Neuroscience of Education PDF

Session 1: a. ***[Neuroscience of Edu]*** -functional magnetic resonance imaging (fMRI)= measure areas of increased blood flow in brain, correlates to neural activity -electroencephalography (EEG) =electrical activity in brain (can identify children at risk of dyslexia) -magnetic source imaging...

Session 1: a. ***[Neuroscience of Edu]*** -functional magnetic resonance imaging (fMRI)= measure areas of increased blood flow in brain, correlates to neural activity -electroencephalography (EEG) =electrical activity in brain (can identify children at risk of dyslexia) -magnetic source imaging (MSI) = MEG (magnetic fields by brain electrical activity)+MRI -Transcranial Magnetic Stimulation (TMS) = electromagnet to pulse magnetic field stimulate neurons -brain's information processing: not narrow sequential processing, but large-scale parallel processing and extraction of stable patterns -dyslexia is not a qualitatively different syndrome but has degree of deficits **Reading:** -map orthographic forms using alphabetic orthography\ -on grapheme-phoneme (letter-sound) level, decode writing into known words\ -dependent on development of amodal processing structures in temporoparietal and left ventral cortex, integrating phonology, orthography and semantics (verbal word form area VWFA)\ -metaphonological awareness: capacity to deliberately utilize knowledge of word sounds **Mathematics:** -two innate perceptual systems:\ distinguish without counting, small exact quantities up to 3 (subitization)\ judge relative differences in large quantities (like 16 vs 32) -Dyscalculia: deficit to learn arithmetic, due to functional and structural abnormalities in parietal lobe known for enumeration\ -relationship between arithmetical skills and innate abilities -- math weaker than for reading **Science:** -brain basis for organization of causal perceptions\ -2y/o children can retrospectively assess cause and outcome for anticipating future events -require mapping of explicit concepts onto implicit perceptions (distinct types)\ -constant calibration between observable regularities and inferred causal relationships\ -working memory and executive function: dorsolateral prefrontal cortex ***[B) Educational Neuroscience (Why is Neuroscience Relevant to Edu?)]*** 8 different neural systems in brain: (in general destination: automacity)\ -**hippocampus** is for episodic/autobiographical **memory** and **spatial** awareness\ -**classical conditioning** in **limbic** structures\ -**control** context-specific functions in **prefrontal** cortex\ -**conceptual** learning in perception+motor, spatial+temporal (in cortices)\ -**rewards**, dopamine in **amygdala, ventral tegmental area\ **-**procedural** learning/memory: **basal ganglia** and **thalamus**\ -**social** learning: **prefrontal** for imitating manners\ -**language** system: **temporal** and **occipital** ***Seminar:*** Ethical Issues in Neuroscience in Edu: experience-dependent, artificial neuromodulation, psychostimulants, use of TMS, parental consent **Session 2:** ***[C) An introduction to brain and cognitive development]*** B&B RECAPS:\ -cell body contains nucleus, **dendrites** as branches, long arm **axon** end in **synapses** and **synaptic cleft**\ -50% brain cells -\> neurons, other 50% **-\> glial cells for nutrients, maintenance** Glial cells: **oligodendrocytes** (central -- brain) and **Schwann** (Peripheral -- body) cells\ **White matter**: axons covered in myelin and glial cells\ **Grey matter**: neuronal cell bodies and glial cells (Cortex) Neurotransmitters: excitatory and inhibitory (Most common excitatory = glutamate; most common inhibitory: GABA) Bump = gyrus, Trough = sulcus ***[Functions:]***\ **Cerebellum**: motor control/balance\ **Fusiform face area (FFA), inferior temporal cortex**: social cognition (processing faces)\ **Amygdala**: Emotions of faces\ **Medial prefrontal cortex + Superior temporal sulcus (STS)**: receive input from visual and auditory cortices to process sights and sounds -\> emotional cues\ **Medial prefrontal cortex + Temporo-parietal junction**: Mentalising (infer people's thoughts)\ **Lateral prefrontal cortex, anterior insula, superior parietal cortex and intraparietal sulcus, anterior cingulate cortex, pre-supplementary motor area** = a chain/connection for cognitive functions/doing tasks\ **Striatum**: both memory and emotion/motivation\ **Ventral Striatum**: decision making and reward-related behaviour (a lot of dopamine)\ **Anterior insula:** self awareness, gut feelings\ **Orbitofrontal cortex:** associate actions with reward/emotion **Declarative** memory: semantic (word, concepts) and episodic (events) = HIPPOCAMPUS\ **Non-declarative** memory: Priming and procedural memory Language:\ -**premotor cortex along precentral gyrus**: overt speech articulation regions for repeating non-words\ -**middle temporal gyrus + angular gyrus** = word activation \> non-word activation\ -**Middle temporal gyrus + Superior temporal sulcus** = sentence comprehension\ -**Wernicke's Area**: semantic/syntactic (comprehension) -- closer to auditory cortex\ -**Broca's Area**: Word retrieval (production) Brain **imaging** Methods!! a. **Structural** **MRI**: noninvasive doughnut magnet!\ -structural development of brain over time, changes in grey and white matter volumes, thickness and folding, volume of subcortices -brightness of tissue in image = how tissue absorbs and gives off energy (T1 scans: gray matter look grey, white matter look white, T2 is opposite!) -first few months and years after birth = **dendrites** and **synapses** and brain **vol** **increase** rapidly.\ -then throughout childhood, adolescence: **decreases** **thickness**, **grey** matter, timing is **region** **specific** Shows **synaptic pruning**: infrequent connections get eliminated to fine tune (cortical thinning is adaptive but can lead to ADHD and schizophrenia)\ Conflicting: less grey matter = more efficient, mature but more grey matter: increased processing power! b. **DTI**: different MRI technique\ -detailed study of white matter tracts\ -see how water diffuses in brain tissue to reveal location and orientation of white matter tracts (so brain regions can communicate)\ -water meander slowly through grey matter multi-directional, but faster in white matter less diffusible c. **fMRI**: study of brain activity in blood oxygenation and blood flow (functional -- when doing task)\ -functional connectivity (Spatial resolution)\ -indirectly measure changes in location of neuronal firing like fNIRS d. **EEG**: measurement of brain function (Hz frequency) -- neuron firing -\> electrical activity The higher the frequency when measuring, the better the temporal solution\ -electric potential at surface of scalp using electrodes\ =measure difference in voltage between 2 electrode sites at regular time intervals 1\) send brainwaves by electric signal, to see **frequencies** among neuron populations (=different alertness states)\ -slower = relax, sleep (1-13Hz); faster = alert, problem solving (13-100Hz)\ 2) event-related potentials (**ERPS**): ***average*** fluctuations in electric signal after stimulus (time-locked, epoch)\ -positive or negative waveforms (change in voltage over time; check **amplitude**, **latency**, **distribution** to see nature timing and organisation of neural systems) e.g.N170: delay of 170millisecs when process faces in posterior temporal lobe (FFA)\ -positive peak initially (\~100ms) may = oh I know this thing (awareness, knowledge); then the negative trough N170 = oh this is A FACE! **Latency**: time which max + or -- activity occurred (Longer latency = slower processing (higher cognitive process), usually within first 200ms of stimuli)\ **Amplitude**: larger amplitudes = more neural activity, number of neurons firing\ **Distribution**: where it is detected is NOT correlated to exact localization of underlying source (observed in left electrode sites does not mean generated in left brain region) -- poor spatial resolution, very vague **Procedures** for EEG/ERP:\ placement of electrodes -\> amplifying and filtering EEG -\> determining sampling rate -\> selecting reference -\> time locking EEG to stimuli -\> detect artefacts -\> reject or correct artefacts -\> average and plot ERP **Filters** = frequencies they allow to pass for recording (e.g. high-pass filters: eliminate contamination of low-pass frequencies)\ -bandpass filters have both high and low settings **Types of waves** (from lowest frequency to highest, busiest)\ Delta: calm, lowest frequencies (sleep, dreaming)\ Theta: drowsiness Alpha: reflective, restful\ Beta: busy, active\ Gamma: problem solving, concentration **Strengths**: not much cognitive process involved, just need to react to stimuli, longitudinal (Can be), direct and instantaneous, good temporal resolution, identify temporal sequence of cognitive processes (very instantaneous) **Limitations**: blinking directly affects electrode (is an artefact), need filters and might misinterpret, poor spatial resolution (hard to localization, anatomical area) e. **NIRS**: near infrared spectroscopy to map hemodynamic responses (local blood flow, volume) with neural activity (here optodes = sensors)\ -detect light changes as a consequence of blood concentration -change in local hemoglobin concentrations when neurons fire\ -biological tissues' transparency to near infrared light (to see cortical activation)\ -light that migrates from source to detectors to see brain tissue + measure blood oxygenation -can use for resting states, or regions of interest Big/overall changes in waves = change in blood oxygen consumption levels\ Small oscillations in between = heartbeat! **Strength**: less susceptible to artefacts, spatially resolved image for cortical localization, tolerant to movement e.g. they found infants activate posterior superior temporal for human dynamic actions, but not activating for non-human movements! (bias for human specific)\ **Limitations**: poor spatial and temporal resolution **Stimulation** brain methods!!! f. **TMS**: eight magnetic coils to induce electric field to interfere neural processes\ -trial by trial at specific times, repetitive manner g. **tDCS**: constant, low current by electrodes\ -change neural excitability for long term changes in function\ -treat depression and cognitive enhancement Brain development\ -neurons **born** -\> **migrate** -\> **differentiate**\ e.g. in cerebral cortex, neurons move along radial glial cells, move past old cells to surface of brain (migration!)\ -**programmed cell death**: during migration, 20-50% cells die due to errors in cell division or just to eliminate surplus neurons\ -synaptic pruning + cell death can **stabilize behaviours**! -**differentiation**: growth/branches of dendrites, myelination\ -when delayed = delay in development Controversy: differentiation of cortex into areas/regions: **domain-specific or activity-dependent (nature nurture)\ **1^st^ suggestion: by protomap/blueprint: pre-specification of tissue, intrinsic markers\ 2^nd^ suggestion: protocortex is initially undifferentiated, but they outcompete each other for functions **Neuronal recycling:** circuits with most appropriate functions are repurposed through practices\ e.g. fusiform gyrus named VWFA when processing words, but can also process objects and faces **Ways to understand brain development:\ 1. Genes** -- probabilistic epigeneist\ -development are bidirectional external and internal through gene, neural, behaviour and environment\ e.g. Activity A activates gene, while the gene can also be used for activity B **2. Encellment:** neural activity for both progressive elaboration and stabilisation\ -e.g. closing one eye during postnatal -\> shrink columns in closed eye and expansion in the opened eye (activity based competition) **3. Embrainment: context sensitive**\ -e.g. blind people from an early age: braille reading correlated to visual cortex in sighted ppl\ -interactive specialisation view, narrower set of circumstances -\> functional specialisation **4. Embodiment**\ -body acts as a filter (e.g. when babies have physical restrictions, limits potential complexity in representations of environment)\ -manipulate environment (infants move arm to beam until block light, generate new sensory experiences) -\> feedback loop proactive approach\ -not passively absorb information but manipulate (real time interaction) **5. Ensocialment\ **-synchronous interactions have strong effect***\ *** Session 3 Lecture: **Language Development** -form of human communication in either spoken, written, signed forms; using words in a structural and conventional way From **womb**: Last 3 months gestation start to have **sensitivity** to language\ After birth: **Cooing (interactional vocalisation**), physical limitations in vocal tract\ First few years: **Phonological** development, babbling: put vowels with consonants\ Start school: Semantic/Syntactic development: word production (simple sentences)\ Formal Edu: Literacy, pragmatic development, complex language and second language Mainly left hemisphere, Broca's area and Wernicke's Area (production and comprehension respectively)\ Left temporal cortex: Verbal sounds (wernicke's area) **Phonological development -- categories of sounds\ **-when we are born, no specialization of sounds, we absorb all of them\ -first 5 months of our life, we **perceptual narrow** our specialisation into our own language\ -matured connectivity in temporal cortex from 6 months old! e.g. using EEG/ERPs with 7.5 months children,\ Stimuli = using phonological sounds, some are native language of infants, others are not\ Instruction = Mismatch Response (MMR)\ DV: size of difference in brain response to native vs non-active language phonemes **Result** in EEG: +100ms positive peak! = recognise native sound\ -150 to 200ms negative trough! = recognise non-active sound The larger the amplitude for native sound (higher familiarity) -\> the smaller the amplitude for the negative trough for non-active sound (less care)\ =we optimise the phonological representation from early on and we discard the things we don't need Lecture: a. Size of MMR to native negatively correlated to size of non-native b. Positive language outcomes at 24 and 30 months correlated with large MMR! **Neurobiological Maturation\ **-Grey matter (body): exponentially decreases due to specialization throughout life\ -White matter (neurons, axons): branches, networks reach peak during 30-40yo, then decrease again (quadratic) **Interactive Specialization theory:**\ -cortex area is specialised due to experience and input\ -not driven by chronological age (quality of parents and caregivers interaction very important) **SES:\ **-explains phonological awareness and receptive vocab most\ -cognitively rich environment allows inherent differences to emerge\ -input becomes dominant constraint on development in less rich environments **Neural system specialisation:\ E.G. fMRI studies examine speech regularities (statistical learning)\ **-Speech regularities: patterns and rules that characterise sounds, words, sentences of a particular language. To predict e.g. subject, verb, object (syntax), or e.g. pinyin (phonological level)\ -Statistical learning: implicitly extract statistical regularities or patterns from sensory environment -fMRI study on 156 individuals from 5 to adulthood\ Results: temporal cortex more and more specialised but lateralization varied\ 6yo: right temporal lateralization (opposing to common conception of left hemisphere being sole area)\ 10 and 13 yo: bilateral lateralization\ Adults: left\ = changes in lateralization through development! **Conclusion:**\ -6 and 10 yos signal increase for speech stream containing weak statistical regularities\ =behavioural **decrease in ability to acquire other language** when native proficiency increases with age\ =language learning is not fixed across development through biological maturation and relevant learning experiences **Developmental Language Disorders (DLDs)/Specific Language Impairment (SLI)\ **-in expression, comprehension, vocab, reading etc. **Adolescence:**\ -start idioms, proverbs, figurative language (pragmatic language)\ -auditory processing, EEG\ -maturation of phonological categorization\ -motor control\ -verbal fluency **Neuro changes:**\ maturation of white matter tracts\ synaptic pruning (removing connections no longer needed)\ less grey matter volume\ -lateral temporal last to mature\ -maturing arcuate fasciculus = language comprehension and production\ -white matter in frontal = pragmatics of behaviour and language\ -grey matter changes in left parietal: vocab knowledge **Experiment: British Picture Vocabulary Scale (BPVS)**\ -experimenter say a word related to one out of 4 pics, participants choose d. ***[Neuroscience and Reading]*** -no indication that areas of brain dedicated from birth for reading\ -anatomical result of development after successful instructional experiences **Decoding processes**\ -bilateral activation (both hemispheres)\ -audition, vision, spatial, cross-modal processing\ -spoken-language areas (posterior superior temporal cortex, occipitotemporal cortex, temporal, parietal, frontal cortex) **Implicit Reading Task (visual search task)**\ -aims to dissociate reading, meaning-making from printed symbols, processing visual sequences\ = use false fonts (meaningless hieroglyphic-type symbols visually matched to letters) -\> pick out target visual features like ascenders in b,d,k.\ -compare to same task using real words and pick out visual features **Results:** activation is left-lateralized and on occipitotemporal and posterior superior temporal cortices (for letter identification) **Visual Word Form Area (VWFA)**\ -active during nonsense words/any sequence of printed letters, not purely for word forms\ -activation increases when orthographic strings are more familiar, support print-sound connections **Experiment example:** detect repetition of real words vs meaningless symbol strings\ -measurement of word-specific neural processing: N170: negative deflection in brain electrical activity after stimulus onset\ -**N170 means brain detects nonwords from words within 160-180ms once presented** **Results:\ **-no reading instruction: no N170 despite knowing individual letters (can't differentiate nonwords vs words)\ -1.5 yrs reading instruction: typically developed have N170\ -dyslexia at risk: no N170 until second grade (reduced N170) but not absent\ =reduced N170 is clear neural correlate of visual word-processing deficit **Letter with Sounds:** **Experiment example:** ask whether participants heard vowel sounds in a forced-choice auditory task using degraded stimuli -they either just heard speech sounds or sounds + visually presented letters\ -either congruent or incongruent Results:\ -**better at auditory-visual** condition than auditory-alone for **congruent**, significantly **worse for incongruent**\ -auditory-visual condition have **different brain activity** than auditory-alone condition (differ in speech recognition brain areas but not occipital like VWFA).\ -**congruent: activity increased**; incongruent: decreased Those with dyslexia: incongruent letter-sound pairs do not suppress neural activity for auditory-visual condition, enhancement in processing for congruent (still weaker than controls) Logographic stage: holistic visual stimuli -\> whole spoken words\ -like £ to pound **-high-risk dyslexic people**: significantly delayed brain activity (320 vs control group 210 ms) when seeing nonsense words\ -atypical activation in left inferior frontal gyrus when letter-sound task\ -little differentiation in activation time between occipitotemporal and temporoparietal regions (usually correlated with recoding print to sound for logographic stage) **Diffusion Tensor Imaging (DTI):** measure water in brain tissue (track axonal fibers)\ -white matter tracts "information highways" (axons connecting different neurons)\ -white matter integrity (axonal coherence and density), left temporal lobe, temporoparietal cortex = **word identification!** **Classic** model of visual word forms -- information flows from visual areas of brain -\> speech related area\ **Updated** model -- angular gyrus (spoken language areas) modulated/controlled dorsal occipital activity (so language/speech -\> visual) **Reading**\ 1. Orthography to Meaning (Spelling -\> Meaning)\ 2. Orthography to Phonology (Spelling -\> Pronunciation)\ 3. Phonology to Meaning (Pronunciation -\> Meaning) **Semantic-Relatedness Judgment task:** matching related vs non-related words, use ERP\ -degree of word knowledge influenced recognition of word\ -orthography-meaning condition most powerful recognition effect! **Morphological Processing\ Experiment: brain activation in contrast priming of word pairs** between\ a) opaque morphological relationship (e.g. archer, arch) or transparent morphological relationship (e.g. bravely, brave) VS\ b) meaning only (e.g. stop, halt) or form only (e.g. catalog, cat) Results: left frontal areas for morphological processing is **distinct** from word form recognition/word meaning identification **Syntax and Semantics** = top-down effect on word meaning (language meaning does not only derive from word meaning + grammatical marker)\ Syntax =identification of grammatical functions and interrelationship of words (left frontal gyrus/broca's)\ Semantics = intention of words, phrases, idioms (temporal-parietal, wernicke's and basal language areas) -for anomalies, are distinct operations\ -word by word basis, early brain activation is for word and morphosyntatic identification\ =linear theories of syntactic processing Consistent Activation across age and modality in left superior temporal sulcus = "comprehension cortex"\ -younger children: diffusely distributed activation pattern which included right temporal pole and right cerebellum\ -older children: left inferior frontal cortex for structurally complex texts **Emotion in meaning making**\ -anterior temporal, inferior prefrontal areas = regulation and socioemotional response\ -basal ganglia, endocrine system = emotional states for memory reconstruction\ -hypothalamus! **Higher order cognitive and discourse-level processes**\ -inference from textual info -\> prior world knowledge\ -text tracking, metaphor analysis **Experiment Results:**\ -hear text multiple times, neural activation from auditory+language comprehension brain areas -\> frontal, parietal and subcortical areas for modelling, memory and recognition (right parietal, discourse representation) Sentence-level comprehension vs Passage-level comprehension\ -similar brain area for sentences but distinct areas for situation model (posterior parietal for constructing model, anterior temporal for maintaining model) Story comprehension -- narrative structures and comprehension (medial parietal cortex) c. ***[Teaching of Reading]*** **Whole-word vs Phonics Instruction** a. whole-word = "look-say" method (50-100 words initially, then learned as wholes too)\ -congruent with spoken language as a continuous stream of sound\ -can be detected as meaningful units\ -avoid low reliability of letter-to-phoneme mappings\ -use contextual clarification b. phonics = graphemes (printed letters) and phonemes (associated sounds)\ -limited set of correspondences between letter and speech sounds\ -letters -\> consonant digraphs (th, ch) -\> consonant clusters (st, tr)\ -taught to blend letter-sounds, alphabetic principle, analytic approach\ -but would be boring! (most children unconsciously learn phonic patterns with oral language) **Word units:** -phonemes (b sound in bat)\ -syllables (crow, beau-ti-ful)\ -morphemes (minimal unit with meaning and grammatical form, "cook" in "cooking", "s" in "cooks" or "ing" in "cooking") Writing Systems -- **alphabetic, syllabary and morpho-syllabic**\ =different variations in orthography -- details of mapping between graphic and language units English -- **alphabetic**, where graphic units (letters) are associated with phonemes (pronunciation)\ -allows alphabets to be productive = few letters can already build indefinitely large number of words (like using t p s o for spot, pots, tops etc.) Japanese -- **syllabaries**, graphic units correspond to syllables. Chinese -- **logographic**, or morpho-syllabic, units correspond to specific words/phonemes; characters map onto syllable units/morphemes **2 major difficulties in alphabetic principle:** 1. **Abstract nature of phonemes** -problem in consonants more than vowels (vowels are easily heard/isolated in words)\ -position of letter may result in different phoneme/pronunciation 2. **Most alphabets don't code each vowel with a unique symbol** -e.g. saw and say, vowel has different sounds -Reading is a modification of mental lexicon for being **print addressable** = use conventional forms of printed language to obtain meaning from word **Language development**\ -phonology, grammar, word meaning, pragmatics basics acquired by age 4\ -basic syntactic structures learnt by age 2 (then understand semantic notions mapped onto syntactic structures)\ -mental representations of abstract phonological structure further refined when learn writing **Perceptual narrowing:**\ -newborns can discriminate phonemes(sounds) in all spoken languages\ -by 12 months they can only discriminate sounds of the native language\ =lexicon becomes larger, forcing finer discriminations, increased speech production **Naming explosion:** dramatic increase in word knowledge age 2 (while using multiple word phrases) -first grader higher **SES** \> double vocab size \> lower SES\ -social use of language (pragmatics) throughout preschool, basic conversational functions and speech acts Altogether = **metalinguistic awareness**: Linguistic subsystems, the morphological and phonological and grammatical! **Emergent literacy**\ -developmental continuity between cognitive tasks during preschool and learning to read\ -e.g. young child's activities around books -\> later opportunities for reading\ =literacy emerges in various forms in development before conventional reading+writing\ e.g. oral reading by teacher, idea sharing by children, writing, drawing, but not much on letter-sound relationships **Learning to Read Alphabetic Reading System** -differs in orthographic depth / consistency of mapping letter-sound\ e.g. shallow orthography=consistent mapping (finnish, Italian, dutch shallow than eng)\ -eng irregularities in graphemes and phonemes, but in short, high-frequency words\ -eng irregularities in larger, subsyllabic "rimes" (vowel+syllable ending) -important to have context-sensitive mapping between grapheme-phoneme (and larger units) e.g. know, now, low different pronunciations\ -more experience in orthography = reduce impact of dialect variations **Phonological Awareness**\ -knowledge of internal sound structure of spoken words\ -usually tied to reading than speaking (know read supposedly will know speak/hear)\ -acoustic phonetic info (voice onset time, lag between release of consonant and onset of vowel like /b/ to /p/) -IQ and early reading weak relationship, nonspecific in first second grades\ -many children can't read but have above-average IQs Theories of Read Learning\ 1. Stage theory: association between visual graphic forms and spoken words -\> graphic-phonological decoding -\> letter-sound associations\ -qualitative characterizations of changes\ 2. Incremental acquisition of individual word representations (and NOT discrete stages)\ -connectionist theory -- how stage theory changes are from basic mechanisms **Bootstrapping hypothesis**: decoding unfamiliar word is a self-teaching method 1. **Stage theory (Gough):**\ -selective association (visually printed word with name of word)\ -cipher stage (after reaching limits of associations, using alphabetic principle) 2. **Stage theory (Ehri)**\ -no purely visual stage, but use letter names to identify words (like j in jay helping to read jail)\ -phonetic cue reading: association between printed letter form and phonetic cues 3. **Nonstage incremental theory**\ -qualitative shifts in strategy (by the amount and complexity of info acquired)\ -progression from **logographic** (printed words associate to meaning and pronunciations) to **alphabetic** stage (letters/phonemes)\ this change in behaviour = **increased sensitivity to internal structure** of words and correspondences between subwords and pronunciations\ -acquisition of more word representations, accessed by **spellings, positional specificity and phonological redundancy (both quantity and quality of word representations)\ **-**functional** lexicon -\> **autonomous** lexicon -Phonological **errors** (mistaken similar sounding words) in first grade predicted end-of-year reading achievement\ -Nonphonological errors (like shared letters with phonemes at wrong position like vs milk) predicted low end-of-year achievement\ -when phonological errors \> nonphonological = functional phonological skill (know at least half of the alphabet with phonological sensitivity/awareness) **Phonological recoding:** spelling into pronunciations (self-teaching) **Fluency**: automatic word-identification (functional to autonomous lexicon)\ -input-output mappings practice\ -retrieving word forms and meanings (output) from printed words (input) **Spelling:\ **-easy to underestimate children's potential grasp on alphabetic principle / speech sounds are associated with letters (if only consider decoding)\ -(pronunciation -\> spelling) = less consistent than (spelling -\> pronunciation) (reading more reliable than spelling!)\ e.g. snear sneer snere all same pronunciation = backward consistency effect (phonology on orthography)\ Orthography -\> Phonology inconsistent: EER not always pronounced as EER\ Phonology -\> Orthography inconsistent: ERE EAR EER all can pronounce EER -spelling: production ; reading: recognition (success in reading not automatically = success in spelling) **Comprehension:** -language understanding, not unique feature of reading\ -skills in oral setting not directly transferred to spoken ver of written texts\ -correlated to comprehending spoken texts a\) Application of nonlinguistic (conceptual) knowledge\ b) Application of general language comprehension skills to written texts Syntactic parsing problems: processing capacity limitation, not intrinsic syntax problem\ -working memory factors produce individual comprehension differences Structure-building framework: activate and enhance relevant concepts while suppressing irrelevant ones\ -poor reading skills = deficient suppression mechanism **Dyslexia**\ **a) Developmental Dyslexia**\ -normal intelligence without sensory/neurological impairments\ -discrepant from potential implied by IQ **b) Acquired dyslexia**\ -brain injury -big problem: reading words and pseudowords (nonwords) -- sublexical processing **Causes:**\ -phonological-deficit hypothesis\ -abnormal hemispheric dominance\ -secondary to subtle impairments in auditory info (speech specific/general)\ -speech processing\ -some research show phonological knowledge is at normal range **Visual impairments:**\ -dorsal visual pathway by magnocellular in lateral geniculate nucleus (rapid changes in illumination, motion) **Developmental delay:**\ -environmental insufficient reading experience, inefficient teaching... Reading was viewed as **psycholinguistic guessing/hypothesis-testing** activity\ -figure out meaning by 3 cuing systems: semantic, syntactic and graphophonic\ **Graphophonic** = general knowledge of spelling-sound relations\ **Syntactic**= Syntactic patterns and markers like function words and suffixes\ **Semantic** = Word meanings and topic Eye-movements in reading: like a slide show\ -**fixations**: 200 to 250 ms (acquire info)\ -**saccades**: 20-40ms (suppressed vision = no new info)\ -**regressions**: re-read material **Acuity Limitations** in visual system (best in center of vision (fovea)), so need to move eyes back and forth\ -**Perceptual Span** (which useful info is acquired) is restricted Low frequency words fixated longer\ -unskilled readers have very short saccade, very long fixation, more regressions, smaller perceptual spans Identify words as wholes? (**parallel**?) or letter by letter? (**serial**?)\ -**word superiority effect:** letters in words identified more accurately than in isolation\ a) serial cannot be correct, must be parallel identification\ b) all letters must've been processed Word meaning: directly from **print** (direct access) vs **letter-string phonological** code (phonologically mediated access) Direct doesn't always work:\ -spelling associated with meaning: very arbitrary\ -unfamiliar letter-string = meaning not yet established\ -if can "sound it out" phonologically recode letter string, can be derived from spoken language\ =recoding is self-teaching mechanism Division of labour between two mechanisms:\ a) a race between them, whichever is fastest (depending on familiarity, spelling-sound consistency, reading skill, writing system etc.)\ b) complementary Brain parts!\ -movement of mouth in **oral** reading: primary **motor**\ -**phonological, orthographic and semantic** in word identification**: inferior frontal** cortex, **left temporoparietal** cortex, **left basal temporal** cortex\ -orthography/object **recognition**: **occipital**\ -**word-form area**/printed words/non-words/pseudowords**: occipito-temporal** border, **fusiform gyrus** Phonological deficit may have impact deep orthography (eng) \> shallow ortho (Italian)\ -but shallow ortho perform better in pronouncing words + nonwords **Connectionist** model: both constitution and experience\ -knowledge can deviate from intuitive cognitive phenomena\ -words are not entries but patterns of activation\ -basic components of knowledge information processing + representation\ = distributed representations, in which spelling/sound/meaning are small sets of units in many words!\ -include hidden units for more complex mappings between codes =letter string input, activates units then spread to other info, adjusting weights based on experience\ Backpropagation: output is compared with target pattern **Traditional**: cuz irregularities in phonology, phonological recoding doesn't always work\ **Connectionist**: gave and have = irregularity in "ave"; but same mechanism in learning overlapping words have, had, has... ("inconsistent words" have own patterns) **Orthography-phonology mapping training**\ a) phonological knowledge from speech\ b) orthography-phonological system; clearer representations of subword units like phoneme, onset, rimes\ -**important**: not having full phonemic representations before reading, but having the **capacity** **to** **develop** such representations **Phonological pathway**: predominate in early reading!\ Orthography-semantics association too arbitrary, only for cleaning up patterns activated by phonology. -Learning to read vs speak can be different, not always automatic and effortless\ -Some say no child needs to be taught phonemes, but need the symbols that make up writing system, have alphabet song but not phoneme song **Direct teaching/Prescriptive Teaching: scope and sequence**\ - basal reading emphasizes whole-class instruction and independent practice, not much assessment\ - Reading Mastery: placement tests and in small groups\ - Success for All has whole class, partner reading, independent... **Prescriptive techniques**: review letter sounds, introduction, practice blending of sound into words, decodable texts, read-alouds, language arts...\ e.g. read decodable text: children signal thumbs up when hear a phoneme/consonant in text, then add more blended words so harder to differentiate **Responsive Teaching: Scaffolding (good for heterogeneous readers)\ **-use three cuing systems from inherited oral language abilities\ -Reading Recovery: provide feedback on errors, respond errors and breaking words down\ -but lack of sequential instruction+practice makes it difficult for transferring decoding skills\ -Guided Reading: whole class discussion and small group, independent reading discussions\ -leveled text: ordered by difficulty, in contrast with decodable texts as they are selected for sense of story and predictable syntactics, selected for word frequency not sound-spelling\ -activities: analogical reasoning around onsets and orthographic rimes **Lab studies:\ **-letter-sound correspondences are productive, than whole-words (even though letter-sound slower at first)\ -direct teaching might be more efficient than inferring teaching\ -letter-phoneme group read more new words than whole-word group ***[D) Basic Calculation Proficiency and Mathematics Achievement]*** ================================================================================ **Traditional education view** on basic calculation proficiency:\ -solutions stored and readily retrieved in LTMemory (memorization + practice)\ -discourage finger use -- indicates reliance on backup strategies **Conventional Wisdom** **view** on basic calculation proficiency:\ **-conceptual knowledge** for computational fluency\ -three phases: counting fingers -\> use principles/knowledge for decomposition (8 is 5+3) -\> retrieve **Number Sense view** on basic calculation proficiency:\ -accurate solution by ANY efficient strategy, **conceptual knowledge**\ -understanding number operations, patterns, principles Conceptual knowledge types: **Natural number system** and **knowledge of calculation principles** Traditional + Conventional Wisdom: knowing every solution to basic problems (number sense view does not) -\> accurate retrieval correlated with math achievement **Assessing** basic calculation proficiency:\ a) Strategy Assessment Tasks: any method to solve problem\ b) Forced-Retrieval Tasks: retrieval within 3secs Challenging views:\ a) retrieval not that frequently used in single additions!\ -**perfectionist** use **less retrieval**, but whenever they do, is extremely accurate (due to **higher confidence threshold** for retrieval)\ -decomposition correlated with **faster** responses (forced retrieval reflect retrieval + decomposition) b\) accuracy may increase when using fingers!\ -accurate solutions increase likelihood of subsequent retrieval (so both methods can be interrelated, opposing traditional views) c\) children use **multiple strategies** across problems\ -no fixed sequence like conventional wisdom suggests\ -decomposition -\> found to have higher accuracy Other **cognitive functioning** influences: a. Working Memory\ -phonological loop (PL) and visuo-spatial sketchpad (VSSP), together -\> central executive (CE)\ -test PL&VSSP: simple span tasks like digital recall / Corsi blocks\ -test CE: complex span tasks like storage and processing e.g.backward digit recall b. Processing Speed c. Oral language (phonological processing related to basic calculations) d. Literacy ***Aim of study:***\ a) development of basic calculation proficiency from 2^nd^ to 3^rd^ grade\ b) how much proficiency relates to conceptual knowledge and cognitive factors ***Methods*:** both forced retrieval and strategy assessment task -- code behaviour and self report methods ***Results:**\ *-retrieval = **least common** strategy\ -counting = most common in 2^nd^ grade, **lowest accuracy**!\ -decomposition = most common in 3^rd^ grade -high degree of **consistency**: 2^nd^ and 3^rd^ grade both same strategy for same problem\ -supporting conventional wisdom: counting-\>decomposition-\>retrieval\ -less supporting conventional wisdom: **variability** in strategies -correct retrieval in strategy assessment = less common than rapid correct in forced retrieval\ -**rapid** solutions in forced retrieval **NOT just from retrieval** (many from decomposition too)\ -retrieval not always rapid -Forced retrieval -- reflect higher knowledge levels than strategy assessment task\ -**forced retrieval overestimates knowledge**, as many answers come from decomposition instead of solely retrieval ***Discussion:***\ -conceptual knowledge linked with basic calculation skill (limited though)\ -even with imperfect basic calculation solutions, math achievement can be excelled\ -math and reading are mediated by many factors Adaptiveness:\ -variability is a fundamental characteristic of computational strategy development\ e.g. finger counting, even discouraged by traditional views, help accuracy\ -choices migrated in the direction of efficiency and accuracy\ -30% questions solved by retrieval in grade2 is not solved same way in grade 3: reflect variability of strategy/experience, maybe other challenges in multiplication/division, classroom environments -rapid decomposition solutions -- evidence of applying "combination knowledge" and "conceptual knowledge"\ -both decomposition + retrieval -- consistent with Number Sense view of proficiency\ -basic calculation proficiency predicted later concpetual knowledge -processing speed very limited effect/explanation for basic calculation, but oral language and working memory mediated a little\ -VSSP and CE contributed more than PL measures in working memory (mediating reading and math, as well math itself)\ -better readers are superior at maths not just because of language and memory skills -**ignoring basic number combinations NOT barrier** to math achievement (contrary to traditional and conventional wisdom view)\ -**conceptual knowledge does NOT completely mediate** basic calculation and achievement (contrary to number sense view) ***[E) Everyday conceptions of object fall: Explicit and tacit understanding]*** a\) moving objects predicted to fall vertically/ diagonally/ travel backward/ continue horizontal\ -seldom predicted in its actual parabolic paths b\) heavy items predicted to fall faster, while it should be same max velocity -people usually do better in recognition (tacit) than prediction tasks (explicit)\ -**explicit** (prediction tasks) **need conceptual knowledge** (scenarios related to underlying relations to draw inferences) = **deliberation, reflection**\ -**tacit** (recognition tasks) only need scenario-concept **matching** (no need consideration/inference) = more unconscious, no need processing Omission Hypothesis: reason why there is explicit-tacit processing gap\ -omission at the explicit elements which are tacitly appreciated\ e.g. not considering forward velocity when predicting direction, or considering one moment only (instead of across time) when predicting speed Lit reviews on usual predictions:\ -vertical fall associated with heavy objects\ -continued horizontal then 90degree fall associated with light objects (esp with rapid pre-fall velocity like bullets) -from 2y.o., less willing to accept vertical fall compared to parabolic\ -but no rejection at any age of horizontal/diagonal motion -- still incomplete understanding STUDY 1 (DIRECTION) Results: 1. Prediction task\ -stationary dropping easier than moving-dropping scenarios\ -in stationary: performance improved with age, varied with type of ball\ -moving: poor performance in general (no correlations with age/ball) 2. Recognition task\ -stationary easier than moving\ -recognising correct motion easier than incorrect ones\ -correctness x medium x ball significance -proved consistency in omission hypothesis (lack of conceptual understanding, omitting explicit information)\ -backward and forward errors exist in all ages but backward increased, forward decreased with age (usually in moving prediction tasks)\ -regardless of age children believe stationary balls fall vertically, sometimes deflected. 1. Prediction task\ -poor performance with air-only scenarios, better in air-plus-water\ -medium x ball and medium x ball x age interactions 2. Recognition task\ -better performance in air than air-plus-water\ -medium x correctness and medium x motion interactions\ -significant improvement with age\ -ball x correctness, ball x correctness x medium, ball x correctness x motion and ball x medium x age -children: all errors involved predicting deceleration (but they judged it incorrect in recognition tasks), but undergrads: constant speed errors instead\ =**developmental change (not improvement) with age**\ -**recognition** tasks may **display conceptions reverse** of those in **prediction** General discussion:\ -**prediction** task accuracy (STUDY 1 + air-only in STUDY 2) **\< recognition task accuracy**\ -predictions improved with **age**\ -recognition marginally **better** in **correct** scenarios than **incorrect** ones\ -**insufficient** evidence for **omission hypothesis:**\ -conceptual elements were guiding predictions which were not used in recognition = **inclusion at the explicit level, not omission!\ **-some task performances decreased with age -hemispheric differences for causal perception (tacit) of motion and causal inference (explicit)\ -media images/sociocultural representation may be a cause of why prediction of backward trajectories increase with age, prediction accuracy decreases\ -initial understanding may be more precocious than later ones ***[F) Sensori-motor experience leads to changes in visual processing]*** -**functional specialisation:** different areas of brain have specialised focus for stimuli (emerges from extensive experience)\ e.g**. single-letter** and **word perception** in adult visual system\ -**anterior** **left fusiform gyrus** for individual letters than letter strings, words, digits...\ -**posterior** left fusiform (Visual Word Form Area (**VWFA**)) for strings/words than individual -overlap between VWFA and Lateral Occipital Complex (area for objective selective) -specialisation can be **category specificity** (e.g. this area = for face recognition), but can also mean specialised configural process for whatever object (**skill/process specificity**)\ e.g. letters perception can be category specificity (letter area), or analytic skill specificity How we process words/reading:\ -physical interaction with environment help visual processing\ -not encapsulated modules but receiving feedback constantly\ -integration across sensory/motor systems\ e.g. history of writing letters stored -\> re-activated when visually perceive letters (even pseudo-letters) **This experiment:** use short cartoon and have two groups of training: visual only (only read/identify) and visual+ motor (identify+ copy letters)\ -through short story while viewing text **Results:**\ -imaging data shows b4 training, **left fusiform gyrus** activated during letter perception\ -after training, activation increased even more! (not due to familiarity, cuz only for sensori-motor training and not visual)\ -**right anterior fusiform gyrus** also activated after sensori-motor training Discussions: -even **for pre-literate** children, there are **hemispheric differences** for **letter vs shape** perception\ -**anterior left fusiform gyrus**, more activation to letters than shapes/pseudo-letters\ -**right fusiform** similar activation to letters, shapes, pseudo\ -hemispheric specialisation may be due to familiarity with stimuli\ -sensori-motor training increased activation in *visual* area\ -saying word out loud also a motor response (in both trainings), so may be manual motor works instead of oral motor here\ -visual training no increase VWFA activation, but sensori-motor triggered VWFA -perceiver as an active participant for embodied cognition