System 51 (2015) 65e76
Contents lists available at ScienceDirect
System
journal homepage: www.elsevier.com/locate/system
Out-of-school digital gameplay and in-school L2 English
vocabulary outcomes*
€m 1
Pia Sundqvist*, Peter Wikstro
Faculty of Arts and Social Sciences, Karlstad University, SE-651 88 Karlstad, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 February 2014
Received in revised form 1 April 2015
Accepted 7 April 2015
Available online
The aim of the present study is to examine the relation between out-of-school digital
gameplay and in-school L2 English vocabulary measures and grading outcomes. Data were
originally collected from a sample of 80 teenage Swedish L2 English learners and comprise
a questionnaire, language diaries, vocabulary tests, assessed essays, and grades. Using an
observational post-hoc design, three Digital Game Groups (DGGs) were created based on
frequency of gameplay: (1) non-gamers (0 h/week), (2) moderate gamers (<5 h/week), and
(3) frequent gamers (5 h/week). Results show that DGG3 had the highest rated essays,
used the most advanced vocabulary in the essays, and had the highest grades, closely
followed by DGG1, while DGG2 trailed behind. For the vocabulary tests, DGG3 was followed by DGG2 and DGG1, indicating that gameplay aligns more directly with vocabulary
test scores than vocabulary indicators drawn from essays. Due to the gender distribution of
non-gamers (predominantly girls) and frequent gamers (exclusively boys), a subsidiary
aim is to investigate how gameplay correlates with outcomes for boys and girls: significant
correlations were found for gameplayevocabulary tests/English grades for the boys.
© 2015 Elsevier Ltd. All rights reserved.
Keywords:
L2 vocabulary acquisition
Informal learning
Incidental learning
Gender
CALL
COTS games
Digital games
1. Introduction
The present study taps into the fields of second language acquisition and computer-assisted language learning (CALL).
More specifically, the main aim is to shed light on the relation between out-of-school digital gameplay activity and second
language (L2) English vocabulary as well as various L2 English grading outcomes. Recently, there have been calls for more
research on digital gaming and language learning (see, e.g., Cornillie, Thorne, & Desmet, 2012).
In contrast to laboratory-based research, Spada (2005, p. 330) has argued for language research “in real classrooms with
real learners.” She argues that such research has great potential in informing classroom practice. In light of the fact that
language learning beyond the language classroom has grown increasingly common (for a collection of papers on this topic,
see Benson & Reinders, 2011b), there is a need for more studies in out-of-school settings e and what Spada refers to as real
learners should be included in such studies. Further, from the perspective of ecological validity, this would be a step in the
right direction (cf. Brewer, 2000). However, in current research, studies on digital gameplay and language learning carried out
exclusively in educational contexts still seem to be more common than studies attempting to combine out-of-school gaming
*
Data used in the present study was originally collected for a doctoral dissertation, published as Sundqvist (2009).
* Corresponding author. Tel.: þ46 54 7001508 (office), þ46 76 8496226 (mobile).
€m).
E-mail addresses: [email protected] (P. Sundqvist), [email protected] (P. Wikstro
1
Tel.: þ46 54 7002178 (office).
http://dx.doi.org/10.1016/j.system.2015.04.001
0346-251X/© 2015 Elsevier Ltd. All rights reserved.
66
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
with in-school language scores (Benson & Reinders, 2011a). In order to at least partially fill this gap in research, the present
paper contributes with findings about teenagers’ time spent on playing commercial off-the-shelf (COTS) games in their spare
time and how this investment in digital gameplay is linked to different measures of L2 English in school.
The focus is on L2 English learners in Sweden in the last year of compulsory school. As for the level of proficiency among
these learners, a passing grade in English would correspond to being an Independent L2 English user (~B1) according to the
Common European Framework of Reference for Languages (or CEFR, Council of Europe, 2001; The Swedish National Agency of
Education, 2012). At the core of each CEFR scale, or level, lies the ability to use L2 vocabulary, which is one reason why several
measures of vocabulary are used here. In our analyses, then, several sets of vocabulary data from tests and essays are used
along with grading outcomes. In what follows, we give a literature review concerning (i) the role of English in the lives of
young people, more specifically among Western and in particular Swedish youth, (ii) digital gameplay, L2 English, and gender,
and (iii) advanced vocabulary use in learner English, before we state our research questions.
2. Literature review
For a long time, formal instruction in school was how most students usually learned languages, but over the last decades
informal language learning in out-of-school contexts has grown increasingly common (Forsman, 2004; Kuppens, 2010;
~ oz & Lindgren, 2011; Olsson, 2011; Sundqvist, 2009; Sylve
n, 2010; Sylve
n & Sundqvist, 2012; Turgut & Irgin, 2009).
Mun
Clearly, opportunities for language learning outside of school have increased much thanks to the multimodality afforded by
technology (see, e.g., Coniam & Wong, 2004 as regards grammar development in online chat; Martinez & Schmitt, 2010; as
regards vocabulary learning). Sundqvist (2009; see also Sundqvist, 2011) proposes an umbrella term, extramural English (EE),
for activities such as watching Anglophone TV-shows or films, playing digital games using English as a lingua franca, listening
to music with lyrics in English, reading books or magazines in English, etc. The present study focuses on the EE activity digital
gameplay.
EE is clearly linked to incidental or naturalistic learning of English. Incidental language learning has been defined by Laufer
and Hulstijn (2001, p. 10; cf. Schmidt, 1994) as “the learning without an intent to learn, or as the learning of one thing, e.g.
vocabulary, when the learner's primary objective is to do something else, e.g. to communicate.” With regard to the relation
between digital gameplay and language learning, vocabulary acquisition would be a by-product of a desire to understand the
game or to communicate with other players. Benson (2011, p. 77) instead uses the term naturalistic language learning for this
type of learning and, when set outside of school, he suggests out-of-school learning. In the present paper, extramural English
and out-of-school learning of English are used interchangeably.
2.1. English as a part of daily life for youth
Access to authentic English input and involvement in productive interactions in English are part of everyday life for youth
growing up in what Kachru (1985) refers to as the expanding circle, that is, countries where English is taught as a foreign
language e but English is not very “foreign” anymore, at least not in all expanding circle countries. In France, for instance,
Sockett and Toffoli (2012, p. 149) describe a situation where learners are involved in English language use daily and where
they learn English “perhaps without ever being enrolled in a formal language course.” The situation is similar in Sweden,
where the prevalence of English makes it possible to claim that English has the status of a second as opposed to a foreign
language (Viberg, 2000). The picture is much the same in Finland (Forsman, 2004), Norway (Simensen, 2010), and Belgium
(Kuppens, 2010), to give only three examples. However, it needs to be emphasized that the same may not be true for youth
growing up in some other countries that are also part of the expanding circle, but where the digital infrastructure is more
limited, such as in Mexico or Malawi (International Telecommunications Union, 2013). Furthermore, there are countries
where English language digital media may be blocked or not as easily accessed as digital resources in major local languages,
such as in China (Schwankert, 2007). Nevertheless, technology has facilitated the informal learning of English in many
expanding circle countries since Kachru published his well-known paper almost 30 years ago.
2.2. Research on digital gameplay, L2 English, and gender
Research that combines gameplay and L2 vocabulary are particularly relevant to this study. In an early study about
incidental vocabulary acquisition, the participants improved in terms of game-specific lexical items thanks to intensive gamework (Cheung & Harrison, 1992). More recently, by adding material to the game The Sims in order to make vocabulary input
more comprehensible for players, two separate studies revealed positive results regarding the participants’ L2 English
vocabulary (Miller & Hegelheimer, 2006; Ranalli, 2008), and considerable L2 vocabulary gains were found in a study involving
intermediate and advanced L2 English students playing EverQuest 2 (Rankin, Gold, & Gooch, 2006). In an experimental study
among Japanese university undergraduates (deHaan, Reed, & Kuwada, 2010), the interactive aspect of playing digital games
was examined by comparing groups whose members either played or watched a music-centered English digital game. The
purpose was to see to what extent interactivity would help or hinder the noticing and recall of L2 vocabulary. Results showed
that both the players and the watchers recalled in-game words, but the players significantly less so than the watchers. The
former group perceived the game and the English to be much more difficult than the latter group did, a finding supposedly
connected with the cognitive load induced by game interactivity, and both groups forgot significant amounts of words over
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
67
the course of the two-week study. These findings suggest that a great deal of time probably needs to be invested in gameplay,
if L2 words are to be stored in long term memory. In an Iranian study (Aghlara & Hadidi Tamjid, 2011) among primary school
learners with no prior knowledge of English, an experimental group learned vocabulary through a digital game (SHAIEx) and a
control group via traditional teaching methods; the experimental group significantly outperformed the control at the end of a
s playing games in English. They
45-day treatment. Turgut and Irgin (2009) investigated Turkish children at Internet cafe
argue that the games allowed the participants to incidentally make use of vocabulary for their own purposes and in complex,
pleasurable ways, which would promote learning. In a school setting, Cobb and Horst (2011) examined young Francophone L2
English learners in Canada who were instructed to play a mini-game. The researchers found vocabulary gains and increased
speed of lexical access. Notably, they conclude that longer periods of gameplay were necessary to consolidate learning.
However, it should be noted that this study was set in school and did not involve COTS games. Finally, in a case study about
transcultural communication in World of Warcraft, Thorne (2008) concludes that there were both frequent and highly
meaningful linguistic activities between the players, including discussions about lexical items.
In the Swedish context, Sundqvist (2009) found significant correlations between learners' total amount of EE and
vocabulary size. Furthermore, she concludes that of all EE activities investigated in the study, extramural digital gaming along
with using the Internet were the most important activities for vocabulary size. She also found that the boys spent significantly
more time on these two activities than the girls. However, it was not within the scope of that study to examine the relation
between gameplay and vocabulary in any detail; thus, the present study can be viewed as a follow-up. In another project
n and Sundqvist (2012) saw that frequent gamers (self-report data; 5 h/week)
about young learners in 5th grade, Sylve
scored significantly higher on a test of vocabulary than moderate gamers, who in turn scored higher than non-gamers. This
pattern was repeated in listening and reading comprehension tests. Moreover, in a third Swedish study, Olsson (2011)
examined EE and 9th-graders’ writing of letters and news articles. Her findings corroborate Sundqvist (2009) in that the
group of learners who had the most frequent EE contacts (mainly boys and many via digital games) had the highest scores for
the writing tasks and also the highest final grades. Finally, in an international study among 15- and 16-year-olds in 14 European countries, the Swedish participants showed top scores for English proficiency. Scores were compared with the CEFR
and a great deal of the students performed at the B2 level and some even as high as C1; heavy involvement in EE activities is
provided as a main explanation for the outcomes among the Swedish learners (The Swedish National Agency of Education,
2012).
In sum, previous research has established links between playing digital games and L2 English, but there is more to
investigate pertaining to the specifics of this relation. For instance, researchers have suggested that the affordances of some
types of digital games, especially massively multiplayer online role-playing games (MMORPGs), may be particularly beneficial
for L2 learning (see, e.g., Kuppens, 2010; Peterson, 2012; Reinders & Wattana, 2011). There is also recent research providing at
least partial empirical evidence for such a claim, in particular when time spent playing MMORPGs is compared with time
spent playing single-player games (Sundqvist, 2013). Also as regards the specifics of gamingeL2 learning, game-based
research has revealed significant gender-related differences that merit more attention from CALL-researchers. For
instance, boys tend to play more frequently and for longer stretches of time than girls do and, further, boys and girls tend to
have different game preferences (for the US, see, e.g., Lenhart et al., 2008; Lucas & Sherry, 2004; for Sweden, see, e.g., The
Swedish Media Council, 2010). It is beyond the scope of this paper to examine in detail the possible role of game genres
for L2 vocabulary, but gender in relation to gameplay and L2 English will be touched upon.
Interestingly, boys and girls also tend to have different preferences as regards how foreign languages should be taught and
learned in school, which can be explained by powerful stereotypical narratives about what boys and girls respectively are
‘good’ at (Carr & Pauwels, 2006). An example related to the topic of this paper is that boys more than girls prefer “using
computers” and “escaping (…) the normal classroom space” (p. 80). However, as for the game-related gender differences, it
must be acknowledged that female gaming worldwide is “variable enough to suggest that gender is not a reliable predictor of
gaming habits” (Carr, 2005, p. 465), making potential connections between gameplay, gender, and L2 vocabulary complex to
investigate. Contributing to the complexity is the issue of sexism within certain gaming contexts (see, e.g., Dietz, 1998; Jansz &
Martis, 2003), where females might face harassment (Sarkeesian, 2012, December 4), something which undoubtedly has
implications for their playing habits and possible L2 learning. There is still little research on achieved language learning
outcomes due to relations between gender and habits of out-of-school gameplay or, as phrased by Higgins (2010, p. 370), of
“how gendered identities relate to language learning.”
3. Advanced vocabulary in learner English
There are many ways to measure learner vocabulary. One measure is lexical sophistication, commonly defined as the
percentage of advanced words in a text, where “advanced” would depend on the levels of the learners tested (Laufer, 1991). In
general, the term lexical richness is used in reference to the quality of the lexis learners use, and measures of lexical richness
thus attempt to quantify the degree to which learners use a varied and large vocabulary (Laufer & Nation, 1995, p. 307).
Further, the results of Daller and Phelan (2007, pp. 234e235) suggest that in “an educational setting the advanced parts of the
lexicon play a crucial role in the construct of foreign language proficiency.”
As pointed out by Milton (2009, p. 125), there is little agreement within the research community as regards which
methods might be considered most suitable when measuring vocabulary knowledge. However, one fairly straightforward
approach to measuring advanced vocabulary use originates from Zipf's law, which states that word length is inversely
68
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
n, 2004, p. 202). By adopting such an approach, use of
proportional to frequency of usage (Malvern, Richards, Chipere, & Dura
long words in writing would thus be indicative of learners' vocabulary size, since long words are generally more common in
peripheral than core vocabulary, and use of long words has indeed been shown to correlate with proficiency (Grant & Ginther,
2000; Reid, 1986, 1990). Different definitions of word length are found, such as ‘mean number of orthographic letters per
word’ (Biber, 1988) and ‘three syllables or more’ (Sundqvist, 2009); the common denominator is that learners who are able to
produce long words show mastery of advanced vocabulary. The present study uses polysyllabic words of three syllables or
more as a unit of measure for advanced vocabulary. While not all polysyllabic words are advanced, frequency studies of the
English lexicon show that the majority of words that are part of basic vocabulary is monosyllabic and the majority of words
that are part of advanced vocabulary is polysyllabic (Minkova & Stockwell, 2006; see also Nation, 2001).
In a study about teacher ratings of English learner essays and aspects of lexical richness, Daller and Phelan (2007) found a
high correlation between overall rating and advanced vocabulary, defined as words beyond the 2000 word level, enabling
them to conclude that reliable overall rating was “closely connected to the occurrence of low-frequency, rare words” (p. 244;
the 2000 level has been used before by others, e.g., Laufer, 1995; West, 1953). This finding clearly indicates that teachers are
sensitive to learners’ use of advanced vocabulary. Daller and Phelan close their article by stressing the need for further
research on proficiency and lexical richness in settings outside the English classroom. Finally, with regard to the length of
learner essays, several studies show that for the same task, more-proficient learners tend to write longer texts than their lessproficient peers (see, e.g., Ferris, 1994; Reid, 1986, 1990).
4. Research questions
The present study poses two main research questions concerning the relation between digital gameplay and L2 English
vocabulary on the one hand and various grading outcomes on the other. We use a number of variables as measures of L2
English vocabulary, namely two vocabulary tests e the Productive Levels Test and the Vocabulary Levels Test (Laufer & Nation,
1995; Nation, 2001) e and five quantitative variables derived from free written essays: overall tokens, overall types, polysyllabic tokens, polysyllabic types, and “own” polysyllabic types (see Methods). Further, the grading outcomes comprise the
learners’ grade for the above mentioned essay, their English grade upon finishing 8th grade, their English final grade (one year
later, upon finishing 9th grade), and, finally, their overall grade (the sum of grades of all school subjects). The main research
questions are:
1. To what extent is there a positive relation between digital gameplay and the L2 English vocabulary measures?
2. To what extent is there a positive relation between digital gameplay and the grading outcomes?
To answer the questions, based on available data about gameplay habits, three Digital Game Groups were created post hoc
in order to allow for comparisons between learners who do not play digital games at all, learners who play some, and learners
who play extensively. We discovered a gender distribution of non-gamers (predominantly girls) and frequent gamers
(exclusively boys; the in-between moderate gamers were a mixed group). Thus, the three groups are also partially defined by
gender. For this reason, we added analyses of correlation between time spent playing digital games (hours per week) and the
target variables (that is, the vocabulary measures and grading outcomes) for cases selected by gender, that is, for only boys
and only girls. Accordingly, we pose a subsidiary third research question:
3. To what extent does digital gameplay correlate with the L2 English vocabulary measures and the grading outcomes for
boys and girls respectively?
5. Material and methods
5.1. Participants
Participants were part of Sundqvist (2009), a study investigating the impact of EE on Swedish 9th-grade L2 English
learners’ oral proficiency and vocabulary (age 15e16; N ¼ 80; 36 boys, 44 girls). Written forms of consent were collected from
the guardians prior to the start of the study, which ran for a school year, 2006e2007. Six participants dropped out during the
year, explaining the differing totals below.
5.2. Material
Empirical data used in the present study include a questionnaire, language diaries, vocabulary tests, assessed essays, and
final grades. The questionnaire was distributed in August (first week of school), included 30 questions, and yielded information about background variables (L1, experience of travels abroad, computer/Internet access etc.) and various EE activities
(for details, see Sundqvist, 2009, pp. 231e238). Two sets of one-week language diaries were used to collect self-report data
about exposure to and involvement in eight predetermined EE activities: reading books, reading magazines, watching TV,
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
69
watching films, using the Internet, playing digital games, listening to music, and “other” (see Sundqvist, 2009, p. 239). In the
language diaries, participants were also instructed to contribute with the same type of information for extramural activities in
Swedish and “Other language(s)”. The first diary was filled out during a week in September and the second in March/April;
mean values (hours/week) for digital gameplay are used in the analyses below. The diaries were to be filled out daily in the
homes by the learners, who reported times and titles (of books, TV-programs, games, etc.) for each activity and day. They were
reminded about the diary daily in school by their teacher and at home by their guardians, who had been carefully informed
about the diary (orally and in writing) by the researcher. The diary was developed by Sundqvist in collaboration with Liss
n, University of Gothenburg, and was used by the former in her PhD study and by the latter in another project
Kerstin Sylve
n, 2006).
(Sylve
Regarding the vocabulary tests, shortened versions of the Productive and Vocabulary Levels Tests were used (Laufer &
Nation, 1995; Nation, 2001). The Productive Levels Test (PLT) measures productive vocabulary, requiring test-takers to produce the target word with the help of some initial letters and a context sentence: “He was riding a bi……..” (bicycle); the PLT
was taken in September. The Vocabulary Levels Test (VLT) measures receptive vocabulary (recognition knowledge) and each
question consists of words in groups of three accompanied by six distractors; test-takers are supposed to pair each of the
three words with the most accurate distractor (see http://www.lextutor.ca/tests/levels/recognition/2_10k/); the VLT was
taken in April.
The essays were written and assessed as part of the mandatory Swedish national test in English in March, and the
essay grades were collected as part of, but never used in, Sundqvist (2009). The essays were graded by the participants’
own English teacher, who is obliged to follow test administration, assessment and grading instructions/guidelines from
the national agency for education; thus, the test is centrally administered but locally assessed and graded e external
raters are not used. Each year, the scores/grades of all 9th graders are reported to the national agency for education, the
institution officially in charge of the quality and evaluation of all national tests used in Swedish schools. In addition to
these reports, a random selection of complete English tests (that is, not only the scores/grades, but also hard copies of the
essays as well as two comprehension tests) are sent to the test constructors for further examination as regards, for
instance, reliability and validity (for more information, see http://www.nafs.gu.se/english). It should be added that essay
topics are carefully selected through extensive piloting by the test constructors. The students had 80 min at their
disposal to compose their text, and they were to choose between two topics: a letter to a summer camp or an essay about
making the world a better place. No dictionaries were allowed. As mentioned, the essays ended up not being used in
Sundqvist (2009), but constitute a major source of data for this study. Moreover, the final English grades were collected
at the end of the school year, in June. Finally, as part of the present study, the English grades from one year before (the
end of 8th grade) were collected from official records, in order to provide a starting point and to allow for comparisons
over time.
The essays were transcribed, with problematic passages checked by a second transcriber. The Compleat Lextutor
VocabProfile tool was used to generate an automatic count of tokens and types. Since spelling errors, some Swedish words,
and occasional minor artefacts such as smiley faces and peace symbols were kept and represented in the transcripts, the type
counts do not represent English words exclusively. However, this was deemed to be a relatively insubstantial validity concern.
Further, advanced vocabulary was assessed using the frequency of polysyllabic types (defined as words of three or more
syllables) as an indicator (Grant & Ginther, 2000; Reid, 1986, 1990). Polysyllabic items were counted manually, using essay
wordlists generated by AntConc 3.2.4w (concordancing software). Some polysyllabic words occurring in the instructions for
the essay (24 types) were frequently repeated by the students in their texts. To control for this, the essay wordlists were
compared with the list of polysyllables from the instructions, in order to generate data about each student's own polysyllabic
types e that is, types that the student assuredly did not copy from the instructions.
5.3. Design
Considering the findings from previous gaming and L2 English studies reported on above, it is reasonable to hypothesize
that there is a positive relation between digital gameplay activity and L2 English vocabulary in our sample. Therefore, as
mentioned, three Digital Game Groups (DGGs) were formed based on diary data for the number of hours spent playing digital
games per week (see Table 1).
The division into three groups, as opposed to for instance four or five groups, is somewhat arbitrary, but justified on a
few main grounds. The category of participants who spent zero hours per week gaming constitutes a fairly “natural” group
Table 1
The three digital game groups.
Digital game group
DGG1 (non-gamers)
DGG2 (moderate gamers)
DGG3 (frequent gamers)
Total
n (%)
35
26
19
80
(44%)
(32%)
(24%)
(100%)
Time interval (hours/week)
From
To
0
>0
5
0
0
<5
42
42
Mean (hours/week)
SD
0
2.0
13.9
3.9
0
1.5
9.5
7.3
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
70
Table 2
Gender distribution across the digital game groups.
Digital game group
Total n (%)
Boys n (%)
Girls n (%)
DGG1 (non-gamers)
DGG2 (moderate gamers)
DGG3 (frequent gamers)
Total
35
26
19
80
4
14
18
36
31
12
1
44
(44%)
(33%)
(24%)
(100%)
(11%)
(39%)
(50%)
(100%)
(71%)
(27%)
(2%)
(100%)
of its own. Further, dividing the remainder of the sample into subgroups is necessary since it is inappropriate to identify
those who spend a small handful of hours gaming and those who spend upwards of 40 h per week equally as “gamers”,
especially in light of previous findings that long periods of gameplay are necessary for learning (Cobb & Horst, 2011). Two
subgroups of moderate and frequent gamers were considered appropriate for the purposes of allowing a finer grained
statistical analysis without generating group sizes too small to allow for meaningful statistics. Five hours per week was
settled on as a reasonable cutoff point due to the distribution of the diary data and its correspondence to the distribution of
answers to a questionnaire item about frequency of gaming. The cutoff also aligns well with Cobb and Horst (2011), where
(younger) frequent gamers averaged 14 h per week and infrequent gamers about 3; a five-hour cutoff was also used in
n and Sundqvist (2012). Finally, the group design has pedagogical merits among others,
Bytheway (2011) as well as in Sylve
but the between-group analyses are also complemented with linear regression in the results pertaining to the third
research question.
The DGGs were generated solely based on frequency of gaming, but as can be seen in Table 2, the groups turned out also to
be partially defined by gender, with non-gamers being predominantly female and frequent gamers exclusively male
(crosstabulation of gender and DGGs: c2 35.750, df ¼ 2, p ¼ .000, 4c ¼ .668; see Analytic Procedure). The sole girl in DGG3 fell
out during the course of the study, limiting data collected from her to the questionnaire, the diary, and the PLT. For all other
measures, DGG3 represents boys only.
5.4. Analytic procedure
All tests were run in IBM SPSS Statistics 20. We keep to the convention of regarding p < .05 as significant, however, we
r's V (4c) were used for tests of
report exact p-values since they facilitate interpretation. Pearson's chi-squared (c2) and Crame
association between nominal variables, while one-way analysis of variance (ANOVA) together with classical eta squared (h2)
were used to calculate significance and effect sizes for tests with numeric variables. Cohen's conventions for interpreting
r's V),
effect sizes were used (see Aron, Aron, & Coups, 2005). In line with Cohen's convention for d (also often used for Crame
4c ¼ .2 is a small effect size, 4c ¼ .5 is medium, and 4c ¼ .8 is large (Aron et al., 2005, p. 192). In line with Cohen's convention
€ rnyei, 2007, p. 221).
for r2, h2 ¼ .01 is a small effect size, h2 ¼ .06 is medium, and h2 ¼ .14 is large (Do
For ANOVA, Gabriel's post-hoc test was used to provide additional indications of which groups differed from which within
the general between-groups differences. Gabriel's test was preferred over, for example, SeNeK, since it avoids conservative
bias resulting from somewhat unevenly-sized groups, while still being conservative in correcting for multiple comparisons.
Paired samples t test was used was used to analyze progression over time, from the end of 8th to the end of 9th grade, as
regards the grade in English. Finally, Spearman's rank order correlation coefficient (rs) was used in correlation analyses (linear
regression) and preferred over Pearson due to the non-normal distribution of the digital gameplay variable.
5.5. Background variables controlled for
While our third research question involves gender as a background variable, diary and questionnaire data also enabled
examination/tests of other background variables. Thus, in total we examined eight such variables from the perspective of the
DGGs; none of them co-varied significantly with indices of vocabulary or outcomes:
1)
2)
3)
4)
5)
6)
7)
8)
Experience of travel abroad (outside of Scandinavia)
Speaking English regularly (with family, friends, relatives, etc.)
Reading habits (in Swedish and English)
Television viewing habits
Film/Movie watching habits
Music listening habits
Parental level of education (maternal, paternal)
Family cultural capital (indicated by number of books in the home)
Nevertheless, it is conceivable that other factors, unrecorded in the present dataset, could intrude into the present
findings. This possibility is always a consideration for social scientific research and should be kept in mind in the interpretation of the analyses below.
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
71
6. Results
6.1. Digital gameplay and the L2 English vocabulary measures
6.1.1. Two vocabulary tests
An internal consistency measure showed that the PLT and VLT were reliable (r ¼ .843, p ¼ .000; Cronbach's a ¼ .832). The
results of the tests are reported in Table 3. The frequent gamers (DGG3) scored the highest on the PLT, followed by the
moderate gamers (DGG2) and the non-gamers (DGG1). Analysis of variance indicated that there was a significant difference of
mean scores among the three groups (ANOVA: F (2, 77) ¼ 12.08, p ¼ .000). Gabriel's post-hoc test showed that the score for
DGG3 was different from the scores of the other two groups (p ¼ .000), but that DGG1 and DGG2 were indistinguishable from
one another. As for the VLT, the pattern was repeated, with DGG3 scoring the highest, followed by DGG2 and DGG1. Again,
ANOVA revealed significant differences with regard to the mean scores (F (2, 75) ¼ 10.85, p ¼ .000). Similarly, Gabriel's posthoc test showed that the score for DGG3 was different from the scores of the other two groups (p < .005); scores for DGG1 and
DGG2 were indistinguishable. Effect size measures showed a large effect for both tests.
6.1.2. Vocabulary use in essays
The second set of vocabulary data was drawn from the essays; the results are presented in Table 4. As mentioned, polysyllabicity was used as the indicator of advanced vocabulary.
The overall number of tokens illustrates mean essay length. Table 4 shows that the longest essays were written by the nongamers (363.4 words on average). In contrast, the shortest essays were found among the moderate gamers, who wrote
slightly fewer than 300 words on average. In between are the frequent gamers with a mean of almost 330 words. However,
ANOVA revealed that these differences were not statistically significant (F (2, 74) ¼ 2.33, p ¼ .104).
As for overall types, DGG3 had the highest mean (155.2), followed by DGG1 (153.3), again with DGG2 trailing behind
(130.0). In contrast to the vocabulary test results, the association gameplayeoutcomes is not linear. The differences between
the groups were statistically significant (ANOVA: F (2, 74) ¼ 3.54, p ¼ .034). Gabriel's post-hoc test revealed that the score for
DGG2 was different from the scores of the other groups (p ¼ .057), but that groups DGG1 and DGG3 were indistinguishable
from one another.
The examination of the use of advanced vocabulary is reported with regard to polysyllabic tokens, types, and the learners'
own polysyllabic types (Table 4). For all three sets of measures, DGG3 scored the highest, followed by DGG1 and DGG2. The
differences between the groups did not reach statistical significance with regard to polysyllabic tokens (p ¼ .219). However,
for both the polysyllabic types and the participants' own polysyllabic types, significance was reached. First, for polysyllabic
Table 3
Productive levels and vocabulary levels test scores across the DGGs.
Productive levels test (Max: 45)
Mean
SD
n
Vocabulary levels test (Max: 90)
Mean
SD
n
DGG1
DGG2
DGG3
Tot
Significance/Effect size
13.7
6.3
35
14.4
7.2
26
22.7
7.2
19
16.1
7.7
80
p ¼ .000
h2 ¼ .239
55.2
12.9
34
58.4
13.2
26
71.6
9.2
18
60.1
13.7
78
p ¼ .000
h2 ¼ .224
Table 4
Essay vocabulary token and type counts across the DGGs.
Overall tokens
Mean
SD
Overall types
Mean
SD
Polysyllabic tokens
Mean
SD
Polysyllabic types
Mean
SD
Own polysyllabic types
Mean
SD
DGG1 (n ¼ 34)
DGG2 (n ¼ 25)
DGG3 (n ¼ 18)
Tot (N ¼ 77)
Significance/Effect size
363.4
115.5
298.0
133.0
326.7
88.2
333.6
118.1
p ¼ .104
h2 ¼ .059
153.3
34.5
130.0
42.4
155.2
34.0
146.2
38.4
p ¼ .034
h2 ¼ .087
17.2
9.5
14.1
11.1
19.7
10.9
16.8
10.4
p ¼ .219
h2 ¼ .040
12.2
6.3
9.5
6.4
15.4
9.0
12.0
7.3
p ¼ .029
h2 ¼ .091
9.1
5.3
6.6
5.2
11.6
8.9
8.9
6.5
p ¼ .041
h2 ¼ .083
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
72
Table 5
Grading outcomes across the DGGs.
DGG1
9th grade,
Mean
SD
n
9th grade,
Mean
SD
n
8th grade,
Mean
SD
n
9th grade,
Mean
SD
n
DGG2
DGG3
Tot
Significance/Effect size
Essay grade (1e10)
p ¼ .035
6.03
2.107
34
5.31
1.960
26
6.94
1.939
18
6.00
2.086
78
2.74
.701
35
2.58
.758
26
3.06
.725
18
2.76
.738
79
p ¼ .104
h2 ¼ .058
2.46
.505
35
2.42
.578
26
2.83
.618
18
2.53
.574
79
p ¼ .037
h2 ¼ .083
h2 ¼ .085
English final grade (1e4)
English final grade (1e4)
Overall grade (0e320)
217.3
67.9
35
216.4
63.9
26
218.6
51.4
18
217.3
62.4
79
p ¼ .993
n/a
types, ANOVA indicated different uses amongst learners in the three groups (F (2, 74) ¼ 3.71, p ¼ .029). Gabriel's post-hoc test
showed that the scores of DGG3 differed from those of DGG2 (p ¼ .023), but not from those of DGG1 (p ¼ .300). The scores for
polysyllabic types were indistinguishable between DGG1 and DGG2. Second, for the differences regarding learners' own
polysyllabic types (ANOVA: F (2, 74) ¼ 3.34, p ¼ .041), once more DGG3 differed from DGG2 (p ¼ .035), whereas no significant
difference was found between DGG3 and DGG1, nor between DGG2 and DGG1. Finally, effect sizes were slightly below
medium for polysyllabic tokens (h2 ¼ .040) and medium for polysyllabic types (h2 ¼ .091) as well as for learners' own
polysyllabic types (h2 ¼ .083).
6.2. Digital gameplay and the grading outcomes
There are four sets of grading outcomes in the material: (i) learners' essay grade, (ii) learners' final grade in English (9th
grade), (iii) learners' grade in English from 8th grade, and (iv) learners’ overall grade (all school subjects) upon leaving school
(see Table 5).
An overview of the various grading outcomes reveals a similar picture for three of the grades. For the essay grade and the
two grades in English (8th and 9th grade), DGG3 had the highest means, followed by DGG1 and DGG2. Out of these three
variables, significant differences were found for the essay (p ¼ .035) and the English grade from 8th grade (p ¼ .037). For the
essay, Gabriel's post-hoc test indicated one significant between-group difference, DGG3 as compared with DDG2 (p ¼ .029).
For the English grade (8th grade), on the other hand, no particular between-group difference was significant. Regarding the
overall grade, all three groups had similar means.
Further, for both grades in English, the paired samples t test revealed that the mean grade improved significantly at sample
level over the period of 12 months (p ¼ .000); that is, from the end of 8th grade to the end of 9th grade (see Table 5). As for the
DGGs, DGG 1 and DGG3 improved significantly (p ¼ .006 and p ¼ .042 respectively), whereas DGG2 only had a small, nonsignificant, improvement (p ¼ .103).
6.3. Correlations between digital gameplay and L2 English vocabulary measures/grading outcomes, for boys and girls
Table 6 shows the results of correlations between digital gameplay and L2 English vocabulary measures for the total
sample as well as for boys and girls. Table 7 reveals the corresponding results for the grading outcomes.
For most of the measures, there were no correlations between gaming and measures/outcomes in the total sample.
However, for both vocabulary tests, there were indeed statistically significant correlations at sample level, though these were
more pronounced for the boys in particular (PLT: rs ¼ .532, p ¼ .001; VLT: rs ¼ .540, p ¼ .001). The correlations for the girls
alone on these two tests were non-significant. Regarding the essay vocabulary measures, no significant correlations were
found. Nevertheless, there is a tendency (although not reaching the .05 threshold for significance) for gameplay to correlate
with overall types, polysyllabic tokens, and polysyllabic types for the boys, while nothing of the kind holds for the girls.
For the grading outcomes (Table 7), there were positive and statistically significant correlations between gameplay and
grades for the boys regarding the three outcomes that specifically have to do with English, but not for the overall grade. In
contrast, for the girls and the total sample, there were no significant correlations.
7. Discussion
First, using DGGs turned out to be a fruitful way of examining L2 English data. The present study yields some highly
relevant results related to the relation between out-of-school digital gaming behavior, or lack thereof, and results as
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
73
Table 6
Correlations by gender: digital gameplay and L2 English proficiency measures.
Proficiency measure
Productive levels test (Max: 45)
Vocabulary levels test (Max: 90)
Overall tokens
Overall types
Polysyllabic tokens
Polysyllabic types
Own polysyllabic types
Tot
Boys
.358***
.001
rs
p
N
rs
p
N
rs
p
N
rs
p
N
rs
p
N
rs
p
N
rs
p
N
80
36
.429***
.000
78
.182
.114
77
.072
.534
77
.018
.874
77
.009
.935
77
.041
.725
77
Girls
.532***
.001
.080
.607
44
.540***
.001
36
.159
.314
42
.141
.418
35
.326
.056
35
.311
.069
35
.285
.096
35
.162
.351
35
.061
.703
42
.022
.888
42
.044
.783
42
.095
.551
42
.152
.338
42
Note. * alpha ¼ .05; ** alpha ¼ .01; *** alpha ¼ .001.
Table 7
Correlations by gender: digital gameplay and grading outcomes.
Grading outcome
9th grade, Essay grade (1e10)
9th grade, English final grade (1e4)
8th grade, English final grade (1e4)
9th grade, Overall grade (0e320)
Tot
rs
p
N
rs
p
N
rs
p
N
rs
p
N
Boys
.068
.555
78
.537***
.001
36
.104
.362
79
.442**
.007
36
.165
.145
79
.072
.528
79
.438**
.008
36
.230
.177
36
Girls
.030
.853
42
.125
.424
43
.142
.365
43
.118
.450
43
Note. * alpha ¼ .05; ** alpha ¼ .01; *** alpha ¼ .001.
measured by different tests and grades in school. While the study design does not allow for any claims of causality, there is a
clear pattern: on several measures of L2 English vocabulary, the frequent gamers (DGG3) scored highest followed by the nongamers (DGG1) and the moderate gamers (DGG2). Evidently, frequent gamers show the strongest results for all vocabulary
measures. It is particularly noteworthy that they have the highest mean scores of polysyllabic types as well as own polysyllabic types in the essay, which shows that these learners are able to produce advanced vocabulary in a high-stake
examination. Of the examined groups, DGG3 has the best results for lexical richness.
Interestingly, although DGG3 also received the highest essay grade when the three DGGs were compared (thereby
also corroborating findings in Olsson, 2011), they wrote shorter essays than DGG1, even though it should be noted that
this difference was non-significant. Text length has previously been found to be a useful indicator of L2 writing proficiency (Ferris, 1994; Reid, 1986, 1990); the results of this study e although limited to examining only lexical aspects of L2
writing e suggest that other measures, particularly the use of advanced vocabulary, may also function as a valuable
indicator of L2 writing proficiency. It is possible that the teachers in this study valued learners’ use of advanced
vocabulary highly when deciding upon an overall rating for a text, which would be in line with previous research (Daller
& Phelan, 2007). While DGG1 wrote the longest texts, DGG3 had the most advanced vocabulary and, also, received the
highest essay grades.
Since all three DGGs were on par with one another in terms of the overall grades (p ¼ .993), it is noteworthy that the
learners in DGG3 in 9th grade were awarded higher English final grades as compared with the learners in DGG1 and DGG2
(p ¼ .104; h2 ¼ .058). This latter between-group comparison was non-significant but, nevertheless, the medium effect size
suggests that gameplay may be beneficial for English. This suggestion is at least backed up by the positive correlations found
between gameplay and vocabulary measures/grading outcomes for the boys, who played digital games to a much larger
extent than the girls. In short, it seems likely that the boys benefitted from their investment in gameplay time.
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
74
Moreover, the results corroborate previous studies on the positive relation between digital gameplay and L2 English
vocabulary acquisition (Aghlara & Hadidi Tamjid, 2011; Cheung & Harrison, 1992; Cobb & Horst, 2011; Miller & Hegelheimer,
2006; Ranalli, 2008; Rankin et al., 2006). That the frequent gamers, who had the highest vocabulary test scores, were followed
by the moderate gamers rather than the non-gamers is an important finding pertaining to this specific relation. Phrased
differently, digital gameplay aligns more directly with L2 English vocabulary test scores than it does with the vocabulary
indicators drawn from essays. It is important to bear in mind, though, that the types of games played were not investigated in
this study. Also, as indicated by the large standard deviation for DGG3 (9.5), gaming patterns in this group clearly vary. For
example, gamers close to the low end of the scale (5 h/week) are presumably less likely to play games that require extensive/
repeated stretches of game time.
As regards the grades in English and progression over time, there was only negligible improvement for DGG2. In contrast,
both DGG1 and DGG3 improved significantly. They were at different levels at the end of grade 8, and also at different levels a
year later; both groups developed in a similar fashion, but from different starting-points. As for the two groups that included
gamers, DGG2 and DGG3, if gameplay plays a role in L2 English learning, the results indicate that there seems to be a threshold
for a possible gaming effect, since the moderate gamers did not improve significantly.
8. Conclusion
A recurrent pattern in our findings is that the frequent gamer and non-gamer groups both perform well. In addition,
gameplay correlates with vocabulary and grading outcomes for the boys, but not for the girls (for the simple reason that much
fewer girls were gamers). Possibly, these findings together may be seen to conform to a common stereotype of, on the one
hand, studious girls who do not play digital games and, on the other hand, ‘nerdy’ boys who enjoy doing so (cf. J. Carr &
Pauwels, 2006). However, such an interpretation is certainly not directly supported by the present data. Criticism against
simplistic descriptions of gamer identity, and considerations of the role of gender in gaming communities, should be borne in
mind (D. Carr, 2005; Dietz, 1998; Jansz & Martis, 2003). Nevertheless, the present findings indicate a positive relation between
gameplay and L2 English e at least for boys. Many of the boys from Australia, England, Wales, New Zealand, and Scotland in
Carr and Pauwel's (2006) study said that they prefer to learn foreign languages in game-like conditions, dissimilar to
traditional classroom teaching practice; the Swedish boys in our study seem to be doing just that.
Still, there are a number of limitations to this study. For instance, the number of participants in DGG3 was rather small
(N ¼ 19). Another limitation is the fact that the study design does not allow for any conclusions as regards causality. Thus, even
though the results describe the relation between out-of-school digital gameplay and L2 English vocabulary and grading
outcomes respectively, these findings should not be taken as conclusive but rather as indicative. Despite these reservations,
the study makes an important contribution to the field and, clearly, there is a need for more studies on the topic.
The fact that many (but certainly not all) youth in expanding circle countries such as Sweden devote a great deal of their
spare time to digital gameplay has implications for the teaching of English in those countries. As all teachers know, in order
for teaching to be successful, it is crucial to meet the needs and interests of individual learners and one way to do so would be
to incorporate a systematic use of language diaries (or something similar) to map learners’ out-of-school language-related
contacts and based on the given information, design or recommend suitable learning tasks. Such an approach would, most
likely, facilitate the fostering of learner autonomy (cf. Benson, 2011) in line with the CEFR (Council of Europe, 2001).
For future research, a more ecologically valid study where individual learners are observed while playing COTS games in
their homes is recommended; simultaneous collection of data from in-game interaction should be made. Moreover, there is
also a need for an experimental study with pre- and posttest design by which causality may perhaps be established. Further,
the present study did not take types of games played into consideration. Conceivably, types of games played may be at least as
important as time spent playing and is, therefore, a suitable area for future inquiry. Finally, the findings pertaining to gender
warrant more research into the complex interrelationship between digital gameplay, aspects of L2 English, and gender.
Acknowledgments
We would like to thank the three anonymous reviewers for valuable comments on previous drafts of this article.
References
Aghlara, L., & Hadidi Tamjid, N. (2011). The effect of digital games on Iranian children's vocabulary retention in foreign language acquisition. Procedia e
Social and Behavioral Sciences, 29, 552e560.
Aron, A., Aron, E. N., & Coups, E. J. (2005). Statistics for the behavioral and social sciences: A brief course (3 ed.). London: Prentice Hall International.
Benson, P. (2011). Teaching and researching autonomy (2 ed.). Harlow: Pearson Education.
Benson, P., & Reinders, H. (2011a). Introduction. In P. Benson, & H. Reinders (Eds.), Beyond the language classroom: The theory and practice of informal language
learning and teaching (pp. 1e6). Basingstoke: Palgrave Macmillan.
Benson, P., & Reinders, H. (2011b). Beyond the language classroom: The theory and practice of informal language learning and teaching. Basingstoke: Palgrave
Macmillan.
Biber, D. (1988). Variation across speech and writing. Cambridge: Cambridge University Press.
Brewer, M. B. (2000). Research design and issues of validity. In H. T. Reis, & C. M. Judd (Eds.), Handbook of research methods in social and personality
psychology (pp. 3e16). Cambridge: Cambridge University Press.
Bytheway, J. A. (2011). Vocabulary learning strategies in massively multiplayer online role-playing games. Wellington: Victoria University of Wellington. MA.
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
75
Carr, D. (2005). Context, gaming pleasures and gendered preferences. Stimulation and Gaming, 36(4), 464e482. http://dx.doi.org/10.1177/
1046878105282160.
Carr, J., & Pauwels, A. (2006). Boys and foreign language learning: Real boys don't do languages. Basingstoke: Palgrave Macmillan.
Cheung, A., & Harrison, C. (1992). Microcomputer adventure games and second language acquisition: a study of Hong Kong tertiary students. In M. C.
Pennington, & V. Stevens (Eds.), Computers in applied linguistics (pp. 155e178). Clevedon: Multilingual Matters.
Cobb, T., & Horst, M. (2011). Does word Coach coach words? CALICO Journal, 28(3), 639e661.
Coniam, D., & Wong, R. (2004). Internet Relay Chat as a tool in the autonomous development of ESL learners' English language ability: an explorative study.
System, 32(3), 321e335.
Cornillie, F., Thorne, S. L., & Desmet, P. (2012). ReCALL special issue: digital games for language learning: challenges and opportunities. ReCALL, 24(3),
243e256. http://dx.doi.org/10.1017/S0958344012000134.
Council of Europe. (2001). Common European framework of reference for languages: Learning, teaching, assessment. Cambridge: Cambridge University Press.
Daller, H., & Phelan, D. (2007). What is in a teacher's mind? Teacher ratings of EFL essays and different aspects of lexical richness. In H. Daller, J. Milton, & J.
Treffers-Daller (Eds.), Modelling and assessing vocabulary knowledge (pp. 234e244). Cambridge: Cambridge University Press.
Dietz, T. L. (1998). An examination of violence and gender role portrayals in video games: implications for gender socialization and aggressive behavior. Sex
Roles, 38(5/6), 425e442.
€ rnyei, Z. (2007). Research methods in applied linguistics. Oxford: Oxford University Press.
Do
Ferris, D. (1994). Lexical and syntactic features of ESL writing by students at different levels of L2 proficiency. TESOL Quarterly, 28(2), 414e420.
Forsman, L. (2004). Language, culture and context: Exploring knowledge and attitudes among Finland-Swedish EFL-students with particular focus on extracurricular influence. Vasa: Åbo Akademi. Licentiate's thesis.
Grant, L., & Ginther, A. (2000). Using computer-tagged linguistic features to describe L2 writing differences. Journal of Second Language Writing, 9(2),
123e145.
deHaan, J., Reed, W. M., & Kuwada, K. (2010). The effects of interactivity with a music video game on second language vocabulary recall. Language Learning
and Technology, 14(2), 74e94.
Higgins, C. (2010). Gender identities in language education. In N. H. Hornberger, & S. Lee McKay (Eds.), Sociolinguistics and language education (pp.
370e397). Bristol: Multilingual Matters.
International Telecommunications Union. (2013). Percentage of individuals using the internet 2000-2012. From http://www.itu.int/en/ITU-D/Statistics/Pages/
stat/default.aspx Accessed 18.09.13.
Jansz, J., & Martis, R. G. (2003). The representation of gender and ethnicity in digital interactive games. In M. Copier, & J. Raessens (Eds.), Level up: Digital
games research conference (pp. 260e269). Utrecht: Utrecht University.
Kachru, B. B. (1985). Standards, codification and sociolinguistic realism: the English in the outer circle. In R. Quirk, & H. G. Widdowson (Eds.), English in the
world: Teaching and learning the language and literatures (pp. 11e30). Cambridge: Cambridge University Press for the British Council.
Kuppens, A. H. (2010). Incidental foreign language acquisition from media exposure. Learning, Media and Technology, 35(1), 65e85.
Laufer, B. (1991). The development of L2 lexis in the expression of the advanced learner. The Modern Language Journal, 75(4), 440e448.
Laufer, B. (1995). Beyond 2000: a measure of productive lexicion in second language. In L. Eubank, L. Selinker, & M. Sharwood-Smith (Eds.), The current state
of interlanguage: Studies in honor of William E. Rutherford (pp. 265e272). Amsterdam: Benjamins.
Laufer, B., & Hulstijn, J. H. (2001). Incidental vocabulary acquisition in a second language: the construct of task-induced involvement. Applied Linguistics,
22(1), 1e26.
Laufer, B., & Nation, P. (1995). Vocabulary size and use: lexical richness in L2 written production. Applied Linguistics, 16(3), 307e322.
Lenhart, A., Kahne, J., Middaugh, E., Rankin Macgill, A., Evans, C., & Vitak, J. (2008). Teens, video games, and civics. Washington, DC: Pew Internet & American
Life Project.
Lucas, K., & Sherry, J. L. (2004). Sex differences in video game play: a communication-based explanation. Communication Research, 31(5), 499e523. http://dx.
doi.org/10.1177/0093650204267930.
Malvern, D. D., Richards, B. J., Chipere, N., & Dur
an, P. (2004). Lexical diversity and language development: Quantification and assessment. Basingstoke: Palgrave
Macmillan.
Martinez, R., & Schmitt, N. (2010). Invited commentary: vocabulary. Language Learning & Technology, 14(2), 26e29.
Miller, M., & Hegelheimer, V. (2006). The SIMs meet ESL. Incorporating authentic computer simulation games into the language classroom. Interactive
Technology & Smart Edcuation, 3(4), 311e328.
Milton, J. (2009). Measuring second language acquisition. Bristol: Multilingual Matters.
Minkova, D., & Stockwell, R. (2006). English words. In B. Aarts, & A. McMahon (Eds.), The handbook of English linguistics (pp. 461e482). Oxford: Blackwell.
~ oz, C., & Lindgren, E. (2011). Out-of-school factors e the home. In J. Enever (Ed.), ELLiE: Early language learning in Europe (pp. 103e124). London: British
Mun
Council.
Nation, P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University Press.
Olsson, E. (2011). Everything I read on the internet is in English” e On the impact of extramural English on Swedish 16-year-old pupils' writing proficiency.
Gothenburg: (Lic.), University of Gothenburg.
Peterson, M. (2012). Learner interaction in a massively multiplayer online role playing game (MMORPG): a sociocultural discourse analysis. ReCALL, 24(3),
361e380. http://dx.doi.org/10.1017/S0958344012000195.
Ranalli, J. (2008). Learning English with The Sims: exploiting authentic computer simulation games for L2 learning. Computer Assisted Language Learning,
21(5), 441e455.
Rankin, Y. A., Gold, R., & Gooch, B. (2006). 3D role-playing games as language learning tools. Paper presented at the EuroGraphics 2006, September 4e8, 2006,
Vienna, Austria http://www.thegooch.org/Publications/PDFs/Rankin_Gold_Gooch.pdf Accessed 27.12.10.
Reid, J. (1986). Using the writer's workbench in composition teaching and testing. In C. Stansfield (Ed.), Technology and language testing (pp. 167e188).
Alexandria, VA: TESOL.
Reid, J. (1990). Responding to different topic types: a quantitative analysis from a contrastive rhetoric perspective. In B. Kroll (Ed.), Second language writing:
Research insights for the classroom (pp. 191e210). Cambridge: Cambridge University Press.
Reinders, H., & Wattana, S. (2011). Learn English or die: the effects of digital games on interaction and willingness to communicate in a foreign language.
Digital Culture and Education, 3(1), 4e28.
Sarkeesian, A. (2012, December 4). The mirror. Retrieved from http://tedxwomen.org./speakers/anita-sarkeesian-2/ Accessed 18.09.13.
Schmidt, R. (1994). Deconstructing consciousness in search of useful definitions for applied linguistics. AILA Review, 11, 11e26.
Schwankert, S. (2007, September 6). Wikipedia blocked in China again. Retrieved from http://www.pcworld.com/article/136879/article.html, Accessed 18.09.
13.
Simensen, A. M. (2010). English in Scandinavia: a success story. In D. Wyse, R. Andrews, & J. Hoffman (Eds.), The Routledge international handbook of English,
language and literacy teaching (pp. 472e483). Milton Park, Oxon: Routledge.
Sockett, G., & Toffoli, D. (2012). Beyond learner autonomy: a dynamic systems view of the informal learning of English in virtual online communities.
ReCALL, 24(2), 138e151. http://dx.doi.org/10.1017/S0958344012000031.
Spada, N. (2005). Conditions and challenges in developing school-based SLA research programs. The Modern Language Journal, 89(3), 328e338.
Sundqvist, P. (2009). Extramural English matters: Out-of-school English and its impact on Swedish ninth graders' oral proficiency and vocabulary. Karlstad:
Karlstad University. Diss.
Sundqvist, P. (2011). A possible path to progress: out-of-school English language learners in Sweden. In P. Benson, & H. Reinders (Eds.), Beyond the language
classroom (pp. 106e118). Basingstoke: Palgrave Macmillan.
76
€m / System 51 (2015) 65e76
P. Sundqvist, P. Wikstro
€sny (Eds.), 20
Sundqvist, P. (2013). Categorization of digital games in English language learning studies: Introducing the SSI Model. In L. Bradley, & S. Thoue
years of EUROCALL: Learning from the past, looking to the future. Proceedings of the 2013 EUROCALL Conference, Evora,
Portugal (pp. 231e237). Dublin/
Voillans: © Research-publishing.net.
n, L. K. (2006). How is extramural exposure to English among Swedish school students used in the CLIL classroom? VIEWS e Vienna English Working
Sylve
Papers, 15(3), 47e53.
n, L. K. (2010). Teaching in English or English teaching? On the effects of content and language integrated learning on Swedish learners' incidental vocabulary
Sylve
acquisition. Gothenburg: Acta Universitatis Gothoburgensis.
n, L. K., & Sundqvist, P. (2012). Gaming as extramural English L2 learning and L2 proficiency among young learners. ReCALL, 24(3), 302e321. http://dx.
Sylve
doi.org/10.1017/S095834401200016X.
The Swedish Media Council. (2010). Ungar & Medier 2010. Stockholm: Regeringskansliet.
The Swedish National Agency of Education. (2012). Internationella språkstudien 2011. Rapport 375, 2012. Stockholm: The Swedish National Agency of
Education.
Thorne, S. L. (2008). Transcultural communication in open Internet environments and massively multiplayer online games. In S. S. Magnan (Ed.), Mediating
discourse online (pp. 305e330). Amsterdam: John Benjamins.
Turgut, Y., & Irgin, P. (2009). Young learners' language learning via computer games. Procedia e Social and Behavioral Sciences, 1(1), 760e764.
€ r skolan, 2000 Språk, 18, 27e41. Stockholm: Skolverket.
Viberg, Å. (2000). Tvåspråkighet och inl€
arning av språk i och utanfo
West, M. (1953). A general service list of English words. London: Longman, Green & Co.