Individual differences in incidental language learning: The role of learning styles, personality, and working memory

Page 1

Learning and Individual Differences 38 (2015) 44–53

Contents lists available at ScienceDirect

Learning and Individual Differences

j ourna l homepage: www.e lsev ie r .com/ locate / l ind i f

Short report (original research)

Individual differences in incidental language learning: Phonologicalworking memory, learning styles, and personality

Sarah Grey a,⁎, John N. Williams b, Patrick Rebuschat c

a Georgetown University, United Statesb University of Cambridge, United Kingdomc Lancaster University, United Kingdom

⁎ Corresponding author at: Department of Psychology,Park, PA 16802, United States. Tel.: +1 814 865 9514.

E-mail address: [email protected] (S. Grey).

http://dx.doi.org/10.1016/j.lindif.2015.01.0191041-6080/© 2015 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 April 2014Received in revised form 12 December 2014Accepted 26 January 2015

Keywords:Language learningLearning stylesPersonalityWorking memory

We investigated whether learning of word order and morphological case interacts with three individualdifferences: phonological working memory, learning styles, and personality. Thirty-six participants en-gaged with a semi-artificial language during incidental exposure. Learning was assessed by acceptabilityjudgment and picture-matching tasks immediately after exposure and two weeks later. Participants alsocompleted learning style and personality surveys as well as two assessments of phonological workingmemory. The immediate results showed a significant learning effect on acceptability judgment only. No re-lationships were found for phonological working memory though effects did emerge for the extraversionpersonality trait and several learning styles. At delayed testing, results showed maintenance of learningon acceptability judgment and significant improvement on picture-matching. At delayed testing no rela-tionships between performance and individual differences were found. Overall, the results indicate thatlanguage learning under incidental exposure is durable and is not strongly constrained by individual differ-ences tested here.

© 2015 Elsevier Inc. All rights reserved.

1. Introduction

For many researchers in the field of second language acquisition,there are two key questions: (1) which of the external conditions towhich we expose learners promote learning (Norris & Ortega, 2000;Spada & Tomita, 2010) and (2) how do these conditions interact withindividual variation in learners (Dornyei, 2003, 2005). Regarding thefirst issue, a large body of research has examined the effectivenessof input conditions on second language (L2) learning, specificallyfrom the perspective of explicit or implicit contexts. Generally,such research has found explicit contexts, such as exposure to gram-mar rules or instruction to search for rules, to be more effective thannon-explicit contexts, which do not provide rules or instruction tosearch for them (see Norris & Ortega, 2000; Spada & Tomita, 2010for meta-analyses).

However, explicit conditions often provide learners withmore inputand/or more time-on-task (R. Ellis et al., 2009; Norris & Ortega, 2000;Sanz, 2005), which likely creates an artifact favoring explicit contexts.Research on the scope of learning under non-explicit exposure condi-tions is informative for theoretical accounts of adult language learning(e.g., Gass, 1997; Segalowitz, 2003; Schmidt, 2001; VanPatten, 2004)

112 Moore Building, University

not least becausemost language likely cannot be acquired intentionally.And though individual differences (IDs) are believed to be critical inadult language learning (Dornyei, 2003, 2005) there is little researchon individual variation in non-explicit exposure conditions. Here, weinvestigate the roles of several IDs in language learning under incidentalexposure, which we define as a learning context in which subjects arenot informed about the learning target, nor that they will be subse-quently tested (Williams, 2009).

1.1. Additional language learning under incidental exposure

While some researchers conduct between-group comparisons onthe efficacy of more or less explicit conditions (for review, see Spada &Tomita, 2010), another strand of research focuses strictly on the adultL2 user's ability to learn from exposure to language during a meaning-focused task, without explicit instruction (Williams & Kuribara, 2008;Rebuschat & Williams, 2012; Leung & Williams, 2011, 2012). Williamsand Kuribara (2008), for example, used a semi-artificial language,Japlish, to investigate the acquisition of non-native syntax in nativeEnglish speakers under incidental exposure. They provided subjectswith written instructions about the functions of case markers -ga(subject), -o (direct object), and -ni (indirect object) and participantsthen read and listened to Japlish sentences during a semantic plausibilityjudgment task. The results of a grammatically judgment test revealed

Page 2

45S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

a clear learning effect, with the exposure group outperforming theno-exposure group.

Rebuschat and Williams (2012) investigated whether the learningof non-native syntax under incidental exposure is accompanied byimplicit (unconscious) knowledge. In two experiments, native Englishspeakers were exposed auditorily to a semi-artificial language whichcombined German syntax with English vocabulary while they madesemantic plausibility judgments. The results on an acceptabilityjudgment task with confidence and source ratings (subjective mea-sures of awareness, Rebuschat, 2013) in experiment 1 showed thatonly learners who demonstrated awareness of their acquired knowl-edge also evidenced learning. In experiment 2, which tested fewerrules than experiment 1, there was a learning effect across all learners,and it was shown that exposure had resulted in both implicit and ex-plicit knowledge.

These studies provide evidence for the learning of non-native syntaxas a result of incidental exposure. However there remain several unad-dressed issues. First, delayed effects of learning were not assessed. Thelearning effects found in these studies are likely dynamic and to betterunderstand language learning under such contexts, delayed effectsmust be examined. Second, each study tested the learning of wordorder (syntax), but we know little of how other domains of grammarmight fare with incidental exposure. For example, Robinson (2005b)measured the learning of three morphosyntactic rules of Samoanunder incidental exposure and found clear learning effects for onlyone. This suggests that the learning of morphosyntax in an incidentalexposure condition may be less straightforward than the findings forsyntax1. Finally, the observed learning outcomes across subjects werevariable. In Williams and Kuribara (2008) only a subset of theparticipants evidenced learning of syntactic scrambling and inRebuschat and Williams (2012, exp. 1) only learners that showedawareness also showed learning. Such variability likely stems fromindividual differences, which may be the most reliable predictor of L2learning success (Dornyei, 2005). To date, however, there has beensurprisingly little work on the role of IDs in language learning underincidental exposure.

1.2. Individual differences

The role of individual differences in L2 learning following incidentalexposure has recently received attention from the perspective ofworking memory capacity, though its role in such a context is unclear.Tagarelli, Borges Mota, and Rebuschat (2011) found no relationshipbetween working memory and learning under incidental exposure toa semi-artificial German-like language (exposure and test measurespresented aurally). Working memory was assessed by an operationword span task and a letter–number ordering task (both written). Incontrast, Robinson (2005b) found that working memory positivelycorrelated with an aural grammaticality judgment test (GJT) and aguided production task following written incidental exposure toSamoan. Robinson suggested that the auditory GJT may have involvedworking memory because learners listen to the sentence and hold it inmemory to make a judgment about grammaticality but notes that “themeasure of working memory was a reading-span test and is not ameasure of aural phonological working memory capacity” (p. 260).Thus, a clear connection between working memory as assessed in thewritten modality (Robinson, 2005b; Tagarelli et al., 2011) and learningfollowing incidental exposure as assessed in the aural modality is diffi-cult to make.

1 Note that Robinson (2005b) did in fact administer 1 week and 6month delayed post-tests to his Samoan learners. However, no results of the performance at these testing ses-sions are reported, only correlations between task performance and the individualdifferences assessed in the study (Modern Language Aptitude Test andworking memory)are discussed. Thus, this study cannot inform on delayed effects of incidental exposure.

To address this, the present study investigated the role of phonolog-ical working memory (PWM) in language learning under an incidentalexposure context. In Baddeley and Hitch's (1974) multi-componentmodel of working memory, PWM is maintained by the phonologicalloop in the working memory system. The phonological loop consists oftwo separable sub-components: a temporary storage system whichholds memory traces for a period of seconds and a sub-vocal rehearsalsystem (see Baddeley, 2003; Baddeley & Logie, 1999). Phonologicalworking memory underlies the storage and processing of familiarand novel verbal and acoustic information and is considered criticalfor language learning, with implications for L2 vocabulary learning(Baddeley, Gathercole, & Papagno, 1998; French, 2006) as wellas L2 grammar learning (Speciale, Ellis, & Bywater, 2004). It mayalso be a key component of general language aptitude (Juffs &Harrington, 2011; Kormos & Sáfár, 2008; Robinson, 2005a; Sawyer& Ranta, 2001). A sizable portion of language learning takes place as aresult of exposure to auditory sequences but the role of PWM in thelearning of auditorily-presented language during incidental exposurehas yet to be addressed.

Phonological working memory, however, is just one ID that may beimportant. There exists a broad spectrum of IDs that interact with vari-ability in language learning outcomes (Dornyei, 2003, 2005), but howthese IDsmay interact with learning under incidental exposure is large-ly unknown. Therefore, the study included additional ID measures tobetter characterize the potential impact of individual differences onlearning under incidental exposure.

The first of these IDs is personality. In 1978, Naiman and colleaguesdescribed the “good” language learner as likely beingmore extraverted,but following their null effects for Extraversion (Naiman, Frohlich, Stern,& Todesco, 1996) personality research remained relatively unattendedin SLA. In the last decade, however, personality has been increasinglystudied in L2 speech production research (Dewaele 2005; Dewaele &Furnham, 1999; van Daele, Housen, Pierrard, & Debruyn, 2006) whereExtraversion has been found to positively correlate with fluency andcomplexity measures. Interestingly, a recent study in psychology byKaufman et al. (2010) found that intuition, openness to experience,and impulsivity personality traits were related to implicit learningability, suggesting that personality may interact with learning in addi-tion to L2 production.

We also investigated learning styles. Learning styles, or “anindividual's natural, habitual, and preferred ways of absorbing,processing, and retaining new information and skills” (Kinsella, 1995,p. 171), seem to vary systematically as a function of L2 proficiency. Forexample, early work by Hyland (1993) and Violand-Sanchez (1995)found that higher proficiency learners tended to be field-independentand intuitive, with a reliance on their own judgment. This researchassessed the impact of learning styles on the ongoing languagelearning process: low proficiency L2 learners were compared totheir higher proficiency peers. To our knowledge, no study has ex-pressly examined whether learning styles influence learning fromno knowledge of the language to a learning effect —and this studyaddressed that gap.

The current study brought together work on IDs and learningunder incidental exposure by investigating the role of PWM, learningstyles, and personality on the learning of additional language (L32)morphosyntax in late L2 learners of Spanish. We investigated the dura-bility of IDs on such learning by assessing the relationship betweenour IDs and performance immediately after exposure and following atwo-week delay.

2 Third language acquisition has also been referred to as TLA (Cenoz, 2011; De Angelis,2007). In this paper, the term L3 refers to any language acquired after the first and second,but does not strictly refer to the third sequential language acquired by the subjects. For adetailed discussion on multilingualism and third language terminology see Hammarberg(2010).

Page 3

Table 1Participant background information.

N = 36

M SD

Age at time of testing 19.4 2.3Years education 13.8 1.1Age of FL classroom exposure 13.2 4.2Semesters of FL classroom study 10.9 6.5Examples of foreign languages studied French, Spanish, Polish, Mandarin, Latin

Note. Age is expressed in years. M=mean, SD= standard deviation, FL= foreign language.

46 S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

2. Methods

2.1. Participants

Thirty-six university undergraduates participated in the study(Table 1). Participants were recruited from advanced (n = 21, 15females) and beginning (n = 15, 9 females) L2 Spanish classes, andthey all had English as their first language (L1). One original goal ofour study was to investigate the role of amount of prior L2 Spanish lan-guage learning experience in L3 learning (Grey, Williams, & Rebuschat,2014). Though the participants differed significantly in semesters of L2Spanish study, their total semesters of foreign language study (in addi-tional foreign languages beyond Spanish) were not significantly dif-ferent [t (34) = 1.78, p = .084, d = .62] and we therefore did notseparate these learners into distinct groups. Analyses are reportedon the entire group of learners. We refer readers to Grey et al.(2014) for more information on these groups as well as additionalsets of analyses that are beyond the focus and scope of this paper.3

A language background questionnaire indicated that no participantshad experience with Japanese, the language whose morphosyntacticfeatures were the target of the study.

Table 2Examples of Japlish items presented during the incidental exposure phase.

Pattern Examples

Plausible

2.2. Materials

2.2.1. Target languageTo assess learning of L3 morphosyntax, a semi-artificial language,4

Japlish, was adapted fromWilliams and Kuribara (2008). Japlish followsthe word order and case-marking rules of Japanese, but uses Englishvocabulary. In the current study a total of eight sentence structures(four simple, four complex) were created that model a subset of thoseused in Williams and Kuribara (2008). The word order is verb-final inboth simple and complex sentences and nouns are case-marked forsubject (-ga), object (-o), and indirect object (-ni). Verb-final wordorder and case-marking were the target structures in this study. Adetailed description of Japlish can be found in Grey et al. (2014) andsample Japlish sentences used during the exposure task can be foundin Table 2.

3 In Grey et al. (2014) we report analyses of performance on immediate and delayedtests for both assessment tasks as well as endorsement rates for each item-type (simple/complex, grammatical/ungrammatical) on each task at each testing session. We also re-port analyses for confidence ratings and source attributions for each task at each sessionas well as analyses that assessed potential effects of verbalizing a grammar rule for wordorder and case marking. Additionally, in that paper we confirm that the advanced and be-ginning learners were not statistically different in performance on our assessments, forgrammatical or ungrammatical items, at either session, or for the item-types. Finally, wereport post-hoc analyses on total semesters of language study and performance on bothtasks at both testing sessions.

4 The use of a semi-artificial language has several advantages. First, it allows for controlover prior knowledge of the structures aswell as language transfer. Second, it enables con-trol over the amount and type of exposure participants receive. Finally, given that previousresearch has demonstrated learning effects using semi-artificial systems (Williams, 2005;Leung&Williams, 2011, 2012), it is one of themost usefulmethods for investigating issuesin language learning that are difficult to isolate in the complex context of natural languageresearch.

2.2.2. Incidental exposure taskAll Japlish sentences were pre-recorded by a female native speaker

of American English and tasks were designedwith Superlab 4.0 (CedrusCorp, San Pedro, CA). The exposure stimuli consisted of 128 Japlishsentences (Appendix B). There were 16 exemplars of each word orderin Table 2, eight semantically plausible items and eight semanticallyimplausible ones. Participants heard each sentence once and presenta-tion was randomized. Participants were told they were going to listento sentences in a new language that contained English words andwere asked to decide if the sentences they heard were semanticallyplausible or not (Appendix A). Thus, participants' attentionwas directedtowards meaning as opposed to being directed to discover a rule orpattern. If the participant's judgment during the Japlish exposure wasincorrect, a brief tone was presented and the next item automaticallybegan. Participants were informed that the tone meant that their judg-ment was incorrect. Participants were not informed that they wouldbe tested, nor that they were expected to learn the morphosyntax ofanother language. That is, participants were receiving exposure underincidental learning conditions. This task lasted approximately 20 min.

2.2.3. Assessment tasksThere were two tasks designed to measure learning effects: accept-

ability judgment and picture-matching tasks. Both tasks were timed(participants were given 5 s to respond). The acceptability judgmenttask (AJT) consisted of 80 new Japlish sentences. Of these, 40 wereungrammatical and 40 were grammatical. Ungrammatical itemsexhibited incorrect word order (n = 20, five of each word orderviolation; Table 3) or had correct word order but contained a nounthat was missing its case-marking or had a case-marked verb (n = 20,five of each violation; Table 3). Grammatical items consisted of newJaplish sentences following the same eightword order patterns presentedduring exposure (n = 40, five each of the eight word orders in Table 3).For this portion of the task, participants indicated with a button press ifthe sentence sounded ‘good’ or ‘bad’. Two versions of this task weredesigned, with different Japlish sentences in each version, and thesewere counter-balanced within and across participants during immediateand delayed testing. Items were presented randomly and the AJT lastedapproximately 15 min. No feedback was provided.

The second assessment was a picture-matching task (PMT). Thisassessed whether participants learned the thematic content of themorphological case markings, i.e., not just that -ga needs to be on anoun, as indexed by responses to items on acceptability judgmenttask, but that -ga marks the subject. The task consisted of 40 picturespaired with either congruent or incongruent Japlish sentences in

OSIV Hammer-o Matt-ga brother-ni handedOSSIVV+ Dress-o Ann-ga sister-ga mother-ni brought that saidOSSVV+ Drug-o Brian-ga doctor-ga prescribed that saidOSV That coffee-o Tim-ga drankSIOV Brian-ga cat-ni food-o gaveSOV Matt-ga money-o gambledSSIOVV+ Ann-ga teacher-ga student-ni detention-o gave that saidSSOVV+ Detective-ga suspect-ga that car-o stole that saidImplausibleOSIV That toothbrush-o Stacey-ga secretary-ni quotedOSSIVV+ Essay-o Ann-ga computer-ga assistant-ni bathed that saidOSSVV+ Keys-o Ann-ga shoes-ga threw that suspectedOSV Towels-o Karen-ga puppy-ni overpaidSIOV Matt-ga girlfriend-ni those candles-o stapledSOV Tim-ga that movie-o drankSSIOVV+ Baker-ga cake-ga child-ni that frosting-o recited that saidSSOVV+ Lindsey-ga driver-ga that bus-o caused that said

Note. S = subject, O = object, I = indirect object, V = verb. Complex sentences aremarked with +.

Page 4

Table 3Grammatical and ungrammatical patterns used in the acceptability judgment task.

Pattern Examples

GrammaticalOSIV Debt-o Larry-ga bank-ni paidOSSIVV+ Car-o Lindsey-ga parents-ga son-ni loaned that announcedOSSVV+ Scotch-o bartender-ga guest-ga ordered that saidOSV That trail-o Larry-ga hikedSIOV Teacher-ga principal-ni grades-o reportedSOV Stacey-ga bag-o boughtSSIOVV+ Judge-ga lawyer-ga client-ni deal-o proposed that saidSSOVV+ Baker-ga customer-ga bread-o ordered that saidUngrammaticalSVO Tim-ga!called that police-oSIVO Brian-ga doctor-ni!showed those wounds-oSSVOV+ Driver-ga ciclist-ga!hit tree-o that saidSOVSV+ Lindsey-ga puppy-o!washed Karen-ga that said-ga !Stacey brother-ni tickets-o sold-o Tim-ga nurse-ni!symptoms explained-ni Ann-ga!landlord check-o wroteVerb Tim-ga Lindsey-ni news-o!delivered-ga

Note. ! indicates point of ungrammaticality onword order, case, or grammatical constituent.Complex sentences are marked with +.

47S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

terms of the case-markings on the nouns. Four listswere created and foreach list the same set of pictureswasused. Across the four lists, however,one of four possible variations of a particular Japlish sentencewas pairedwith its picture (Latin-square design). All of the picture-matchingsentences were new and participants were assigned to lists semi-randomly. Specifically, the lists were assigned in sequential order asparticipants initially arrived for the study, but at delayed testing partici-pants completed the ‘reverse-item’ list. For example, if the list containingvariation (1) for the kitten picture in Table 4 was given at immediatetesting, the list that contained (3) (for that particular picture) wasgiven at delayed testing. Participants indicated whether the sentencethey heard matched the picture on the screen: 20 pictures were pre-sented with matching sentences and 20 with mismatching sentences.For each set of 20, 10 items depicted a subject, object, and indirectobject and 10 depicted a subject and object only. Items were presentedrandomly and the task took approximately 10 min; there was nofeedback.

2.2.4. Phonological working memory tasksTwo non-word repetition tasks (NWRTs) assessed the learners' L1

and L2 PWM capacities. These tasks were used to match the auditorymodality of language exposure and testing to the modality of workingmemory assessment. The L1 NWRT provides a baseline measure of theparticipants' phonological processing and storage ability in their native

Table 4Sample pictures used in the picture-matching task with the four variations of Japlishsentences distributed across the four task versions.

(1) Kitten-ga tree-o climbed(SOV case, SOV word order)(2) !Tree-ga kitten-o climbed(SOV case, OSV word order)(3) !Kitten-o tree-ga climbed(OSV case, SOV word order)(4) Tree-o kitten-ga climbed(OSV case, OSV word order)

(1) Man-ga woman-ni necklace-o gave(SIOV case, SIOV word order)(2) !Necklace-ga man-ni woman-o gave(SIOV case, OSIV word order)(3) !Man-o woman-ga necklace-ni gave(OSIV case, SIOV word order)(4) Necklace-o man-ga woman-ni gave(OSIV case, OSIV word order)

Note. ! indicates ungrammatical case marking.

language. The L2 NWRT provides information on participants' phono-logical processing and storage ability for non-native phonotacticpatterns, which could be important during exposure to a foreignlanguage in an auditory context. The two tasks were identical with theexception of the language they were designed to approximate. In theL1 NWRT, the non-words obeyed English phonotactics (melistok,nutolon) while in the L2 NWRT, the non-words obeyed Spanish phono-tactics (asaco, filomo). Each task contained 16 pairs of non-words. Frompair 1 to pair 16 the number of syllables per word increased from threeto eight (e.g., by pair 16 learners had to hold 16 syllables in memory).The subjects listened to each pair and then repeated the words aloudafter a tone. Verbal repetitions were digitally recorded and later codedfor accuracy by the lead author. Repetitions were assigned one point ifthe participant repeated the item correctly without missing more thanone syllable, for a possible score of 32 points on each task. The itemsfor the L1 NWRT were pre-recorded by a female native speaker ofAmerican English (Lado, 2008). The items for the L2 NWRT werepre-recorded by a male native speaker of Spanish (non-peninsulardialect; Grey, Cox, Serafini, & Sanz, 2015). Each NWRT tookapproximately 5 min.

2.2.5. Individual difference questionnairesTwo questionnaires assessed learning styles and personality traits.

For personality, a paper-and-pencil version of the Big Five personalityquestionnaire was given to participants (Costa & McCrae, 1992a;McCrae & Costa, 2003). This questionnaire consists of 44 statements(Appendix C). For each statement, participants are asked to indicatetheir “strength of agreement” on a Likert-scale where 1 = stronglydisagree and 5= strongly agree. Sample statements include I see myselfas someone who is depressed, blue; I see myself as someone who generatesa lot of enthusiasm; and I seemyself as someonewho remains calm in tensesituations. The 44 statements index the personality traits of Extraversion,Neuroticism, Openness to Experience, Conscientiousness, and Agree-ableness. This five factor model of personality represents a robustparadigm in personality research and is accepted as themajor taxonomyfor classifying IDs in personality, largely as assessed through self- orother-reports (Chamorro-Premuzic & Furnham, 2005; Furnham,Monsen, & Ahmetoglu, 2009; McCrae & Costa, 2003).

To measure learning styles, participants were given the LearningStyle Survey (Cohen, Oxford, & Chi, 2006). This questionnaire assesseslearning style in 11 parts (110 total items, see Supplementaldocuments) and subjects are asked to rate “what they generally dowhen learning” on a 5-point Likert scale where 0 = never and 4 = al-ways. Some of the styles this survey identifies include Inductive/Deduc-tive, field dependence/independence, and Metaphoric/Literal styles.Sample statements include I learn better when I work or study with othersthan bymyself; I like to start with rules and theories rather than specific ex-amples; and I ignore details that do not seem relevant. This surveywas de-veloped from Oxford's (1995) Style Analysis Survey and though it is notwithout its critiques (Tseng, Dörnyei, & Schmitt, 2006) it is a representa-tive of commonly used instruments in thefield and is therefore likely ap-propriate for learning style research (Tight, 2010).

3. Procedure

Each participant was tested individually. Following informed con-sent, participants were taken to a sound-attenuated room where theycompleted the exposure task; they were not informed they would betested following this task. Following exposure, participants completedthe acceptability judgment and picture-matching tasks, the orderin which participants completed these tasks was counter-balancedwithin subjects and across the two sessions. After finishing the assess-ments, participants filled out the Big Five personality questionnaire(Costa & McCrae, 1995b), the Learning Style Survey (Cohen et al.,2006), and completed the twoNWRTs aswell as a language backgroundquestionnaire.

Page 5

48 S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

At delayed testing two weeks later, subjects returned to the testingroom and again completed acceptability judgment and picture-matching tasks. At this session participants filled out a debriefing ques-tionnaire that served as a retrospective report of knowledge gainedduring the study. Note that from immediate to delayed testing subjectshad no exposure to or practice with the target language.

4. Results

Mean accuracy was calculated for each participant on the AJT andPMT. However, accuracy may not capture potential response bias topredominantly respond ‘good’(or ‘bad’) to items at random, so partici-pants' performance was converted to d′ (d-prime) values, whichprovide an unbiased measure of the participants' ability to discriminatebetween good and bad items (Macmillan & Creelman, 2005; Stanislaw& Todorov, 1999). All data were analyzed using SPSS, version 18.0(IBMCorp). The ratings from each participant's personality and learningstyle questionnaire as well as the scores on the NWRTs were entered inSPSS for correlation analyses. For the delayed testing, analyses weredone on 34 learners since two participants did not return to the studyafter immediate testing. In our reporting of results for performance onthe AJT and PMT, we include effect sizes and follow the guidelines inOswald and Plonsky (2010) for interpreting Cohen's d: d = .40 is asmall effect size, d = .70 is a medium effect size, and d = 1.00 is alarge effect size. For the correlation analyses, we used a Bonferonni cor-rection to correct for multiple comparisons (alpha set at .002). BecauseBonferonni corrections are a highly conservative procedure and in-crease the likelihood of Type II errors (Bender & Lange, 2001;Perneger, 1998), we report significant correlations at p b .002 and alsomarginal significance at p b .05.

4.1. Acceptability judgment task

4.1.1. Immediate AJT and IDsAverage accuracy on the AJT at immediate testing was 57.5%

(SD = 14.9). The average d′ score on the immediate AJT was .244(SD = .527), and was significantly above chance, t(35) = 2.78,p = .009, d = .46 indicating that participants discriminated beyondwhat would be expected by chance between grammatical and un-grammatical Japlish items.

The correlation results for d′ on the immediate AJT are in Table 5.There were several correlations for performance and our ID factors.For learning styles, d′ showed a significant negative correlation withan Impulsive learning style, r (36) = − .608, p b .001, and a marginallysignificant positive correlationwith a Concrete–Sequential, r (36)= .344,p = .040 style. For personality, Extraversion showed a marginallysignificant negative correlation, r (36) =− .329, p = .037.

Analyses of the NWRTs indicated that learners' L1 and L2 PWMscores were significantly different, t(35) = 5.89, p b .001, d = .98(L1 NWRTm=16.5, SD=4.3; L2 NWRTm=13.6, SD=3.2). Scores

Table 5Intercorrelations between acceptability judgment performance with learning styles and person

Visual Auditory Kinaesthetic Ex

Imm. AJT − .031 − .056 .114 −Delayed AJT − .031 .009 .058 −

Closure Open Global PaImm. AJT − .094 − .060 .048 −Delayed AJT − .070 − .080 .021 .09

Leveler Deductive Inductive F-Imm. AJT .125 .218 − .267 −Delayed AJT − .029 .225 − .301 −

Metaphor Literal Agreeableness-P ExImm. AJT .124 − .259 − .155 −Delayed AJT − .034 − .210 − .259 −

Note: Imm. = immediate test; RI = Random-Intuitive; CS = Concrete–Sequential; F-indepeindexed by the Big Five questionnaire. Significance: +p b 0.05. *p b .002 (Bonferonni-adjusted,

on the NWRTs did not correlate with accuracy on the immediateAJT (L1 NWRT, r (36) = − .056, p = .744; L2 NWRT, r (36) = .281,p = .097), but did positively correlate with each other, r (36) = .747,p b .001.

4.1.2. Delayed AJT and IDsAverage accuracy on the AJT after a two-week delay was 59.8%

(SD = 15.8). The average d′ score was .408 (SD = .595) which wassignificantly above chance with a medium effect size, t (33) = 3.99,p b .001, d = .69. This indicates that the learning effects were durableand perhaps stronger, as the immediate AJT results showed a smallereffect size. There was also a trend towards a significant differencefrom immediate to delayed testing, t (33) = −1.71, p = .096, d = .29,which also suggests that learners improved in their discrimination ofgrammatical and ungrammatical Japlish items.

Correlation results for d′ scores on the delayed AJT can be foundin Table 5. For learning styles, two negative correlations weremarginally significant: performance on the AJT and the Extraversionlearning style, r(34) = − .489, p = .003, as well as the Impulsive style,r (36) = − .346, p = .045. For personality as measured by the BigFive questionnaire, the negative correlation between AJT perfor-mance and Extraversion was again a marginally significant effect,and was stronger at delayed testing, r(34) = − .481, p = .004. Nosignificant correlations were found for performance on the delayedAJT and NWRT scores: L1 NWRT, r(34) = − .039, p = .825; L2NWRT, r(34) = .005, p = .979.

4.2. Picture-matching task

4.2.1. Immediate PMT and IDsAverage accuracy on the PMT at immediate testing was 40.4%

(SD = 17.6) and the average d′ was .035 (SD = .381). Learnersperformance was not different from what would be expectedby chance t(35) = .554, p= .583, d= .10 and this indicates that par-ticipants were not discriminating sentence–picture matches frommismatches.

No significant correlations were found for d′ scores on the immedi-ate PMT and personality or learning styles, though PMT performanceand anAuditory learning style showed amarginally significant relation-ship, r(36)= .337, p=.045. The results from this analysis are in Table 6.No significant correlations were found for d′ on the PMT and L1 or L2phonological working memory: L1 NWRT, r (36) = .163, p = 342; L2NWRT, r (36) = .150, p = .383.

4.2.2. Delayed PMT and IDsAt delayed testing, average accuracy on the PMT was 56.3%

(SD = 19.7); the average d′ was .304 (SD = .652) and was signifi-cantly different from chance, t (33) = 2.71, p = .011, d = .47. Thisindicates that learners' responses were beyond what would beexpected for random guessing following a two-week delay. This

ality traits.

traversion Introversion RI CS

.227 .316 .051 .344+

.489+ .306 .037 .265rticular Synthesizing Analytic Sharpener.253 − .105 .297 − .1900 − .045 .283 .081independent F-dependent Impulsive Reflective.126 − .028 − .608* .233.033 − .219 − .346+ .071traverted-P Openness-P Neuroticism-P Conscientious-P.329+ .010 .205 − .006.481+ − .034 .101 − .064

ndent = field independent; F-dependent = field independent; -P = personality trait astwo-tailed).

Page 6

Table 6Intercorrelations between picture-matching performance with learning styles and personality traits.

Visual Auditory Kinaesthetic Extraversion Introversion RI CS

Imm. PMT .000 .337+ .069 .010 .258 − .050 .118Delayed PMT − .017 .059 .159 − .229 .312 − .030 .253

Closure Open Global Particular Synthesizing Analytic SharpenerImm. PMT .203 − .126 − .146 − .198 .058 .006 .110Delayed PMT − .056 .007 − .067 − .127 .110 .055 .095

Leveler Deductive Inductive F-independent F-dependent Impulsive ReflectiveImm. PMT − .243 .094 − .115 .232 − .028 − .052 .009Delayed PMT − .022 .038 − .199 .205 − .018 − .293 − .402+

Metaphor Literal Agreeableness-P Extraverted-P Openness-P Neuroticism-P Conscientious-PImm. PMT − .002 − .236 .217 − .190 .107 .099 .243Delayed PMT .218 − .283 − .177 − .189 − .028 .123 .038

Note: Imm. = immediate test; RI = Random-Intuitive; CS = Concrete–Sequential; F-independent = field independent; F-dependent = field independent; -P = personality trait asindexed by the Big Five. Significance: +p b 0.05.

49S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

improvement was echoed by a significant difference in perfor-mance between immediate PMT (no learning effects) and the de-layed PMT with a small learning effect, t (33) = −2.19, p = .036,d = .38.

At delayed testing, no significant correlations were found for thePMT and learning styles or personality; one marginally significantnegative correlation emerged for the Reflective learning style,r (34) = − .402, p = .019 (Table 6). Again, no significant correlationswere found for performance on the delayed PMT and PWM: L1 NWRT,r (34) = .135, p = .447; L2 NWRT, r (34) = .131, p = .460.

5. Discussion

Our study supports previous research showing that adults can learnnon-native language structures while processing the language formeaning, and without instruction to learn or search for rules(Williams & Kuribara, 2008; Rebuschat & Williams, 2012; Hamrick &Rebuschat, 2014; Robinson, 1995). The results from this study also sug-gest that learning under these conditions is not limited to word orderbut extends to morphological features such as case-marking (Rogers,Révész, & Rebuschat, 2015), as evidenced by significantly improved per-formance on the picture-matching task at delayed testing. The study hasprovided evidence that the learning that takes place is durable andimproves, even in the absence of additional practice with or exposureto the target language. With respect to the IDs examined, thoughthere were several trends towards relationships between perfor-mance on the learning assessments and our ID variables, there wasonly one statistically significant relationship once the results werecorrected for multiple comparisons. These results will be discussedin further detail below.

5.1. Phonological working memory

From a methodological perspective, this study diverged from bothRobinson (2005b) and Tagarelli et al. (2011) by using phonologicalworking memory tasks to assess the relationship between learning ofnon-nativemorphosyntax under both aural exposure and aural learningassessments. Tagarelli et al. (2011) used visual (written) workingmemory tasks with aural learning assessments, as did Robinson(2005b). The two studies found conflicting results regarding therelationship between working memory and learning under an inci-dental context. Our study matched the modality of the workingmemory assessment and outcome measures, and found no evidencethat it played a part in the observed outcomes.

The lack of any significant relationship between learning and PWMis similar to the findings in Tagarelli et al. (2011). In their study, nulleffects for working memory were found for learners in their incidentalexposure condition; only one measure of working memory (letter–

number ordering task) was related to performance on the acceptabilityjudgment task, and only for learners who were explicitly instructed tosearch for rules and became aware of them. On the other hand, the pres-ent study's findings run contrary to the results from Robinson (2005b)who foundworkingmemory to be related to performance on a listeningGJT followingwritten incidental exposure to Samoan. However, the cur-rent study used a task which did not move beyond the word level andtargeted only one component of the workingmemory system: the pho-nological loop. Robinson (2005b) used a sentence–span taskwhichmayhave tapped into aspects of working memory that are involved inlearning under an incidental context but were not measured here.Nonetheless, the results from Tagarelli et al. (2011) together withthe those from the current study suggest that the lack of a relation-ship may not merely be a task–effect, but instead a meaningful effect(or lack thereof) for aspects of working memory in learning underincidental exposure.

5.2. Learning styles

Previous language research on learning styles has compared learnersat different proficiency levels on their performance of an already-known language (e.g., Hyland, 1993; Violand-Sanchez, 1995). Thepresent study was the first to measure whether learning styles arerelated to language learning outcomes, beginning from no knowl-edge of the language structures to a learning effect. Performanceon the acceptability judgment task negatively correlated with anImpulsive learning style at immediate testing and this relationshiptrended towards significance at delayed testing. According to theLearning Style Survey, Impulsive learners react quickly without think-ing situations through. Thus, given the evaluative nature of our learningassessments, it seems unsurprising that impulsive, gut reactions to theJaplish items would lead to poorer performance, especially since theselearning assessments are testing the participants' sensitivity to goodand bad exemplars.

Several relationships showed statistical trends at p b .05 but werenot significant with the Bonferonni correction of p b .002. Thoughthese results did not meet our threshold for significance, they warrantbrief consideration. For the AJT, performance positively correlatedwith a Concrete–Sequential style at immediate testing and negativelycorrelated with an Extraverted learning style at delayed testing. AConcrete–Sequential style indexes how an individual handles possibili-ties. Learners who prefer this style are described as being present-oriented, preferring one-step-at-a-time activities and also wanting toknowwhere they are going in their learning at every moment. Learnersin the present study were not told they would be tested followingexposure and thus no a priori expectation for learning was created, atleast during immediate testing. Perhaps this type of exposure wasfavored by a more Concrete–Sequential style at the beginning of

Page 7

50 S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

learning since these individuals were carrying out the tasks “in thepresent” and one-at-a-time, with little regard for the future learningor testing tasks, and thus were better able to focus on the input athand. However, replications of similar designs are needed to confirmthis suggested relationship.

Regarding the trend for a negative correlation with an extravertedlearning style at delayed testing, Cohen et al.'s (2006) Learning StyleSurvey uses Introversion–Extraversion to index how an individualexposes themselves to learning situations. Introverts are described asenjoying working or studying alone or on a computer, which is theprecise environment in which subjects in this study were exposed toand assessed in Japlish. Extraverts, on the other hand, prefer social andinteractive learning tasks. Thus, an extraverted learning style may be abad longer-term fit for learning situations that involve robust input,such as listening or reading comprehension, and situations in whichthe learning context is receptive and much less social, interactive, ordialogic in nature.

5.3. Personality

This study, like much of the second language research on person-ality, found evidence to suggest that the Extraversion personalitytrait may play a role in the outcome measures (Dewaele & Furnham,1999; van Daele, et al., 2006). However, contrary to the work onL2 speech production, our results suggest a negative relationshipbetween Extraversion and performance, both immediately after ex-posure and following a two-week delay. This negative relationshipmay be related to the nature of incidental exposure and the charac-teristics of the target language in our study. Eysenck and Eysenck(1985) argued that extraverts are capable of high inhibition ofdistracting external stimuli while introverts have poor inhibitioncapabilities, making them more vulnerable to distracting externalstimuli. The target language in this study featured two linguisticcharacteristics that are not present in these learners' L1, namely arich morphological case-marking system and non-L1 word order.Such features may be distracting when language is presented with-out grammar rules or explicit information, as with incidentalexposure, and thus individuals with higher extraversion may inhibitthese distracting features. However, this study is the first study toexplore personality in incidental learning contexts and the interpre-tation of these results is preliminary. Researchers in future studiesshould consider evaluating personality alongside their assessmentsof learning in order to better characterize potential effects of personalityon language learning outcomes.

6. Limitations

As with all scientific research, this study has its limitations. First,we tested only receptive knowledge of Japlish and do not know theextent to which our incidental exposure condition also led to pro-ductive ability in the language, or how production might be relatedto the IDs we investigated. Second, although testing learning-effects for exposure-only groups is very common in artificial gram-mar/language studies (see R. Reber & Perruchet, 2003 for discus-sion), our lack of a no-exposure control group may open our resultsup to the critique that the observed outcomes on our assessmentscannot be isolated from participants having “learned” during the as-sessments as opposed to during the exposure phase. We examinedperformance during exposure by measuring accuracy to plausibilityjudgments in the first, second, third, and fourth quarters of theexposure task. In the first quarter participants responded with ap-proximately 58% accuracy that increased to nearly 70% accuracy inthe second quarter and then leveled around 72% in quarters threeand four. This suggests that the exposure phase was indeed effectiveand the AJT effects at immediate testing are a result of such exposure.Nonetheless, the added exposure inherent in our assessments may

have led to further learning or other processes or strategies (seealso Grey et al., 2014) which could be disentangled in futureresearch.

7. Conclusion

From a language learning perspective, this study has provided valu-able information on learning under incidental exposure and its potentialrelationship with three individual difference variables. Individual varia-tion had some relationship to our learning outcomes but did not appearto be an overwhelming factor, at least for the IDsmeasured here and theincidental learning context we investigated. Our finding for littleinfluence of individual variation in learning outcomes under incidentalexposure aligns with what A. Reber (1989, 1993, 1991) would predictin the related context of implicit learning (incidental exposure whichresults in unconscious knowledge). However, not only these individualdifferences, but also others such as motivation and sex, should be moreactively investigated so as to further our understanding of the interac-tions between learner-external and learner-internal variables in addi-tional language learning.

Appendix A. Exposure task instructions

You are going to hear a series of sentences in an artificial language.Even though it's artificial, it contains English words.

For this task,wewould like you to decide if the sentences are seman-tically plausible or not. For example, a sentence such as “The cat drankthe milk” is plausible, but “The cat barked at John” is not plausible,because cats don't bark.

Press / for any sentence you think is semantically PLAUSIBLE.

Press z for any sentence you think is NOT semantically plausible.Please take a moment to acquaint yourself with these key positions.If your plausibility judgment is not correct, you will hear a buzz toneand the next item will automatically begin.Please respond as quickly and accurately as possible after you heareach item. If you don't respond within the allotted amount of time,the next item will automatically begin. You will hear a total of 128sentences.This task will take you approximately 20 min to complete.If you do not have any questions, please press the space bar to beginthe task.

Appendix B. Incidental exposure items

Italicized items are semantically implausibleSimple structuresSOV

1. Brian-ga that runway-o saw2. Tim-ga whiskey-o tasted3. Larry-ga those boxes-o lifted4. Matt-ga money-o gambled5. Ann-ga those glasses-o washed6. Stacey-ga that picture-o painted7. Lindsey-ga pasta-o ate8. Karen-ga that car-o parked9. Brian-ga those plates-o fried

10. Tim-ga that movie-o drank11. Larry-ga shelf-o fought12. Matt-ga couch-o baked13. Ann-ga mail-o sang14. Stacey-ga those lamps-o watered15. Lindsey-ga that painting-o spent16. Karen-ga that pillow-o walked

Page 8

51S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

SIOV

1. Brian-ga cat-ni food-o gave2. Tim-ga Stacey-ni a letter-o wrote3. Larry-ga mom-ni that present-o sent4. Matt-ga boss-ni those reports-o faxed5. Ann-ga Larry-ni brownies-o baked6. Stacey-ga brother-ni dinner-o bought7. Lindsey-ga that neighbor-ni sugar-o loaned8. Karen-ga boyfriend-ni a message-o posted9. Patio-ga Brian-ni that chair-o put

10. Tim-ga colleague-ni that stapler-o ironed11. Larry-ga library-ni email-o painted12. Matt-ga girlfriend-ni those candies-o stapled13. Ann-ga mother-ni that letter-o bathed14. Stacey-ga cat-ni those issues-o presented15. Lindsey-ga those kids-ni bike-o prescribed16. Karen-ga bird-ni that nest-o assigned

OSV

1. Quiche-o Brian-ga baked2. That coffee-o Tim-ga drank3. Those roses-o Larry-ga watered4. That bonus-o Matt-ga spent5. Tune-o Ann-ga sang6. Puppies-o Stacey-ga walked7. That cold-o Lindsey-ga caught8. Those steaks-o Karen-ga fried9. That sweater-o Brian-ga parked

10. Pencils-o Tim-ga ate11. That mailman-o Larry-ga painted12. Television-o Matt-ga tickled13. That puppet-o Ann-ga gambled14. Bathtub-o Stacey-ga swept15. Those tapes-o Lindsey-ga tasted16. That shirt-o Karen-ga flattered

OSIV

1. That trophy-o Brian-ga student-ni awarded2. Tasks-o Tim-ga intern-ni assigned3. That crime-o Larry-ga volunteer-ni explained4. Hammer-o Matt-ga brother-ni handed5. Those seeds-o Ann-ga parrot-ni fed6. That note-o Stacey-ga Matt-ni handed7. Papers-o Lindsey-ga office-ni supplied8. That file-o Karen-ga roommate-ni showed9. Cake-o Ann-ga boyfriend-ni taught

10. That toothbrush-o Stacey-ga secretary-ni quoted11. .That computer-o Lindsey-ga manager-ni printed12. Towels-o Karen-ga puppy-ni overpayed13. That song-o Brian-ga cousin-ni boiled14. Letters-o Larry-ga desk-ni owed15. That bartab-o Tim-ga bartender-ni built16. Blanket-o Matt-ga dog-ni charged

Complex structuresS[SOV]V

1. Wife-ga jeweler-ga that ring-o lost that said2. Detective-ga suspect-ga that car-o stole that announced3. Bartender-ga customer-ga drink-o spilled that said4. Tim-ga dog-ga leg-o broke that said5. Customer-ga chef-ga food-o ruined that announced6. Stacey-ga mother-ga that play-o watched that said7. Matt-ga teacher-ga those essays-o published that announced8. Teacher-ga student-ga that paper-o plagiarized that said9. Brian-ga inspector-ga that story-o fired that announced

10. Larry-ga realtor-ga that house-o forgave that said11. Matt-ga play-ga show-o asked that announced12. Student-ga teacher-ga assignment-o sat that said13. Ann-ga clerk-ga that salad-o denied that announced14. Lindsey-ga driver-ga that bus-o caused that said15. Karen-ga assistant-ga those kittens-o cost that announced16. Officer-ga driver-ga pedestrian-o bottled that announced

S[SIOV]V

1. Larry-ga Stacey-ga friend-ni that present-o gave that said2. Owner-ga waitress-ga customer-ni that drank-o poured that as-

sumed3. Tim-ga Brian-ga police-ni that license-o handed that announced4. Larry-ga secretary-ga lawyer-ni those memos-o mailed that said5. Lindsey-ga boyfriend-ga blonde-ni number-o gave that suspected6. Karen-ga government-ga refugee-ni aid-o offered that announced7. Ann-ga teacher-ga student-ni detention-o gave that said8. Dean-ga department-ga professor-ni those grants-o awarded that

announced9. Larry-ga blanket-ga girl-ni those stitches-o reported that announced

10. Secretary-ga memo-ga company-ni employees-o frosted that said11. Tim-ga Ann-ga fishtank-ni that telephone-o loaned that assumed12. Matt-ga lawyer-ga case-ni that plaintiff-o laughed that said13. Karen-ga client-ga perfume-ni that man-o fried that suspected14. Stacey-ga banker-ga poster-ni teller-o licked that announced15. Lindsey-ga leader-ga announcement-ni those visitors-o buttered that

revealed16. Baker-ga cake-ga child-ni that frosting-o recited that said

OS[SV]V

1. Drug-o Brian-ga doctor-ga prescribed that said2. Vase-o Larry-ga cat-ga broke that assumed3. Money-o Tim-ga bank-ga loaned that announced4. Canvas-o Matt-ga sister-ga painted that revealed5. Bird-o Ann-ga neighbor-ga found that said6. Trophy-o Stacey-ga nephew-ga won that proclaimed7. Whiskey-o Lindsey-ga cashier-ga sold that said8. Book-o Karen-ga mother-ga bought that announced9. Coffee-o Larry-ga cup-ga spilled that said

10. Garbage-o Brian-ga broom-ga took that realized11. Sweater-o Matt-ga razor-ga made that announced12. Laptop-o Tim-ga disk-ga copied that said13. Keys-o Ann-ga shoes-ga threw that suspected14. Milk-o Lindsey-ga bottle-ga drank that announced15. Carpet-o Stacey-ga lamp-ga vaccumed that said16. Bench-o Karen-ga squirrel-ga cleaned that suspected

OS[SIV]V

1. Phone-o Brian-ga woman-ga son-ni bought that discovered2. Movie-o Larry-ga brother-ga colleague-ni loaned that said3. Pasta-o Tim-ga roommate-ga boss-ni cooked that announced4. Message-o Matt-ga girlfriend-ga co-worker-ni sent that suspected5. Dress-o Ann-ga sister-ga mother-ni brought that said6. Card-o Stacey-ga niece-ga patient-ni gave that announced7. Check-o Lindsey-ga boyfriend-ga company-ni paid that assumed8. Vaccine-o Karen-ga vet-ga kitten-ni gave that said9. Turtle-o Brian-ga sister-ga folder-ni sent that said

10. Water-o Tim-ga plant-ga manager-ni painted that assumed11. Steak-o Matt-ga knife-ga sister-ni cooked that revealed12. Portrait-o Larry-ga gallery-ga artist-ni swallowed that announced13. Essay-o Ann-ga computer-ga assistant-ni bathed that said14. Candle-o Stacey-ga match-ga boyfriend-ni ate that suspected15. Watch-o Lindsey-ga batteries-ga roommate-ni watered that an-

nounced16. Game-o Karen-ga instructions-ga employee-ni told that realized

Page 9

52 S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

Appendix C. Items included in the Big Five personality survey

I see myself as someone who…

1. Is talkative2. Tends to find fault with others3. Does a thorough job4. Is depressed, blue5. Is original, comes up with newideas6. Is reserved7. Is helpful and unselfish withothers8. Can be somewhat careless9. Is relaxed, handles stress well10. Is curious about many differentthings11. Is full of energy12. Starts quarrels with others13. Is a reliable worker14. …Can be tense15. …Is ingenious, a deep thinker16. …Generates a lot of enthusiasm17. …Has a forgiving nature18. …Tends to be disorganized19. …Worries a lot20. …Has an active imagination21. …Tends to be quiet22. …Is generally trusting

23. …Tends to be lazy24.…Is emotionally stable, not easily upset25. …Is inventive26. …Has an assertive personality27. …Can be cold and aloof28. …Perseveres until the task is finished29. …Can be moody30. …Values artistic, aesthetic experiences31. …Is sometimes shy, inhibited32. …Is considerate and kind to almosteveryone33. …Does things efficiently34. …Remains calm in tense situations35. …Prefers work that is routine36. …Is outgoing, sociable37. …Is sometimes rude to others38. …Makes plans and follows throughwith them39. …Gets nervous easily40. …Likes to reflect, play with ideas41. …Has few artistic interests42. …Likes to cooperate with others43. …Is easily distracted44. …Is sophisticated in art, music, orliterature

Notes.

Appendix D. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.lindif.2015.01.019.

References

Baddeley, A. (2003). Working memory and language: An overview. Journal ofCommunication Disorders, 36(3), 189–206.

Baddeley, A. D., & Hitch, G. (1974). Working memory. In G. A. Bower (Ed.), The Psy-chology of Learning and Motivation. Vol. 8. (pp. 47–89). New York: AcademicPress.

Baddeley, A., & Logie, R. H. (1999). Working memory: The multiple componentmodel. In A. Miyake, & P. Shah (Eds.), Models of working memory: Mechanism ofactive maintenance and executive control (pp. 28–61). Cambridge, UK: CambridgeUniversity Press.

Baddeley, A., Gathercole, S. E., & Papagno, C. (1998). The phonological loop as a languagelearning device. The Psychological Review, 105, 158–173.

Bender, R., & Lange, S. (2001). Adjusting for multiple testing—When and how? Journal ofClinical Epidemiology, 54(4), 343–349.

Cenoz, J. (2011). The influence of bilingualism on third language acquisition: Focus onmultilingualism. Language Teaching, 1–14.

Chamorro-Premuzic, T., & Furnham, A. (2005). Personality and intellectual competence.Mahwah, NJ: Erlbaum.

Cohen, A. D., Oxford, R. L., & Chi, J. C. (2006). Learning style survey: Assessing your learningstyles Styles-and strategies-based instruction: A teacher's guide, 15–21.

Costa, P. T. J., & McCrae, R. R. (1992a). Four ways five factors are basic. IndividualDifferences, 13(653–665).

Costa, P. T. J., & McCrae, R. R. (1992b). Revised NEO personality inventory (NEO-PI-R)and NEO five-factor inventory (NEO-FFI): Professional manual. Odessa, Fl: PsychologicalAssessment Resources.

De Angelis, G. (2007). Third or additional language acquisition. Tonawanda, New York:Multilingual Matters.

Dewaele, J. -M., & Furnham, A. (1999). Extraversion: The unloved variable in appliedlingusitic research. Language Learning, 49(3), 509–544.

Dewaele, J. -M. (2005). Investigating the psychological and emotional dimensions ofinstructed language learning: Obstacles and Possibilities. The Modern LanguageJournal, 89(3), 367–380.

Dornyei, Z. (2003). Individual differences in foreign language learning: Effects ofaptitude, intelligence, and motivation. Studies in Second Language Acquisition,25(4), 584–585.

Dornyei, Z. (2005). The psychology of the language learner: Individual differences in secondlanguage acquisition. New York: Lawrence Erlbaum Associates, Inc.

Ellis, R., Loewen, S., Elder, C., Erlam, R., Philp, J., & Reinders, H. (2009). Implicit and explicitknowledge in second language learning, testing, and teaching. Bristol, UK: MultilingualMatters.

Eysenck, H. J., & Eysenck, S. B. G. (1985). Personality and individual differences. New York:Plenum.

French, L. M. (2006). Phonological working memory and L2 acquisition: A developmentalstudy of children learning French in Québec. Lewiston, NY: Edwin Mellen Press.

Furnham, A., Monsen, J., & Ahmetoglu, G. (2009). Typical intellectual engagement, BigFive personality traits, approaches to learning and cognitive ability predictors ofacademic performance. British Journal of Educational Psychology, 79, 769–782.

Gass, S. M. (1997). Input, interaction, and the second language learner. Mahwah, NJ:Lawrence Erlbaum Associates.

Grey, S., Cox, J., Serafini, E. J., & Sanz, C. (2015). The role of individual differencesin the study abroad context: Cognitive capacity and language development duringshort-term intenstive language exposure. The Modern Language Journal, 99(1)(in press).

Grey, S., Williams, J. N., & Rebuschat, P. (2014). Incidental exposure and L3 learning ofmorphosyntax. Studies in Second Language Acquisition, 36, 611–645.

Hammarberg, B. (2010). The languages of the multilingual: Some conceptual and termi-nological issues. International Review of Applied Linguistics, 91–104.

Hamrick, P., & Rebuschat, P. (2014). Frequency effects, learning conditions, and the devel-opment of implicit and explicit lexical knowledge. In J. Connor-Linton, & L. Amoroso(Eds.), Measured language: Quantitative approaches to acquisition, assessment, process-ing, and variation. Washington, DC: Georgetown University Press.

Hyland, K. (1993). Culture and learning: A study of the learning style preferences ofJapanese students. RELC Journal, 24(2), 69–87.

Juffs, A., & Harrington, M. (2011). Aspects of working memory and in L2 learning.Language Teaching, 44(2), 137–166.

Kaufman, S. B., DeYoung, C. G., Gray, J. R., Jiménez, L., Brown, J., & Mackintosh, N. (2010).Implicit learning as an ability. Cognition, 116, 321–340.

Kinsella, K. (1995). Understanding and empowering diverse learners in ESL classrooms. InJ. M. Reid (Ed.), Learning styles in the ESL/EFL classroom (pp. 170–194). Boston: Heinle& Heinle.

Kormos, J., & Sáfár, A. (2008). Phonological short-term memory, working memory andforeign language performance in intensive language learning. Bilingualism:Language and Cognition, 11, 261–271.

Lado, B. (2008). The role of bilingualism, type of feedback, and cognitive capacity in the acqui-sition of non-primary languages: A computer-based study. Doctoral DissertationWashington, DC: Georgetown University.

Leung, J. H. C., & Williams, J. N. (2011). The implicit learning of mappings betweenforms and contextually derived meanings. Studies in Second LanguageAcquisition, 33, 33–55.

Leung, J. H. C., & Williams, J. N. (2012). Constraints on implicit learning of grammaticalform-meaning connections. Language Learning, 62, 634–662.

Macmillan, N. A., & Creelman, C. D. (2005). Detection Theory: A User's Guide (2nd Edition ).New York, NY: Cambridge University Press.

McCrae, R. R., & Costa, P. T. J. (2003). Personality in adulthood: A five-factor theory perspec-tive (2nd ed.). New York: Guildord Press.

Naiman, N., Frohlich, M., Stern, H. H., & Todesco, A. (1996). The good language learner. To-ronto: Ontario Institute for Studies in Education.

Norris, J. M., & Ortega, L. (2000). Effectiveness of L2 instruction: A research synthesis andquantitative meta-analysis. Language Learning, 50(3), 417–528.

Oswald, F. L., & Plonsky, L. (2010). Meta-analysis in second language research: Choicesand challenges. Annual Review of Applied Linguistics, 30, 85–110.

Oxford, R. L. (1995). Style analysis survey. In J. M. Reid (Ed.), Learning styles in the ESL/EFLclassroom (pp. 208–215). Boston: Heinle & Heinle.

Perneger, T. V. (1998). What's wrong with Bonferroni adjustments. BMJ, 316(7139),1236–1238.

Reber, A. S. (1989). Implicit learning of tacit knowledge. Journal of ExperimentalPsychology. General, 118, 219–235.

Reber, A. S. (1993). Implicit learning and tacit knowledge: An essay on the cognitive uncon-scious. Oxford: Oxford University Press.

Reber, R., & Perruchet, P. (2003). The use of control groups in artificial grammar learning.The Quarterly Journal of Experimental Psychology, 56A, 97–115.

Reber, A. S., Walkenfeld, F. F., & Hernstadt, R. (1991). Implicit and explicit learning:Individual differences and IQ. Journal of Experimental Psychology: Learning, Memory,and Cognition, 17(5), 888–896.

Rebuschat, P. (2013). Measuring implicit and explicit knowledge in second language re-search: A review. Language Learning, 63(3), 595–626.

Rebuschat, P., &Williams, J. N. (2012). Implicit and explicit knowledge in second languageacquisition. Applied PsychoLinguistics, 33(4), 829–856.

Robinson, P. (1995). Aptitude, awareness and the fundamental similarity of implicitand explicit second language learning. In R. Schmidt (Ed.), Attention andAwareness in Foreign Language Learning (pp. 303–358). Manoa, HW: Universityof Hawai'i Press.

Robinson, P. (2005a). Aptitude and second language acquisition. Annual Review of AppliedLinguistics, 25, 45–73.

Robinson, P. (2005b). Cognitive abilities, chunk-strength, and frequency effects in implicitartificial grammar and incidental L2 leaning: Replications of Reber, Walkenfeld, andHernstadt (1991) and Knowlton and Squire (1996) and their relevance for SLA.Studies in Second Language Acquisition, 27(2), 235–268.

Rogers, J., Revesz, A., & Rebuschat, P. (2015). Implicit and explicit learning of languages. InP. Rebuschat (Ed.), Amsterdam: John Benjamins.

Sanz, C. (2005). Adult SLA: The interaction between external and internal factors. In C.Sanz (Ed.), Mind and Context in Adult Second Language Acquisition: Methods, Theory,and Practice (pp. 3–20). Washington, DC: Georgetown University Press.

Sawyer, M., & Ranta, L. (2001). Aptitude, individual differences, and instructional design.In P. Robinson (Ed.), Cognition and second language instruction (pp. 319–353).Cambridge: Cambridge University Press.

Page 10

53S. Grey et al. / Learning and Individual Differences 38 (2015) 44–53

Schmidt, R. W. (2001). Attention. In P. Robinson (Ed.), Cognition and second language in-struction (pp. 3–32). Cambridge: Cambridge University Press.

Segalowitz, N. (2003). Automaticity and second languages. The handbook of secondlanguage acquisition, 382–408.

Spada, N., & Tomita, Y. (2010). Interactions between type of instruction and type of lan-guage feature: A meta-analysis. Language Learning, 60(2), 263–308.

Speciale, G., Ellis, N. C., & Bywater, T. (2004). Phonological sequence learning and short-term store capacity determine second language vocabulary learning. AppliedPsycholinguistics, 25(2), 293–321.

Stanislaw, H., & Todorov, N. (1999). Calculation of signal detection theory measures.Behavior Research Methods, Instruments, & Computers, 31(1), 137–149.

Tagarelli, K. M., Borges Mota, M., & Rebuschat, P. (2011). The role of working memory inimplicit and explicit language learning. In L. Carlson, C. Holscher, & T. Shipley (Eds.),Proceedings of the 33rd Annual Conference of the Cognitive Science Society(pp. 2061–2066). Cognitive Science Society: Austin, TX.

Tight, D. G. (2010). Perceptual learning style matching and L2 vocabulary acquisition.Language Learning, 60(4), 792–833.

Tseng, W. -T., Dörnyei, Z., & Schmitt, N. (2006). A new approach to assessing strategiclearning: The case of self-regulation in vocabulary acquisition. Applied Linguistics, 27(1), 78–102.

van Daele, S., Housen, A., Pierrard, M., & Debruyn, L. (2006). The effect of extraversion onoral L2 proficiency. Paper presented at the EUROSLA.

VanPatten, B. (2004). Input processing in SLA. In B. VanPatten (Ed.), Processing Instruction:Theory, Research, and Commentary. Mahwah, NJ: Lawrence Erlbaum.

Violand-Sanchez, E. (1995). Cognitive and learning styles of high school students: Impli-cations for SL curriculum development. In J. M. Reid (Ed.), Learning styles in the ESL/EFL classroom (pp. 48–62). New York: Heinle & Heinle.

Williams, J. N. (2009). Implicit learning in second language acquisition. InW. C. Ritchie, &T. K. Bhatia (Eds.), The New Handbook of Second Language Acquisition (pp. 319–353)(2nd ed.). Emeral Group Publishing LImited.

Williams, J. N. (2005). Learningwithout awareness. Studies in Second Language Acquisition,27, 269–304.

Williams, J. N., & Kuribara, C. (2008). Comparing a nativist and emergentist approachto the initial stage of SLA: An investigation of Japanese scrambling. Lingua, 118,522–553.