Infant perception 2 Flashcards
What is the difference between behavioural and imaging techniques?
behavioural require action, imaging is passive- sometimes hard to show novelty or preference
Behavioural (5) techniques
high amplitude sucking (HAS) conditioned head turn (CHT) head turn preference (HTP) preferential looking eye tracking
HAS
suck triggers sound, suck less as habituate to sound then change makes rate change if perceived
+: good for newborns, active indication of interest
-: high attrition and if increased sucking rate- lost attention or not perceived?
CHT
suitable only for older babies; requires movement, brightly coloured toy to keep attention straight ahead, then conditioned to turn to particular stimuli (eg speech), reinforced by picture//toy during conditioning phase
eg Kuhl 2006 144 discounted out of 170 due to time
HTP
child looks between different lights; sound played while looking at to light, stops when stops looking. time spent looking indicates preference of sound – or does it show novelty?!
Preferential looking
measure total looking time and decide which one is preferred
- familiarity or novelty? McMurray and Aslin 2005- inconclusive results trying to distinguish
Eye tracking technique
Quantitative measure of attention to different stiuli or actiosn presented on the screen (pref looking)
- non invasive and used for range of experiments
Imaging (3)
ERPs
fMRIs
Baby MEG
- reduced in cost and availability, less attrition, effortful to run and highly trained, but exps are highly controlled
ERPs
Record brain electro activity as EEG, picked up on scalp and amplified with excellent temporal resolution
- odd ball paradigm; low prob mixed with high prob intermixed. evoke potentials, not spontaneous- correlate directly with stimulus
ERP responses
MMN: mismatched negatively; discriminate neurally- subconscious results
P300: reflect processes involved in stimulus evaluation and categorisation
N300: reflect unexpected result
+: passive, all ages, can compare, spectral waves can be directly linked to cognitive processes
-: poor spatial res, difficult to get info to regions specifically
fMRI
difference in test conditions by blood flow to particular region- very good spatial resolution BUT total absnece of movement, noise or blinking- not suitable. but some success with young children: Biasill (3m to emotional sounds)
Baby MEG
magnetoencephalography- electrical activity up to 100,000 neurones in a specific area
+: combines spatial res and much better than fMRI, non invasive, can be used when sleeping and infant can move head within helmet, can be used for interaction (eg washington labs)
-: not very available,
perceptual pre-requisites of speech learning
high sensitivity to..
sound distinctions for phonemic contrasts, congruence between auditory ad visual cues
- sub-phonemic differences that may be relevant (eg allophonic distinctions)
perceptual pre-requisites of speech learning
low sensitivity to…
variance in speaker tone (perceptual constancy), speech sounds not relevant to ambient language, within category variation
are these present from first weeks?
innate categorical perception= high sensitivty to phonemic distinctions by 6m.
perceptual constancy at 4-6m, 4.5m speechread,
Eimas 1971:
HAS in 104m infants, hear cross cat differences but to within category with longer VOTs- high sensitivity to phonemic distinctions and low sensitivity to within-category distinctions
thus INNATE CATEGORICAL PERCEPTION
Kuhl 1979
Evidence of low-sensitivity/ perceptual constancy: HTP, 4-6m, for vowels; transferred to new tokens differing in speaker and pitch contours- equivalence class
Kuhl and Meltzoff 1982:
attunement to AV correspondences using preferential looking for 4.5m: spend longer looking at face matching articulation
Werker ad Tees 1984
Evidence of attunement to native phon cats at 6m desensitise to non-native changes
- infants are born able to discriminate all contrasts. neural commitment to native language occurs at 6-12m
evidence against presence in first weeks:
adsf
animals:
Kuhl and Miller 1978;
Eimas 1971
Kuhl and Miller 1978; chinchillas can discriminate human phonetic barriers
Eimas 1971: basing this evidence on cochlear rather than innate lingustic neural structures.
lack of early sensitivity to native: Eilers and Minfrie 1975
evidence of early sensitivity to native contrasts- counters universal sensitivity
further attunement: Kuhl 2006
evidence of further attunement to native phonemic contrasts with experience: 6-12m CHT, taught 32 infants in japan and usa r/l contrasts… shows neural commitment to native language
ongoing sensitivity to some non-native contrasts:
Meta analysis in 2006 of 9 experiments on contrasts- not conclusive.
7 studies decline in performance with age; 2 didn’t
No decline for US infants discrimination in african clicks (Best, sithole 1988)
No decline for french infants tested on non native contrasts (falsified hypothesis) (polka et al 2001)
Early sensitivity to within-category distinctions
at 3-4m infants show graded perception of VOT (miller and eimas 1996)
at 6m discrimination of within-category difference along VOT continuum when tested with sufficiently sensitive tasks (mcmurray and aslin 2005)
mechanism?
statistical learning; Maye, werker and gerken 2002: tested 6-8m using preferential looking, only children from bimodal exposure group discriminated
SO infants learn phoneme categories via SL sensitive to distributional properties of phonetic input
