress between the groups was not statistically
As in the previous area, GA (FI + CLIL) starts the treatment with a higher
onset level. This is not paired with higher gains throughout. In Figure 4, the
three phenomena are clearly shown: the one year progress in lexical richness
both in GA (FI + CLIL) and GB (FI), the faster rate of progress made by GB (FI),
and the higher starting level of GA (FI + CLIL).
Figure 3: Average performance in the lexical complexity measure (Guiraud’s
index) at T1 and T2
Figure 4: Progress in one year in GA (FI + CLIL) and GB (FI) lexical com-
plexity (Guiraud’s index) measure
by guest on November 23, 2015 from
The third domain tapped in our quantitative analysis on writing is accuracy.
Accuracy was measured by means of errors per word. In this domain, GA’s (FI
+ CLIL) progress (T2-T1) was significantly higher than gains made by GB
(FI) (T2-T1) [F(1,196) = 4.41, p= .037].
As shown in Figure 5, GA (FI + CLIL) obtained an average performance of
0.12 errors per word at T1 versus a result of 0.078 errors per word at T2, hence
a decrease in errors per word. Therefore, GA improved by 0.042. In contrast,
progress was not so high in GB (FI), as they obtained 0.092 errors per word at
T1 versus 0.086 errors per word at T2. Hence, it only reached a 0.006
In Figure 6, GA’s (FI + CLIL) higher advantage at T2 is evident if we compare
it with the progress made by GB (FI) in the same year. A higher decrease in the
Figure 5: Average performance in the accuracy measure (errors per word) at
T1 and T2
Figure 6: Progress in one year in GA (FI + CLIL) and GB (FI) accuracy
(errors per word) measure
by guest on November 23, 2015 from
number of errors per word in GA (FI + CLIL) is clearly shown. This, combined
with the fact that the group had started with a higher number of mistakes at
T1, makes their improvement in accuracy outstanding.
The last domain scrutinized quantitatively in this study as far as writing is
concerned is fluency. The results of the ANOVA statistical analysis run with
the data obtained from the quantitative measure total number of words, when
comparing GA’s (FI + CLIL) progress (T2-T1) and GB’s (FI) progress (T2-T1),
showed that both groups decreased in fluency but GB more than GA.
However, this difference turned out not to be statistically significant
[F(1,196) = 0.08, p= .7801].
Figure 7 shows the average performance of each group in the domain of
fluency. At T1, GA (FI + CLIL) produced an average of 146.2 words in the
compositions analysed but, surprisingly, after one year’s treatment, the total
number of words produced decreased to 145.1. It decreased by 1.1 points.
Similarly, GB’s (FI) total number of words decreased from 149.1 at T1 to
144.7 at T2. This is a 4.4 point decrease.
Surprisingly, both groups are less fluent after one year’s treatment as mea-
sured by total number of words. It is also remarkable that GB (FI) started with
a higher degree of fluency than GA (FI + CLIL), just as had happened with the
number of errors, where it had an initial advantage over GA (FI + CLIL).
Finally, the advantage of GB (FI) over GA (FI + CLIL) on lexical complexity
is also something deserving attention.
Figure 7: Average performance in the fluency measure (total number of
words) at T1 and T2
by guest on November 23, 2015 from
So far the quantitative results allow us to identify the following trends. First,
GA (FI + CLIL) only outperforms GB (FI) significantly in gains in the domain of
accuracy. However, GA (FI + CLIL) shows a tendency towards surpassing GB
(FI) as far as syntactic complexity goes. Another relevant result is the fact that
in the domain of lexical and syntactic complexity GA (FI + CLIL) has a higher
onset level than GB (FI), whereas in the domain of accuracy and fluency, in
contrast, a lower onset level.
It is interesting at this point to gain an overall appraisal of each context in
turn and how each of the groups progresses. On one hand, after the CLIL
treatment, GA (FI + CLIL) writes shorter texts, which nonetheless are more
accurate, lexically richer, and syntactically more complex. On the other hand,
after the FI treatment, GB (FI) also writes shorter texts which are more accur-
ate and lexically richer, however, less syntactically complex. Hence, they both
make some progress except in fluency, but GA’s (FI + CLIL) progress in accur-
acy significantly outrates that of GB (FI). If we now contrast their gains, GA
(FI + CLIL) gains more than GB (FI) in syntactic complexity and significantly
more in accuracy, having started at a lower level than GB (FI). GB (FI) in turn
gains more than GA (FI + CLIL) in lexical complexity, in turn having started at
a lower level.
Writing: qualitative measures
The results for the qualitative measures used in the compositions written by
the participants: task fulfilment, organization, grammar, and vocabulary, are
presented in this section.
Task fulfilment
In relation to task fulfilment, GA’s (FI + CLIL) progress (T2-T1) was higher
than GB’s (FI) progress (T2-T1). The difference between both groups’ progress,
however, was not statistically significant [F(1,96) = 0.20, p= .6572].
When measuring the compositions according to six behavioural levels on a
scale from 0 (not enough to evaluate) to 5 (very good), GA (FI + CLIL) ob-
tained an average performance of 2.92 at T1, versus 3.29 at T2. This is an
improvement of 0.37. On the other hand, progress made by GB (FI) was not
so high since it obtained an average result of 2.63 at T1, versus 2.87 at T2. This
is a 0.24 improvement.
Figure 8 shows such progress: although being higher in the case of GA (FI +
CLIL), the difference with GB (FI) was not statistically significant.
As far as the organization of the compositions analysed is concerned, GA’s
progress (T2-T1) was again higher than GB’s progress (T2-T1). However, the
ANOVA statistical analysis revealed that the difference was not significant
[F(1,96) = 0.20, p= .6565] as in the previous subsection. When measuring
by guest on November 23, 2015 from
the compositions according to the six behavioural levels explained above, GA
(FI + CLIL) obtained an average performance of 2.84 at T1, versus 3.24 at T2.
This is a 0.4 improvement. In contrast, GB’s (FI) average results were 2.49 at
T1, versus 2.76 at T2. This is a 0.27 improvement.
In Figure 9, these results are displayed: it is interesting to note that in both
this and the previous measure, GA (FI + CLIL) had a higher starting level, as
happened with some of the quantitative measures.
In this area, accuracy, the ANOVA of the participants’ performance revealed
that again GA’s (FI + CLIL) results were higher than GB’s (FI), but that how-
ever no significant differences between GA’s (FI + CLIL) progress (T2-T1) and
GB’s (FI) progress (T2-T1) were to be found [F(1,96) = 0.98, p= .3240].
As can be appreciated in Figure 10, participants in GA (FI + CLIL) obtained
an average performance of 2.4 at T1 versus an average of 3.06 at T2 (from the
six behavioural levels used in these measures). Hence, the improvement
amounted to 0.66. In contrast, progress in GB (FI) was not so high since
Figure 9: Progress in one year in GA (FI + CLIL) and GB (FI) organization
Figure 8: Progress in one year in GA (FI + CLIL) and GB (FI) task fulfilment
by guest on November 23, 2015 from
they obtained an average result of 2.34 at T1, versus 2.7 at T2. This is a 0.36
As far as the vocabulary used in the compositions analysed when qualitative
measures were applied, GA’s (FI + CLIL) progress (T2-T1) was again higher
than GB’s (FI) (T2-T1), but the ANOVA statistical analysis revealed that the
difference between the two groups was not significant [F(1,96) = 2.37,
p= .1256].
Like in the previous subsections, taking into account the six behavioural
levels mentioned, GA (FI + CLIL) obtained an average performance of 2.52
at T1, versus 3.18 at T2. This is a 0.66 improvement. On the other hand, GB
(FI) obtained an average performance of 2.53 at T1, versus 2.74 at T2. This is a
0.21 improvement.
Figure 11 shows higher progress in GA (FI + CLIL) than in GB (FI) from T1 to
T2. However, the difference in progress, contrary to what it may seem, was not
statistically significant.
These results obtained by the subjects in the qualitative measures of their
respective progress in writing can be summarized as follows. When analysed
with qualitative measures, GA (FI + CLIL) outperforms GB (FI) in all areas.
However, results do not reach statistical significance in any of them on any
measure. More specifically, GA (FI + CLIL) consistently tends to write a better
organized, more accurate, lexically richer, and more purposeful composition. It
is also interesting here to highlight that GA (FI + CLIL) always has a higher
onset level except in the domain of vocabulary, albeit being a year younger.
Figure 10: Average performance in the grammar measure at T1 and T2
by guest on November 23, 2015 from
GB’s (FI) improvement in all domains of written competence is always inferior
to GA’s (FI + CLIL).
This article has sought to fill an existing gap in the literature by examining the
effects of a CLIL learning context, in the case of a well-established programme
initiated at primary education level, involving all students at a school located
in an urban area. It focuses on written development and includes quantitative
and qualitative analyses, to our knowledge never used in combination in the
few studies published to date tapping into the CLIL effects on written devel-
opment. The analyses presented above explore the impact of an EFL pro-
gramme including a CLIL component in addition to a FI component (GA), in
contrast with a programme only including a FI component (GB). They try and
pin down whether the differential factor between both groups, that of the
additional CLIL hours in GA, has a significant impact on learners’ linguistic
progress in writing over the course of one year. When measured quantita-
tively, GA’s (FI + CLIL) results, as far as written accuracy is concerned, show
a progress of 0.042 over one academic year. This is significantly higher than
GB’s (FI) results, which improved 0.006 from T1 to T2. Syntactic complexity
also shows evidence of a tendency towards improvement on the part of GA (FI
+ CLIL), however, not to a significant level. Lexical complexity is the only skill
in which GB (FI) outperforms GA (FI + CLIL), albeit non-significantly either.
When we try and picture the results as far as the qualitative measures go, GA
(FI + CLIL) consistently tended to write a better organized, more accurate,
lexically richer, and more purposeful composition, however, not to a signifi-
cant level.
In a nutshell, these results reveal that both groups made progress on all
measures over the year under scrutiny, except for fluency and, in the case
of GB, also except for syntactic complexity. However, none of the differences
were significant, apart from accuracy, where GA (FI + CLIL) outperformed GB
Figure 11: Progress in one year in GA (FI + CLIL) and GB (FI) vocabulary
by guest on November 23, 2015 from
(FI). Therefore, what stands out from this study is that the CLIL addition ac-
tually means a significant gain in accuracy. On the one hand, this result was
not easy to predict, since prior findings in the CLIL literature had revealed that
largest gains tend to be obtained in lexis and pragmatic features as far as oral
production goes. On the other hand, however, the results are consistent with
some prior research on CLIL writing and particularly with regards to accuracy
(Jexenflicker and Dalton-Puffer 2010).
As reviewed in the theoretical background section above, the studies con-
ducted by Whittaker et al. (2011) and Llinares et al. (2012) on written discourse
in history classes had shown development in the control of textual resources,
as well as some increase in nominal group complexity, over the four years of
the study. The authors then suggested that CLIL settings which focus primarily
on the learning of content provide suitable contexts in which to develop writ-
ten discourse, since the secondary students involved in the study seemed to be
able to draw on a solid knowledge base from which to generate their texts. This
is an interpretation which might be extended to explain significant gains in
accuracy in the present study.
Another possible explanation that might justify the significant benefit on the
progress made in the domain of accuracy by GA (FI + CLIL) is the transfer of
knowledge and skills from the FI context to the CLIL context, where mean-
ingful practice provides extra opportunities for linguistic development to take
place. Indeed grammar abilities are often practised in the FI context. This might
confirm the beneficial effect of the additional CLIL hours on the automatiza-
tion of previously acquired knowledge through the extensive practice CLIL
contexts allow (DeKeyser 2007). Empirical studies, probably of an experimen-
tal nature, are needed in order to test such an interpretation.
Indeed, further research on CLIL versus FI contexts in secondary education
should help us find an explanation to the fact that, in the present case, an
addition in terms of hours of CLIL exposure to English has meant a significant
gain in written accuracy. While it should be noted that this finding coincides
with some other studies carried out at other latitudes in the continent (see e.g.
Jexenflicker and Dalton-Puffer 2010), it is also true that contradictory results
in CLIL research are highly frequent, perhaps because of the methodological
concerns mentioned above, but also because research is still in its infancy.
Further empirical enquiry should take into account the fact that in the present
study, the skill of writing has been analysed only through a dialogue and two
short narratives. In order to provide substantial evidence of the learners’ over-
all capacity to write, it would be desirable to also administer descriptive and
argumentative tasks, of a more complex nature, to CLIL learners. Following
Kuiken and Vedder (2007), a more complex task would have probably affected
the accuracy results in writing, particularly with respect to lexical errors.
In conclusion, the results obtained in the present study show that the su-
periority of CLIL cannot be confirmed. Our findings show improvement in
written productive abilities for the FI + CLIL group (GA), as writing shows
development, but significantly so only in the case of accuracy. Consequently,
by guest on November 23, 2015 from
as a final point we can state that, with the sample analysed in the presented
study, a CLIL context of learning over the course of one year, does not suffice
for the participants to improve significantly in all the domains scrutinized.
It seems evident that more research is needed in this subfield of language
acquisition if we wish to obtain a more accurate and precise picture regarding
the linguistic impact of European CLIL programmes. This should also allow us
to address a central issue in SLA research, such as that of the effect of the
language practice which all learners alike, and not only (self-)selected learners,
may avail themselves of in different learning contexts. Having said that, we
would like to argue that CLIL programmes hold the promise of representing an
educational approach which can guarantee successful learning, provided pro-
grammes are carefully designed and developed in each school context, such as
the case of the programme whose data are presented in this study, or the
programmes so carefully and successfully implemented in previously existing
immersion settings.
1 This study has benefitted from grant
2014 SGR 5063, and FFI2013-48640-
2 At tertiary levels, other names are most
commonly used: Integrating Content
and Language (ICL), or, when specific-
ally referring to the overwhelming use
of English as the target language in ICL,
English-Medium Instruction, and,
more recently, English-Taught
Programmes (see Smit and Dafouz
3 Research findings on content-learning
outcomes have been mostly posi-
tive, but critical voices are beginning
to make themselves heard. However,
this is beyond the scope of the
present research study, and no further
reference will be made to this
controversy (see Dalton-Puffer 2011:
188 for a summarized account on this
4 When they were 14 years of age, both
groups entered a CLIL English pro-
gramme, one of the groups was only
taught one CLIL curricular subject,
Social Science for three to four hours
a week, whereas the other two CLIL
curricular subjects Social Science for
three to four hours a week and
Modern English Literature for two
hours a week.
5 This study is part of a broader CLIL
study analysing participants’ oral and
written receptive development in add-
ition to the written productive develop-
ment data presented here (see Roquet
2011 for a full description).
6 We are particularly indebted to the
school in Barcelona which allowed us
to conduct data collection and in par-
ticular its language coordinator (see
´az 2013 for a presentation of the
school project).
7 Production was elicited in writing, and
reception in writing and orally. In add-
ition, lexico-grammatical abilities were
also tapped on. The instruments used to
obtain the data were: (i) a composition
on a given topic measuring written pro-
duction; (ii) a reading task (cloze) and a
dictation measuring written and oral
comprehension; (iii) a sentence trans-
formation test and a grammaticality
judgement test with progressive de-
grees of difficulty in multiple choice
format measuring lexico-grammatical
ability. Data collection in two one-
by guest on November 23, 2015 from
hour sessions was handled by the class
teachers due to institutional
conventions (see Roquet 2011 for a
comprehensive presentation of results).
Admiraal, W., G. Westhoff,andK. de Bot.
2006. ‘Evaluation of bilingual secondary edu-
cation in the Netherlands: Students’ language
proficiency in English,’Educational Research and
Evaluation 12/1: 75–93.
Cenoz, J., F. Genesee,andD. Gorter. 2014.
‘Critical analysis of CLIL: Taking stock and
looking forward,’ Applied Linguistics 35/3:
Cenoz, J. and J. F. Valencia. 1994. ‘Additive
trilingualism: Evidence from the Basque
Country,’Applied Psycholinguistics 15: 195–207.
Collentine, J. and B. F. Freed. 2004. ‘Learning
contexts and its effects on second language ac-
quisition: Introduction,’ Studies in Second
Language Acquisition 26: 153–71.
Cummins, J. and M. Swain. 1986. Bilingualism
in Education. Longman.
Dalton-Puffer, C. 2007. Discourse in Content and
Language Integrated Learning. John Benjamins.
Dalton-Puffer, C. 2011. ‘Content and language
integrated learning: From practice to prin-
ciples?,’ Annual Review of Applied Linguistics
31: 182–204.
Dalton-Puffer, C., J. Hu
¨ttner, V. Schindelegger,
and U. Smit. 2009. ‘Technology-geeks speak out:
What students think about vocational CLIL,’
International CLIL Research Journal 1/2: 17–25.
DeKeyser, R. 2007. Practice in a Second Language.
Perspectives from Applied Linguistics and Cognitive
Psychology. Cambridge University Press.
´az, M. 2013. De la implantacio
´de l’EICLE en
llengua anglesa al multilingu
¨isme. Deu
anys d’EICLE! 2013. ‘Perspectives sobre
l’aprenentatge de llengu
¨es i el multilingu
´Informatiu del Collegi de Doctors
i Llicenciats en Filosofia i Lletres i en Cie
`ncies de
Catalunya 14: 45.
Doughty, C. 2006. ‘Instructed second language
learning: Constraints, compensation and en-
hancement’ in C. Doughty and M. H. Long
(eds): The Handbook of Second Language
Acquisition. Blackwell, pp. 256–311.
Ellis, R. 2001. ‘Memory for Languages’ in
P. Robinson (ed.): Cognition and Second
Language Instruction.Cambridge University
Foster, P. and P. Skehan. 1996. ‘The influence
of planning and task type on second language
performance,’ Studies in Second Language
Acquisition 18: 299–324.
Freed, B. F. 1998. ‘An overview of issues and
research in language learning in a study
abroad setting,’ Frontiers 4: 31–60.
Friedl, G. and M. Auer. 2007. ‘Rating scale used
for assessment of the writing task.
¨uterungen zur Novellierung der
Reifeprufungsverordnung fur AHS, lebende
Fremdsprachen,‘ available at http://www.bifie.
Gass, S. 1997. Input, Interaction and the Second
Language Learner. Erlbaum.
Genesee, F. 1987. Learning through two
Languages. Studies of Immersion and Bilingual
Education. Newbury House.
¨rvinen, H. M. 2010. ‘Language as a meaning
making resource in learning and teaching con-
tent. Analysing historical writing in content
and language integrated learning’in
C. Dalton-Puffer, T. Nikula, and U. Smit
(eds): Language Use and Language Learning in
CLIL Classrooms. John Benjamins,pp. 145–68.
Jexenflicker, S. and C. Dalton-Puffer. 2010.
‘The CLIL differential: Comparing the writing
of CLIL and non-CLIL students in higher col-
leges of technology’ in C. Dalton-Puffer,
T. Nikula, and U. Smit (eds): Language Use
and Language Learning in CLIL Classrooms.
John Benjamins,pp.169–89.
Krashen, S. 1985. The Input Hypothesis.
Kuiken, F. and I. Vedder, 2007. ‘Task complex-
ity and measures of linguistic performance in
L2 writing,’ International Review of Applied
Linguistics 45/3: 261–84.
Lasagabaster, D. 2008. ‘Foreign language com-
petence in content and language integrated
courses,’ The Open Applied Linguistics Journal 1:
Lasagabaster, D. and Y. Ruiz de Zarobe. 2010.
‘Ways forward in CLIL: Provision issues and
by guest on November 23, 2015 from
future planning’ in Y. Ruiz de Zarobe and
D. Lasagabaster (eds): CLIL in Spain:
Implementation, Results and Teacher Training.
Cambridge Scholars Publishing, pp. 278–95.
Llinares, A. and R. Whittaker, 2006. ‘Oral and
written production in social science,’ ViewZ
ViennaEnglish Working Papers 15/3: 28–32.
Llinares, A. and R. Whittaker. 2010. ‘Writing
and speaking in the history class: A compara-
tive analysis of CLIL and first language con-
texts’ in C. Dalton-Puffer, T. Nikula and
U. Smit (eds): Language Use and Language
Learning in CLIL Classrooms. John Benjamins,
pp. 125–44.
Llinares, A., T. Morton and R. Whittaker.
2012. The Roles of Language in CLIL.
Cambridge University Press.
Long, M. H. 1996. ‘The role of the linguistic en-
vironment in second language acquisition’ in
W. C. Ritchie William and T. K. Bhatia
(eds): Handbook of Second Language Acquisition.
Academic Press, pp. 413–68.
Lorenzo, F., S. Casal, and P. Moore. 2009.
‘The effects of content and language integrated
learning in European education: Key findings
from the Andalusian Bilingual Sections
Evaluation Project,’ Applied Linguistics 20/11:
Lyster, R. 2007. Learning and Teaching Languages
Through Content. A Counterbalanced Approach.
John Benjamins.
Moore, P. 2007. ‘Enhancing Classroom
Discourse: a modelling potential for content
teachers’ in F. Lorenzo S. Casal
V. Alba-Quinones and P. Moore (eds):
Models and Practice in CLIL.Revista espan
˜ola de
´stica aplicada (RESLA), 20/1: 141–52.
˜oz, C. 2012. Intensive Exposure Experiences
in Second Language Learning. Multilingual
˜oz, C. and T. Nave
´s. 2007. ‘Spain’ in
A. Maljers, D. Marsh and D. Wolff (eds):
Windows on CLIL. Content and Language
Integrated Learning in the European Spotlight.
European Platform for Dutch Education and
European Centre for Modern Languages, pp.
´rez-Vidal, C. 2002. ‘Spain’ in F. Grenfell
(ed.): Modern Languages Across the Curriculum.
Routledge/Falmer, pp. 19–33.
´rez-Vidal, C. 2009. ‘The integration of con-
tent and language in the classroom: A
European approach to education (The second
time around)’ in E. Dafouz and M. Guerrini
(eds): CLIL Across Educational Levels: Experiences
from Primary, Secondary and Tertiary Contexts.
Richmond Publishing, pp. 25–40.
´rez-Vidal, C. 2011. ‘Language acquisition in
three different contexts of learning: formal in-
struction, stay abroad, and semi-immersion
(CLIL)’ in Y. Ruiz de Zarobe, J. M. Sierra,
and F. Gallardo (eds): Content and Foreign
Language Integrated Learning: Contributions to
Multilingualism in European Context. Peter Lang
Publications, pp. 103–28.
´rez-Vidal, C. 2013. ‘Perspectives and lessons
from the challenge of CLIL experiences’ in
C. Abello-Contesse, P. Chandler, M.
D. Lo
´nez, and R. Chaco
(eds): Bilingual and Multilingual Education in
the 21st Century. Multilingual Matters, pp. 52–
´rez-Vidal, C. 2014. ‘Languages for all in edu-
cation: CLIL and ICLHE at the crossroads of
multilingualism, mobility and internationalisa-
tion’ in M. Juan-Garau and J. Salazar-Noguera
(eds): Content-Based Learning in Multilingual
Educational Environments. Springer, pp. 31–51.
´rez -Vidal, C. and C. Escobar. 2004. ‘Teacher
education for the implementation of CLIL in
the school system’ in B. Wilkinson (ed.):
Integrating Content and Language. Meeting the
Challenge of a Multilingual Higher Education.
Maastricht University, pp. 389–402.
Roquet, H. 2011. Integrating Content and Language
in Mainstream Education in Barcelona. A Study of
the Acquisition of English as a Foreign Language.
Lambert Academic Publishing.
Ruiz de Zarobe, Y. 2008. ‘CLIL and foreign lan-
guage learning: A longitudinal study in the
Basque Country,’ International CLIL Research
Journal 1/1: 60–73.
Ruiz de Zarobe, Y. 2010. ‘Written production and
CLIL: An empirical study’ in C. Dalton-Puffer,
T. Nikula and U. Smit (eds): Language Use and
Language Learning in CLIL Classrooms.John
Benjamins, pp. 191–209.
Ruiz de Zarobe, Y. and D. Lasagabaster. 2010.
CLIL in Spain: Implementation, Results and Teacher
Training. Cambridge Scholars Publishing.
Sanz, C. 2014. ‘Contributions of study abroad
research to our understanding of SLA processes
and outcomes: The SALA Project’ in C. Pe
Vidal (ed.): Second Language Acquisition in Study
Abroad and Formal Instruction Contexts, John
Benjamins, pp. 1–17.
by guest on November 23, 2015 from
Smit, U. and E. Dafouz. 2012. ‘Integrating con-
tent and language in higher education. gaining
insights into English-medium instruction at
European Universities,’ AILA Review 25.
Swain, M. and S. Lapkin. 1995. ‘Problems in
output and the cognitive processes they gener-
ate: A step towards second language learning,’
Applied Linguistics 16/3: 371–91.
´n, L. K. 2006. Extramural Exposure to
English among Swedish School Students and the
CLIL Classroom. Paper presented at the ESSE8
Conference in London, 29 August 2006.
Tavakoli, P. and P. Foster. 2008. ‘Task design
and second language performance. The effect
of narrative type on learner output,’ Language
Learning 58/2: 439–73.
Vila-Moreno, X. 2008. Language-in-education
policies in the Catalan language area. In
J. Cenoz, D. Gorterand, (eds): AILA REVIEW.
Multilingualism and Minority Languages:
Achievements and Challenges in Education, vol.
21. John Benjamins, pp. 31–48.
Vollmer, H. J., L. Heine, R. Troschke,
D. Coetzee, and V. Ku
¨ttel. 2006. Subject-
Specific Competence and Language Use of CLIL
Learners: The Case of Geography in 10 of
Secondary Schools in Germany. Paper presented
at the ESSE8 Conference, London, UK.
Whittaker, R., A. Llinares, and A. McCabe.
2011. ‘Written discourse development in CLIL
at secondary school,’ Language Teaching
Research 15/3: 343–62.
Wolfe-Quintero, K., S. Inagaki, and H. Kim.
1998. Second Language Development in Writing:
Measures of Fluency, Accuracy and Complexity.
University of Hawai’i at Manoa.
Zydatiss, W. 2007. Deutsch-Englische Zu
¨ge in
Berlin (DEZIBEL). Eine Evaluation des bilingualen
Sachfachunterrichts in Gymnasien: Kontext,
Kompetenzen, Konsequenzen. Peter Lang.
by guest on November 23, 2015 from
Helena Roquet is currently a lecturer in the Department of Translation and Language
Sciences at the Universitat Pompeu Fabra (Barcelona) and at the Universitat
Internacional de Catalunya (Barcelona). She also teaches at postgraduate level. In
2003, she finished a Master’s degree at the Universitat Pompeu Fabra ‘Multilingual
Communication: Translation, Literature and Linguistics Studies’. After this, she specia-
lized in the field of Language Acquisition, and she had a research stay at the University
of Vienna’s English Department in order to obtain the European Mention. She com-
pleted her PhD in Second Language Acquisition at the Universitat Pompeu Fabra in
July 2011, when she defended her doctoral thesis. This was on the acquisition of
English as a Foreign Language Integrating Content and Language in Mainstream
Education. Since then, she has started to publish internationally. She is also a
member of the consolidated research group ALLENCAM (Language Acquisition from
Multilingual Catalonia) at the Universitat Pompeu Fabra. Her main research interests
lie within the field of second and third language acquisition, bilingualism, and Content
and Language Integrated Learning. Address for correspondence: Helena Roquet,
Department of Translation and Language Sciences, Universitat Pompeu Fabra,
Barcelona. <>
Carmen Pe
´rez-Vidal is currently a lecturer of English Philology at the Universitat Pompeu
Fabra (Barcelona). She is interested in second/foreign language acquisition by young
and adult learners of English, and child bilingualism—her PhD research (1995)—
Journal of Child Language and Bilingualism Language and Cognition. She has recently
examined different learning contexts, that is, Study Abroad, Content and Language
Integrated Learning, and formal instruction, topics on which she has extensively pub-
lished internationally (EUROSLA Yearbook, IRAL, GURT Proceedings, Research in
Language) and also teaches at postgraduate level. <>
by guest on November 23, 2015 from
... For many years, CAF metrics have been used for measuring second language development (Larsen-Freeman, 2009). Nine studies in this review used at least one CAF metric to measure writing (Ackerl, 2007;Celestén Pérez and Basse, 2015;Gené-Gil et al., 2015a, 2015bLahuerta, 2020;Lahuerta Martínez, 2017;Navés, 2011;Roquet and Pérez-Vidal, 2017;Tai, 2015). Complexity in these studies often examines syntactic or lexical complexity. ...
... Unlike the measures described above that focus either exclusively on a language-or content-related aspect, nine studies included in this review utilized rubrics in order to evaluate both language and content aspects within writing simultaneously (Chansri and Wasanasomsithi, 2016;Coetzee-Lachmann, 2007;Gené-Gil et al., 2015a;Jexenflicker and Dalton-Puffer, 2010;Lahuerta Martínez, 2017;Miret Pallarols, 2009;Roquet and Pérez-Vidal, 2017;Ruiz de Zarobe, 2010;Sagrario Salaberri Ramiro and Del Mar Sánchez Pérez, 2015). The majority of these studies used a five-point rubric called the ESL composition profile (Jacobs et al., 1981) comprised of scores for content, organization, vocabulary, language use, and mechanics. ...
... Five studies used one or more CAF metrics to examine differences between CLIL/EMI and traditional language classrooms. For syntactic complexity, two studies have found CLIL/EMI groups to score higher than traditional language classrooms (Lahuerta Martínez, 2017;Navés, 2011) and three found no differences between the two groups (Gené-Gil et al., 2015a, 2015bRoquet and Pérez-Vidal, 2017). In terms of lexical complexity, none of the studies found significant differences between the groups. ...
This study utilizes a systematic review approach to examine current research in content and language integrated learning (CLIL) and English-medium instruction (EMI). Studies were examined for their use of language-related and content-related writing measurement and outcomes. Findings on CLIL/EMI writing measurement shows that studies have utilized a variety of methods for measuring language, but measurement of the content dimension is limited. While some studies in the review combine multiple types of writing measures, CLIL writing research often does not simultaneously assess both the language and content dimensions of writing. For writing outcomes, the reviewed studies show CLIL/EMI generally equals or exceeds traditional language classrooms and that CLIL/EMI instruction leads toward growth in some but not all (i.e. accuracy), language-related metrics. However, regarding the content dimension, the findings are mixed and show CLIL/EMI writing generally does not communicate disciplinary content knowledge appropriately. It is recommended that future research gives more attention to measuring both the language and content dimensions of writing while also incorporating new methods for measuring discipline-specific content in writing. It is also recommended that instructors increase their language awareness and reconsider how disciplinary writing is taught in their CLIL/EMI classroom.
... Firstly, as this is a one-shot research study, no control group was assigned to compare the effectiveness of TBI. Although some scholars (e.g., Lahuerta, 2020;Roquet & Pérez-Vidal, 2017;Suzuki, in press) have investigated the effects of CBI on learners' writing performance with comparative groups, Hinkel (2020) maintains that the number of comparative studies on the effects of the instruction are still scarce in the second language writing field. Thus, in order to explore the effectiveness of TBI on upper-intermediate English proficiency level Japanese university students' LD in their academic writing, it would be necessary to compare them with those of non-TBI groups in future research. ...
Full-text available
Theme-based instruction (TBI) has been prevalent as one type of effective language teaching pedagogy in various English as a-second/foreign-language-teaching contexts, including Japan. Although some research studies on the effectiveness of TBI on second language learners' language learning have been conducted, the effects of instruction on second language writers' writing abilities, including lexical diversity (LD), have not been fully revealed. Therefore, this research paper aims to explore the effectiveness of TBI on 45 Japanese university students' LD in writing based on their LD scores in their essays and drafts that were analysed with text inspector (2019), Friedman tests with Bayes factors and Conover post hoc comparisons followed by the Bonferroni correction. In addition, the participants' learning reflection notes coded thematically are considered to further investigate the effectiveness of TBI. Findings indicated that, regardless of the assigned themes, the participants developed their LD statistically significantly. Furthermore, students' reflection notes also illustrated that topic familiarity and sociocultural scaffolding provided by the teacher-researcher, their classmates, and materials played an important role in their LD development in writing. At the end of the research paper, limitations to address will be considered for future research. テーマ型内容中心教授法(TBI)は、日本を含めた多くの第二言語もしくは外国語としての英語教 育環境において効果的な言語教授法の一つとして普及している。言語学習者の言語学習におけ るこの教授法の効果は研究されているが、書き言葉における語彙の多様性を含めた、ライティ ング能力においての効果は完全には明らかにされていない。本論文では、text inspector (2019) とベイズ推定量を用いたフリードマン検定とボンフェローニ補正されたコノバー検定で分析し た学習者のエッセイ及び下書きを基に、日本人大学生 45 名のライティングでの語彙の多様性に おける TBI の効果を探究していく。加えて、コード化された学生たちの学習振り返りノートを 用いて TBI の効果を考察していく。結果として、テーマに関係なく語彙の多様性が統計的に向 上し、テーマへの馴染み度や教員、クラスメート、そして教材等によって与えられる足場作り (scaffolding)が重要な役割を担っている事が明らかとなった。最後に、将来的な研究調査のた めに本研究での限界点を考察する。
... Content and language integrated learning (CLIL) is an educational approach that integrates learning content along with a foreign language (Mehisto et al., 2008). Despite the above statement, the last decade has witnessed quite a large amount of CLIL research hinging on the effect of CLIL on language achievement (Admiraal et al., 2006;Dalton-Puffer, 2008;Roquet and Pérez-Vidal, 2017). CLIL research so far has primarily pursued L1 (first language) and L2 (second language). ...
Full-text available
Can engaging university students in Content and Language Integrated Learning (CLIL) increase their achievement in Biology and English language? The current study explored the effectiveness of team teaching enhanced CLIL on student achievement. Framed in interdisciplinary/cross-curricular teaching, we examined the effect of CLIL strategy on student achievement in a quasi-experimental study with a sample of Biology education students (N = 25) assigned to control and experimental groups. The topic taught through CLIL was the digestive system, and this subject knowledge was used to test the students' achievement in Biology. Along with t-test and Mann-Whitney U-test a non-parametric ANCOVA was carried out to reveal group differences. We found that engaging in CLIL significantly improved student achievement both in Biology subject knowledge and English language. We highlight the critical role of CLIL in teaching content and language for science subjects such as Biology at university level.
... Roquet and Pérez-Vidal [129] Exercise of English writing ability in CLIL course. ...
Full-text available
This article presents a review of the literature on Content and Language Integrated Learning (CLIL) from 2011 to 2020. We analyzed 100 research articles based on the First Principles of Instruction (FPI), with the aim of considering the problems of CLIL and exploring the possibilities of improvement based on these principles. We found that the focus of almost all the articles was English as a Foreign Language (EFL) as the target language; rarely was a concern for other languages indicated. Moreover, current studies pertaining to CLIL paid little attention to the FPI. In this article, we offer some suggestions for future research.
... The only reason for this is that writing entails creating ideas based on specific criteria. For example, writers should emphasize the conceptual underpinnings that complexity, accuracy, fluency, organization of ideas, grammar, and vocabulary are reflective componential parts (Kathpalia & Heah, 2008) of writing assessment (Roquet & Pérez-Vidal, 2015). Because writing is often perceived as the most difficult skill for students to master, it is challenging for EFL students to brilliantly strengthen all of these components. ...
Full-text available
English as a foreign language (EFL) university students use reflective journals as learning logs to express or capture their ideas within a scientific conceptual framework. The objectives of this research are to (1) describe the power of reflective journal writing in communicating ideas, and (2) identify the aspects of reflective journal writing that aid learning in an EFL context. The reflective journals were written by 21 EFL university students. This research takes a qualitative approach, with the primary data coming from several reflective journals (N=124) while the secondary data coming from EFL students’ interviews (N=15). The research results showed that reflective journals were useful for students to make critical reflections and self-discovery responses to writing topics. The students learned to focus on writing components such as order, unity, coherence, cohesiveness, content, and organization of ideas through reflective journal writing. The students’ perspectives on aspects of reflective journal writing were primarily concerned with macro-and micro-level linguistic issues, as evidenced from the interview results. Writing a reflective journal necessitated their ability to reformulate thoughts, provide details, and solve problems. Furthermore, critical thinking, metacognitive skills, and self-reflections became increasingly important in helping the students to develop their ability to write reflective journals.
The benefit of Content and Language Integrated Learning (CLIL) for second language (L2) development has been addressed by second language acquisition theory and investigated by empirical studies. However, previous studies have not demonstrated the effectiveness of CLIL precisely as most of their study participants took CLIL and non-CLIL courses concurrently. Additionally, existing studies have focused on the effectiveness of CLIL in primary and secondary education and rarely addressed it in higher education. The present study considered the limitations of the previous studies and investigated the development of L2 English productive skills (i.e., speaking and writing) through CLIL conducted in a university in Japan. This study employed a pre-experimental pretest-posttest design, in which all participants received CLIL instruction for four months undertaking the pretest and posttest. The results showed that their productive skills improved significantly. The analysis of pretest and posttest scores also revealed the specific areas of productive skills that benefitted from CLIL, such as Task Achievement (speaking and writing), Coherence and Cohesion (writing), and Grammatical Range and Accuracy (writing). However, this study revealed that the degree of improvement was small, and that vocabulary and critical thinking skills did not improve significantly. This study’s findings provide empirical evidence for the effectiveness of CLIL itself on L2 development in higher education.
El número de centros de enseñanza bilingüe ha crecido considerablemente en España y la metodología AICLE (Aprendizaje Integrado de Contenido y Lengua Extranjera) es muy popular a nivel nacional e internacional. Debido a esta popularidad, cada vez más alumnos se incorporan a programas bilingües, pero ¿son estos programas efectivos? El trabajo que aquí presentamos pretende analizar, mediante pruebas orales, la diferencia de nivel de expresión oral, en lengua inglesa, entre los alumnos de 4º curso de Educación Secundaria Obligatoria pertenecientes a programas bilingües español-inglés y los que siguen enseñanzas convencionales en la región española de Castilla-La Mancha. Con la información obtenida podremos comprobar si los programas bilingües mejoran o no la competencia oral, en lengua inglesa, de los alumnos inmersos en ellos. A su vez, compararemos y valoraremos los resultados hallados dentro del programa bilingüe de Castilla-La Mancha con otros programas desarrollados en diversos territorios españoles y europeos para conocer las diferencias o similitudes existentes.
Numerous studies of written language development have shown minimal improvement over the course of an instructional period, yet these solely text-based studies offer little explanation for the lack of changes in writers’ language because little is reported about the classroom and the participants. The purpose of this study is to use goal theory to better understand students’ instructional contexts and their writing behaviors as they relate to textual features. We focused on six students in two English for academic purposes (EAP) writing classes, in which we conducted observations and teacher interviews. Students filled out an open-ended survey, wrote two essays using keystroke logging software, and participated in stimulated recall sessions. We first evaluated how instructional goals and practices aligned with students’ goals and writing behaviors. We first found that the instructors’ and students’ language-related goals did not completely match. Students had the goal of improving their language, but teachers focused on more global writing issues. In addition, teachers had a goal that students become independent editors of their language, but students missed opportunities to edit. A secondary yet important methodological finding was that the use of timed writing tasks in text-based studies did not allow students to apply explicit grammatical knowledge that the instructors expected the students to have. Pedagogical, research design, and assessments implications are provided.
Full-text available
This paper aims to verify whether blogging helps to improve the quality of EFL written competence within Content and Language Integrated Learning. We conducted an experiment with learners studying geography using English in the 3rd year of secondary education during the academic year 2017-2018. One group of 11 students participated in the treatment group through blogging, while a second group of 11 learners completed the experiment within the control group using Google Docs. In the case of the treatment group, learners completed 5 different written tasks through blogging and, regarding the control group, students completed 5 tasks via Google Docs. Regarding the treatment group, an interactive digital written task is also analysed. The outcomes showed that there was a relevant difference between the treatment group and the control group. This confirms the research hypothesis, which is that the treatment group would achieve better outcomes than the control group.
Full-text available
Content and Language Integrated Learning (CLIL) has become the focus of Foreign Language (FL) research within the last decades. CLIL provides a more complete, naturalistic, and meaningful context to FL learning, which has proven to brings many benefits to learners, such as a higher motivation and promotion of creativity, and better results in receptive skills, vocabulary, morphology and fluency. Nevertheless, most CLIL research has focused on primary and secondary level students and, thus, more research is needed with younger learners, namely, pre-primary students. The present study examines the learning of FL vocabulary in pre-primary learners following a soft-CLIL program, as compared to their same age peers following Formal Instruction (FI) of English. Over the course of six months, pre-primary students of two grades, namely 4- and 5-year-old students (N=155), took part in such program, aiming at teaching two curricular preschool units, traditionally taught in the mother tongue (L1), in English in the FL sessions. A longitudinal study was conducted, and students were administered a general vocabulary level pre-test, as well as a target words receptive vocabulary post-test after the two units had been worked on. The focus of the research was on receptive vocabulary acquisition, but age and word frequency effects were also analyzed. Results showed positive tendencies in receptive vocabulary development through soft-CLIL, although not statistically significant. A significant frequency effect was found, indicating that high-frequency words are recalled more easily than lower-frequency ones, but no significant differences were found when comparing learners from the two grades.
Full-text available
Within the European strategy towards multilingualism, which has guided linguistic policies in recent years, ‘Languages for all!’ is the prevailing motto. This idea has been extremely well captured by one of the central planks in the strategy, the Content and Language Integrated Learning (CLIL) approach at primary and secondary educational levels, which has quickly become popular both top-down and bottom-up. CLIL has its natural continuation at university level, where the preferred term used is Integrating Content and Language in Higher Education (ICLHE). This chapter underlines the instrumental role played by such approaches in the promotion of internationalisation as a position and as a de facto situation. In doing so, it is argued, they are central in meeting the language demands of education, the preparation for mobility being part and parcel of them. CLIL and ICLHE can help students ‘learn to learn’ in multilingual academic settings, and perhaps eventually work in multilingual settings. And for those who do not get to travel, ‘internationalisation at home’ might be a foreseeable objective.
This chapter examines the effects of CLIL provision on different aspects of written language competence in order to determine which of these areas profit more and which are possibly unaffected by the experience of subject matter teaching in a foreign language. For this purpose, we analysed the written work of students who followed either a traditional EFL curriculum or enjoyed additional CLIL provision. The data were obtained in a double case study at two higher technical colleges in Austria where students were asked to complete a free writing task. The bottom line of the analysis shows the CLIL students clearly ahead of their EFL-only peers on the basis of overall scores, but on closer inspection results tell a more complex story: in the area of lexico-grammar the CLIL students show significant advantages throughout, as they do in vocabulary range and orthographic correctness. On the level of discourse competence and textual organization, however, differences are difficult to discern with the general competence level in this area leaving a great deal of room for improvement. Several explanations for these results suggest themselves.