| Sažetak | Inojezična sposobnost (IS) defnira se kao skup kognitivnih i perceptivnih sposobnosti koje pojedincu omogućuju da brže i lakše savlada proces učenja stranoga jezika (Carroll, 1981). Cilj mjerenja IS-i prema tradicionalnom pristupu bio je predvidjeti ishod učenja na kraju određenoga razdoblja tijekom kojega se odvijao proces učenja stranoga jezika. Konstrukt IS-i odraslih tradicionalno je temeljen na sastavnicama koje počivaju na eksplicitnim kognitivnim procesima te poznavanju prvoga jezika. Novija saznanja u području kognitivne psihologije omogućila su istraživačima IS-i da se razvijanjem pristupa temeljenih na teoriji, a ne praksi, preispita uloga tradicionalnih sastavnica IS-i, a u skladu s time razviju nove mjere prilagođene specifičnim skupinama učenika stranih jezika. U skladu s time cilj istraživanja opisanoga u ovome doktorskome radu jest istražiti konstrukt IS-i u ranoj školskoj dobi, odnosno učenika prvoga razreda osnovne škole. Konstrukt IS-i učenika prvoga razreda osnovne škole te na samom početku formalnoga obrazovanja definiran je onim sastavnicma koje su pokazivale najbolju prediktivnost u istraživanjima s djecom, ali su ujedno pokazale najbolji potencijal za prilagodbu učenicima koji još uvijek razvijaju vještine čitanja i pisanja. To su sastavnice slušne pažnje te sposobnosti jezične analize. Ove su sastavnice operacionalizirane oslanjajući se s jedne strane na postojeće mjere (mjera učenja brojeva kao operacionalizacija slušne pažnje), a s druge strane na sasvim nove mjere (mjera slušne percepcije morfofonološkoga uzorka i mjera slušne percepcije točnosti nastavka za množinu kao dvije različite mjere sposobnosti jezične analize) na jeziku nepoznatom ispitanicima - mađarskom. Nakon prve faze istraživanja (N = 49) te druge faze istraživanja (N = 207) tijekom kojih se provjerila unutarnja i vanjska valjanost kao i latentna faktorska struktura sastavnica uključenih u konstrukt IS-i, uslijedilo je glavno istraživanje (N = 209). Provjeravanjem povezanosti mjera uključenih u konstrukt IS-i s individualnim razlikama u domeni kognitivnih sposobnosti (verbalno radno pamćenje, mentalna tranformacija i fonološka svjesnost) i nekognitivnih sposobnosti (mjere motivacije) te mjere okolišnih čimbenika (socioekonomski status obitelji, jezične navike u obitelji, izloženost stranome jeziku, uključenost djeteta u proces učenja stranoga jezika te stavovi roditelja) provedena je konvergentnodivergentna validacija konstrukta IS-i. Prediktivna valjanost utvrdila se provjeravanjem prediktivnosti svih varijabli (mjera IS-i te kognitivnih i nekognitivnih varijabli) u odnosu na kriterijsku varijablu (ishod na zadatku razumijevanja slušanjem), dok se inkrementalna valjanost vii utvrđivala usporedbom prediktivnosti mjera IS-i u odnosu na prediktivnost svih ostalih mjera uključenih u istraživanje. Rezultati provedenoga istraživanja (N = 209) ukazuju na rezultate koji su značajni u vidu definiranja konstrukta IS-i u ranoj školskoj dobi, ali i rasvjetljavanju njegove prirode. Naime, jednakim postotkom varijance postignuće u učenju stranoga jezika objašnjeno je tradicionalnom sastavnicom IS-i (mjerom slušne pažnje) te mjerom općekognitinivnih neverbalnih sposobnosti (mjerom menatalne transformacije). IS je potvrdila svoju inkrementalnu valjanost objašnjavanjem 4 % varijance uspjeha u učenju engleskog jezika. Mjera mentalne transformacije objasnila je dodatnih 4 % varijance, a motivacija te mjesečna primanja roditelja kao komponenta SES-a objasnili su po 10 % varijance uspjeha na rezultatu mjere razumijevanja slušanjem. Teorijski model sastavljen od čimbenika slušne pažnje, mentalne transformacije, motivacije, mjesečnih primanja roditelja te prethodnoga učenja engleskoga jezika objasio je 28 % varijance uspjeha u učenju engleskoga jezika na kraju prvoga razreda osnovne škole. |
| Sažetak (engleski) | Foreign language (FL) aptitude has been defined as a combination of cognitive and perceptual abilities that predispose individuals to learn foreign languages well and rapidly (Carroll, 1981). More recent definitions of FL aptitude describe it as a set of cognitive abilities that play a major role in both second and FL learning. (Li, 2022). It is considered that for adults, FL aptitude gives incremental predictive validity (Doughty, 2019). This means that FL aptitude explains success in FL learning beyond factors such as previous language learning, motivation, and personality facets. The construct of FL aptitude has been, for much of its history, seen as a stable, unitary construct operating on explicit cognitive processes and domain specific abilities (Carroll, 1959; 1973). Contemporary FL aptitude research, however, considers aptitude to be dynamic in its nature and composed of multiple independent components. As a result of this change in paradigm, the role of aptitude for specific groups of FL learners, such as young learners (YLs), is considered crucial in explaining success in reaching the outcomes of early FL learning programs (DeKeyser, 2000; Li, 2015; Robinson, 2019; Skehan, 2019). The research presented in this dissertation was motivated by three research gaps in the body of literature investigating FL aptitude of YLs. The first gap addresses the nature of the abilities making up the construct of FL aptitude of YLs or whether FL aptitude of YLs is domain general or domain specific. The importance of both domain general and domain specific abilities as predictors of FL aptitude of YLs aged 11 was confirmed by the LAPS project conducted in the Swiss context (Berthele & Lambelet, 2019). In addition to the cognitive variables, the LAPS research project also confirmed the importance of affective variables as well as socioeconomic factors in explaining the role of FL aptitude in FL achievement. Alexiou (2005) tested YLs aged 5-7 to determine the predictivity of domain general tasks originally designed by Esser and Kossling (1986). The results showed high predictive validity for success in YLs’ FL achievement. So far there have been no studies that attempted to compare the predictive validity of both traditional, domain specific measures and non-traditional domain general measures for YLs at the beginning of acquiring literacy skills. This brings us to the second gap related to the applicability of traditional aptitude measures like the MLAT-E for preliterate learners. Most of the tasks included in the MLAT-E do not cater to the needs of YLs at the very beginning of developing literacy skills. In other words, existing measures are not fully applicable for YLs at the beginning of formal education. Keeping in mind that the MLAT-E is designed to measure ix explicit cognitive abilities which are believed to be at the core of the FL aptitude construct, the question to be asked is can explicit cognitive processes underlying the traditional aptitude construct be operationalized without relying on YLs’ reading and writing skills? This question subsumes the third research gap and that is the question of the ways in which explicit cognitive processes play a role in the early stages of FL learning. With challenging the view of YLs purely relying on implicit learning mechanisms (Lichtman, 2016), came the need to address the role of language analytic ability as a possible component of the FL aptitude of young learners who are yet to acquire literacy skills (Alexiou, 2005). From existing language analytic ability measures it is clear that language analytic ability can be tapped into via tasks that measure explicit cognitive processes either in the participant's first language (e.g. grammatical sensitivity task (IV MLAT) in the MLAT-E, Carroll & Sapon, 2002) or via an unknown language to the participants (e.g. grammar inferencing task (LLAMA F in the LLAMA, Meara & Rogers, 2019), however the modality of the tasks needs to suit the learners' current literacy levels. When it comes to YLs who are at the very beginning of acquiring literacy skills, the only applicable measure is the artificial language game from the YLAT which is a measure relying on nonverbal material such as images and colours (Alexiou, 2005). Although applicable for young preliterate learners, such a task does not cater to the ecological validity of the FL aptitude measure by testing YLs’ auditory processing of unknown language material which is of crucial importance for the stage of input processing (Skehan, 1998, 2002, 2012). Such a measure could also provide explanatory power in addition to the predictive power of existing language analytic ability measures. Apart from versions of MLAT-E(elementary) (Carroll & Sapon, 2002), which has shown good predictive validity in several studies conducted in different L1 contexts (Kiss & Nikolov, 2005; Tellier & Roehr-Brackin, 2017), there have been no systematic attempts at designing new measures of FL aptitude of young learners (YLs). Drawing on existing literature as well as existing aptitude tests for YLs, a new aptitude measure was designed following the miniature natural language learning paradigm proposed by Kempe & Brooks (2016). The miniature language learning paradigm was used with the aim of achieving a measure of higher ecological validity in comparison to the existing measures. The ecological validity of the new measure was meant to ensure that the task demands resemble the demands of the particular language learning stage learners are trying to master as well as cater to the cognitive demands of a particular age group. During the validation process two studies (N=49; N=207) were conducted involving YLs x aged 7 to establish the final product of the validation: a novel aptitude test. A measure consisting of both visual and auditory input was designed using a natural language (Hungarian) which was at the same time a new language to the participants whose native language is Croatian. The final version of the aptitude test consisted of two measures: an auditory alertness measure and a language analytic ability measure consisting of two subtasks (auditory morphophonological pattern recognition and auditory perceptual acuity). To address the previously mentioned gaps in literature the present study is set out to investigate the FL aptitude construct of YLs at the very beginning of formal education. To test the convergent-divergent validity of the FL aptitude construct several correlations were explored: correlations between the FL aptitude construct and domain general abilities (verbal working memory and mental transformation), correlations between the FL aptitude construct and phonological awareness, correlations between the FL aptitude construct and motivation and, finally, correlations between FL aptitude construct and environmental factors were explored. All the variables included in the study (FL aptitude measures, verbal working memory, mental transformation, phonological awareness, motivation and environmental factors) were tested for their predictive validity for the results on the listening comprehension test administered at the end on the first year of FL learning. Finally, the study examined the incremental validity of the FL aptitude construct. The following research questions were asked, and subsequent hypothesis formed: 1. Is there a correlation between FL aptitude and domain general abilities (verbal working memory and mental transformation)? 2. Is there a correlation between FL aptitude and phonological awareness? 3. Is there a correlation between FL aptitude and environmental factors (socioeconomic status (SES) of the family (parents’ education, monthly income, number of books in the household), language related activities in the family, extramural exposure to English (including prior FL learning in formal contexts), YLs’ engagement with language learning activities, parental attitudes towards (early) language learning)? 4. Is there a correlation between FL aptitude and motivation? xi 5. Can language analytic ability measures designed for the main study predict YLs' achievement measured via listening comprehension at the end of the participant's first year of learning English? 6. Can auditory alertness designed for the main study predict YLs' achievement measured via listening comprehension at the end of the participant's first year of learning English? 7. Can both language analytic ability and auditory alertness predict YLs' achievement measured via listening comprehension better than each of the tasks individually? 8. What is the incremental validity of the FL aptitude construct? H1: There is a positive correlation between FL aptitude and domain general abilities (measured by verbal working memory and mental transformation). H2: There is a positive correlation between FL aptitude and phonological awareness. H3: There is a positive correlation between foreign language aptitude and environmental factors. H4: There is a positive correlation between FL aptitude and motivation. H5: Language analytic ability can predict YLs' achievement measured via listening comprehension at the end of the participant's first year of learning English. H6: Auditory alertness can predict YLs' achievement measured via listening comprehension at the end of the participant's first year of learning English. H7: Language analytic ability and auditory alertness together can predict YLs' achievement measured via listening comprehension at the end of the participant's first year of learning English better than each of the two FL aptitude measures individually. H8: Language aptitude can explain an additional percentage of variance of success in learning English at the end of first year of learning English than any other measure included in the main study (verbal working memory, mental transformation, phonological awareness, environmental factors, motivation). The results of the main study (N=209) point towards auditory alertness as the one aptitude measure that showed to be positively correlated with both verbal working memory (r(201) = 0.19), mental transformation (r(201) = 0.23, p < 0.001)., phonological awareness (r(201) = 0.32, xii p < 0.001), parents’ attitudes towards FL learning (r(167) = 0.20, p = 0.08), maternal education (τ-b(184) = 0.16, p = 0.014), and motivation (r(201) = 0.14, p = 0.043). Auditory alertness was also the only aptitude measure included in the study to be predictive of listening comprehension in English at the end of the first year of FL learning (r(201) = 0.35, p < .001). Since language analytic ability displayed no correlation to the results of the listening comprehension measure, nor any other measure included in the main study, apart from auditory perceptual acuity task and the mental transformation task (r(206) = 0.13, p = 0.059), it could not be tested further for its predictivity as an aptitude measure. The hypothesized predictivity of language analytic ability for the construct of FL aptitude was not confirmed, thus questioning its importance for the construct of FL aptitude of YLs. The best overall predictor of success on the listening comprehension measure were motivation and parents’ monthly income followed by two cognitive measures: auditory alertness and mental transformation. Both motivation and parent’s monthly income each explained 10% of variance of success on a listening comprehension measure in English. Mental transformation as a domain general measure explained the additional 4% of variance. Finally, the incremental validity of FL aptitude was confirmed with auditory alertness explaining 4% of variance of success on a listening comprehension measure. Overall, 28 % of variance of success in FL learning at the end of the first year of FL learning was explained with the new model of the FL aptitude construct of YLs comprising of auditory alertness, mental transformation, motivation, SES and prior FL learning. |
