8TH INTERNATIONAL CONFERENCE ON SECOND LANGUAGE STUDIES (ICSLS-2022) JUNE, 10 Young learners’ profiles with the MLAT-EC Mª del Mar Suárez Charles W. Stansfield www.EUROKD.com AUTHOR’S PHOTOS 2 AUTHOR’S SHORT BIOGRAPHIES Dr Mª del Mar Suárez is an associate professor and researcher at the Faculty of Education of the Universitat de Barcelona (Catalonia). As a GRAL member, I have taken part in several funded projects on SLA. My current teaching involves undergraduate courses in the Media Studies as well as in Early Years and Primary School Education degrees. I am the Editor-in-Chief of the academic journal Didacticae. Journal of Research in Specific Didactics, too. My main research interests are multimodal input, aptitude and formative assessment. Dr Charles Stansfield is an authority on second language testing. During his 50-year career, he has been a secondary school teacher of Spanish, a tenured professor of Spanish and teacher trainer at the University of Colorado, a test program administrator at Educational Testing Service, director of the ERIC Clearinghouse on Languages and Linguistics, and director of the Division of Foreign Language Education and Testing at the Center for Applied Linguistics in Washington, DC. In 1994, he founded Second Language Testing, Inc., for which he served as CEO until its acquisition by Berlitz Languages. Since 2014, he has devoted himself to managing LLTF. Dr. Stansfield has developed and published proficiency tests in English as a second language and in 15 other languages. He is the author or editor of over a dozen books and research monographs and 60 research articles published in professional journals. He has been a member of the editorial boards of Language Testing, the Modern Language Journal, TESOL Quarterly, and the Journal of Second Language Writing. 3 INTRODUCTION Cluster analysis 4 INTRODUCTION Hierarchical cluster  Dendrogram 5 INTRODUCTION Non-Hierarchical cluster 6 K- means: 1. Identify k clusters to assess how distinct our clusters are 2. Using the cluster centers identified before 3. F-values  how well the dimension discriminates INTRODUCTION Aims of cluster analysis in SLA: 1. Identify natural clusters within a mixture 2. Construct a useful conceptual scheme to classify individuals 3. Identify homogeneous subgroups  patterns useful for prediction 7 INTRODUCTION: WHAT FOR? 8 Learner & aptitude profiles Placement & selection Guidance Learning disabilities diagnosis Strengths and weaknesses Aptitude Treatment Interaction SLA research (f.ex. formal vs. informal contexts) LITERATURE REVIEW Aptitude profiles vs learner profiles ICSLS-2021 9 LITERATURE REVIEW ICSLS-2021 10 • Motivation profiles (Dörnyei et al., 2006) • Vocabulary learner strategies profiles (Kojic- Sabo & Lightbown, 1999) Proficiency as criterion variable • LAA in learner profiles Metalinguistic awareness in L1 & L2 and L2 proficiency over time (Ranta, 2002) • Learner differences in strategy use, will to learn and achievement over time (Yamamori et al., 2003) • Learner cognitive profiles including MLAT-4 Words in Sentences, age and criterion test score (Skehan, 1986) Proficiency included in the cluster analysis LITERATURE REVIEW: LANGUAGE APTITUDE COMPONENTS (Carroll, 1981) ICSLS-2021 11 • the ability to identify and memorize new sounds or strings of sounds Phonemic coding ability • the ability to understand how words function grammatically in sentences Grammatical sensitivity • the ability to infer grammatical rules from language samples Inductive language learning ability • the ability to learn a large number of semantic-symbol and/or sound-symbol associations in a short period of time Rote learning ability for FL materials LITERATURE REVIEW: CONSTRUCTS 12 Parts MLAT-EC (grades 3 to 7) MLAT (adults) Construct 1 Paraules ocultes (Hidden Words) Spelling Clues - vocabulary - sound-symbol association 2 Paraules que es corresponen (Words in Sentences) Words in Sentences - grammatical sensitivity 3 Paraules que rimen(Rhyming words) - - hear and make distinctions between speech sounds 4 Aprenguem números(Number Learning) Number Learning - rote memory - aural comprehension LITERATURE REVIEW: APTITUDE, PROFICIENCY AND AGE ICSLS-2021 13 Aptitude componentLAA PhCA Mem LITERATURE REVIEW: ADULT LEARNER PROFILES 14 •Strong on LAA (Ranta, 2002) •Strong on both LAA and memory OR strong on either LAA (syntactically oriented students) or memory (lexically-oriented students) (Skehan,1986, 1998, 2002) •Auditory abilities more relevant in younger students (Skehan, 1986) High achievers •Average or weak on LAA (Ranta, 2002) •Very poor memory except associative memory, average language ability (Skehan, 1986) • Intelligent but with poor language ability (Skehan, 1986) • High linguistic ability and memory but average IQ (Skehan, 1986) Low achievers THE STUDY: HOW ABOUT YOUNG LEARNERS? 15 1. Concrete operational thinking stage vs formal operational thinking (Piaget) 2. First stages of literacy development RESEARCH QUESTIONS 1. What learner profiles do younger high and low achievers have? 2. What aptitude components is determinant for younger learners, as measured by the MLAT-EC? 16 METHODOLOGY/ PROTOCOL: PARTICIPANTS  Context: 6 schools in Catalonia  Participants: bilingual Catalan-Spanish  English as a FL formal instruction 17 Grade 3 4 5 6 7 All Mean age 8.8 9.9 10.9 11.7 12.9 N MLAT-EC 55 62 30 60 62 269 METHODOLOGY/PROTOCOL: INSTRUMENTS 18 MLAT-EC Part scores  Proficiency measures: cloze passage measure (different versions for different grades)  integrative measure All of them reliable and valid across grades METHODOLOGY/ PROTOCOL: INSTRUMENTS 19 Grade 3 Grade 4 Grade 5 Grade 6 Grade 7 Part 1 .75 .81 .83 .87 .89 Hidden words easy v e r y e a s y Part 2 .32 .50 .65 .67 .74 Matching words difficult mid-difficult e a s y easy Part 3 .55 .72 .78 .83 .83 Rhyming words e a s y v e r y e a s y Part 4 .62 .82 .88 .85 .91 Number learning easy v e r y e a s y 55.62 83.67 93.17 94.20 101.47 0 50 100 150 MLAT-EC x/122 RESULTS: GRADE 3 20 Decoding skills Grammar sensitivity Sound recognition Rote memory FL cloze Cluster 1 N=23 Cluster 2 N= 29 Cluster 3 N=35 Cluster 4 N=11 -High decoding skills and average/high sound recognition ability -EITHER high grammatical sensitivity OR high memory + decoding skills for high achievement - Sound recognition alone, no effect RESULTS: GRADE 4 21 - Decoding skills needed for average and high achievement - Grammatical sensitivity relevant in high overall aptitude profile and in high achievement - High and average memory for average/ high performance Decoding skills Grammar sensitivity Sound recognition Rote memory FL cloze Cluster 1 N=22 Cluster 2 N=16 Cluster 3 N=28 Cluster 4 N=41 Cluster 5 N=12 RESULTS: GRADE 5 22 - 2 high achiever profiles with high overall aptitude and with high grammar sensitivity - Grammar – oriented vs. Memory oriented learner - Low achievers have poor decoding skills, poor sound recognition and average / poor memory Decoding skills Grammar sensitivity Sound recognition Rote memory FL cloze Cluster 1 N=5 Cluster 2 N=8 Cluster 3 N=23 Cluster 4 N=20 Cluster 5 N=18 Cluster 6 N=13 RESULTS: GRADE 6 23 -2 high achiever profiles: with high aptitude, or with high grammar sensitivity in spite of lower memory - Memory oriented learners who are average in the rest are average achievers - Low/average achievers have average or poor memory and are poor in the other abilities Decoding skills Grammar sensitivity Sound recognition Rote memory FL cloze Cluster 1 N=40 Cluster 2 N=7 Cluster 3 N=29 Cluster 4 N=16 Cluster 5 N=9 Cluster 6 N=17 RESULTS: GRADE 7 24 -2 high achiever profiles: with high aptitude or with both high grammar sensitivity and high memory but low sound recognition -Low memory, low achiever -Average sound recognition and spelling not crucial to have average achievement Decoding skills Grammar sensitivity Sound recognition Rote memory FL cloze Cluster 1 N=31 Cluster 2 N=16 Cluster 3 N=65 Cluster 4 N=7 Cluster 5 N=8 RESULTS: ANOVA F-VALUES ICSLS-2021 25 Grade 3 4 5 6 7 Part 1 (decoding skills) 21.067 (2nd) 43.031 (2nd) 39.852 (3rd) 21.468 (4th) 46.721 (3rd) Part 2 17.641 27.589 45.896 25.613 39.749 Part 3 18.898 34.819 32.327 48.318 85.473 Part 4 (memory) 21.361 54.467 43.925 52.375 99.537 DISCUSSION RQ1 High achievers vs low achievers profiles no linear high-aptitude profiles in the lower grades prominence of decoding skills in Grades 3 & 4 for high achievement  high-aptitude even profiles  either high grammar sensitivity or memory-oriented for both average and high achievement at higher levels, not in grades 3 & 4  low overall aptitude in low achievers, but not in Grade 3 26 DISCUSSION RQ2 Dimensions to differentiate clusters Decoding skills for younger learners Memory vs analytic abilities (still early stages of development in relation to FL learning?) 27 IMPLICATIONS The role of sound recognition and memory at lower levels (over decoding skills?)  Memory as differentiating dimension except in grade 5, followed by decoding skills at lower grades: Sparks et al’s LCDH in L1and FL, but in all alphabets? Aptitude test working differently across ages despite tapping into the same abilities 28 SUGGESTIONS FOR FURTHER RESEARCH Use of different proficiency measures tapping different abilities Aptitude profiles using proficiency as a criterion variable 29 30 #Make_Impact_Achieve_Impact www.EUROKD.com Thank you! Gràcies!