PSY2002 SEMESTER 1 - WEEK 3 Flashcards
what is reading in a sciency way
transmitting inputs in form of words into comprehension and meaning
input activates existing mental representations of written words, which is linked to meaning and causes output = comprehension
outline mental representation for comprehension (what we hear- speech)
input (speech) activates existing mental representations of sound and links to meaning, causes output (comprehension)
outline mental representation for reading (what we see- writing)
input of written word, activates existing mental representations of written words, links to meaning
causes output of comprehension
name the building blocks for written word comprehension
written word input
form (phonology) + form (orthography)
syntax + morphology
semantics
what is orthography
word form when written down, mostly represented by alphabetic graphemes (graphemes represent phonemes)
iogographic systems use characters to represent whole word (eg; Mandarin)
some languages can be alphabetic languages that look like logographic system (eg; Korean uses letters that represent words but are grouped by characters)
what is an issue of research into orthographics
- anglocentric, focuses on English-speaking POV
- debated topic- do we access both phonology and orthography during reading or just orthography (some think this happen at unconscious level and accessing phonology aids understanding)
name the DRC route of comprehension
written word input - activate letters - activate orthographic form - activate phonological form - semantics
outline development of phonological process for children
children speak years before read and taught to read using phonetics to form connection, eventually sentence
what is a phonological neighbourhood and apply to reading
differ in only 1 phoneme (gait and bait)
if reading involves phonological processing, word with many neighbours has advantage as fixated for less time but if closely related slow down word recognition as all activate, compete
what can phonological priming be used to assess, and how does it do this
assess role of phonology in word processing, showing words processed faster when preceded by phonological identical non-word primes than control
suggests phonological processing rapid, automatic
name 2 computational models of reading aloud
- dual-route cascaded model
- distributed connectionist approach/triangle model
what is main focus of dual-route cascaded model
reading word/nonword involves different processes
what is main focus of distributed connectionist approach/triangle model (opposition of DRC)
reading processes flexible, involve interactive processes with all relevant knowledge on word sounds, spelling, meaning used in parallel
outline DRC model of visual word recognition and reading aloud
lexical and non-lexical routes used in parallel if reading aloud, with both routes not functioning independently
cascaded, as activation on one level passed onto next before processing at first level complete
in DRC, how are connections adjusted and what do they do
excitatory connections motivate process
inhibitory connections stop process
adjusts strength of connection, provide input, processes input into model, assesses output of model against human performance
evaluate computational model in general
allows us to ask very specific questions and test results but rarely can model all variables and parameters at work, not likely to get absolute answer, but good for testing specific questions
in DRC, what does non-lexical route do
convert letters into sounds to activate phonological representation that links to meaning
in DRC, what does lexical route do
activates orthographic representation (stored written word) that is linked directly to meaning
when is Grapheme Phoneme Correspondences GPC (Spelling-to-Sound Correspondences) used in DRC
when processing via non-lexical route must go via GPC
what is a grapheme, how do they relate to phoneme?
single grapheme represents single phoneme and grapheme can be made up of a number of letters eg; Th, igh, ough
same grapheme can be used to represent >1 phoneme (i in mint/pint) and single phoneme can be represented by >1 grapheme (k = c,k,ck)
define regular words
follow set of rules that dictate how grapheme should be pronounced
explain regular and irregular words and how they are created regarding graphemes
lots of variety in how phonemes are represented, via graphemes
mint is regular, pint irregular
both contain grapheme ‘i’ but prounounced differently
what is transparent language
also known as shallow orthography
spelling of each word maps directly onto its pronunciation (ie, Finnish/Italian)
orthographies with lots of reuglar correspondences (sound as they are written, eg; dog)
is transparent or opaque language easier to learn
children learn transparent language, eg; Italian quicker
what is opaque language
orthography with few regular correspondences eg; yacht
spelling of each word does not map directly onto its pronunciation eg; English
if written word is mint, how would a lexical, or non-lexical process this (DRC)
via GPC if non-lexical
/m/ /i/ /n/ /t/
if lexical, simply
see /mint/
when would non-lexical route not be able to be used (DRC)
irregular word ie; pint
/p/ /i/ /n/ /t/
p’in’t not a word (if said as written)
need to use semantics via lexicon to understand what a pint is to then say it correctly
what route does surface dyslexic rely on
route 1 (non-lexical) GPC
have problem reading irregular word
name 4 advantages of the DRC
- allow 2 routes of processing written word, so account for differences in learning
- account for orthographic lexicon, phonological lexicon
- account for processing regular, irregular word
- account for encountering new/novel word, can be processed via grapheme-phoneme correspondence
give 5 limitations of DRC
- focus on reading of individual words but real life is in sentences
- deemphasise semantic processes in reading (occurs in temporal lobe)
- cannot explain how children learn grapheme-phoneme rules
- assume phonological processing of word is slow with little effect on word recognition or reading
- lack applicability to Non-English language
what is understanding spelling to sound correspondence crucial for (grapheme-phoneme correspondence)
learning to read in alphabetic writing systems
outline self-teaching hypothesis (Share, 1995)
children ‘decode’ words using understanding of how letter correspond to sound (phonics)
explain process of self teaching hypothesis and how phonological representation is used to learn words
- existing phonological representation accessed, used to access phonological lexicon meaning
- phonological representation used to develop orthographic lexicon of whole words (look up phonological lexicon and see correspond to words we read)
- this is why children initially sound out all words they read
read /m/ /i/ /n/ /t/ and learn mint
DRC was applied to self-teaching hypothesis to create self-teaching DRC. explain this
non-lexical route used to decode words, access existing phonological representation
contextual cues used to select target word from list of candidate words
what is necessary to develop orthographic lexicon
consistent readign
how do skilled readers build up their knowledge of words (self-teaching DRC)
generate phonological representation that accesses semantic system
starts to build orthographic lexicon
more sounds we learn more words we build up
know bath, can add, grass, raft
skilled reader bypasses non-lexical and accesses semantic directly via activation of orthographic representation in orthographic lexicon
how does DSM-5 define dyslexia?
difficulties in accuracy or fluency of reading that are not consistent with persons chronological age, education, intellect
define dyslexia by British dyslexia association
specific learning difficulty that mainly affects development of literacy and language related skills
characterised by difficulties that may not match up to an individuals other cognitive abilities
how common is dyslexia
1/10
what did Snowling suggest dyslexia was
learning difficulty with problems in phonological processing, understanding how speech sound correspond to letter
not visual problem as have difficulty identifying phonemes (saying word skip without p is difficult)
gene x environment
name 3 aspects dyslexia can be identified on
- decoding skills
- lexical retrieval
- verbal STM (word span, digit span)
decoding skills can be used to identify dyslexia. what tasks may this include
phoneme deletion/substitution tasks, decoding pseudowords (non-words)
lexical retrieval can be used to identify dyslexia, what tasks does this include
RAN (rapid automatic naming tasks), word identification of regular/irregular words
RAN- presented with list of objects, name as many as can
outline the dyslexic profile - 2 components with 2 subcomponent each
- poor phonological awareness (problem with identifying phonemes, reading non-words)
- slow lexical retrieval (RAN task shows slow retrieval of letters, slower word reading)
dyslexia - how are phonological deficits initially presented, and what can this cause in later life, and why
initial problem linking phoneme and grapheme, cause problem with word reading later in development (slower identification and reading of correct words in later stage of development)
later life reading issues as issues when learning causes less robust orthographic lexicon (less fluidity in reading)
how can phonological deficits in dyslexia lead to social continuation factors
difficulty decoding causes less motivation to read
reading facilitates dev of orthographic lexicon and facilitates skilled reading
what can poor orthographic learning cause (dyslexia)
unexpectedly poor speller, or surface dyslexia
what is surface dyslexia
typical decoding, reading speed but spelling difficulties
unimpaired phonological awareness
non-word reading in normal range
impaired irregular word reading (break read as breeeek) - impaired lexical, intact nonlexical
unable to distinguish between homophones
outline Wybrow (2015) test of lexical route deficits in dyslexia
surface dyslexia shows higher proportion of regularisation errors for irregular words than control
phonological dyslexia shows lower proportion of regularisation errors for irregular words than control
outline Peterson (2013) study into RAN in dyslexia
phonological awareness = phoneme deletion
RAN = images and colours
orthographic coding = which of following is flower, rose or rows
found phonological dyslexia have better phonological awareness than surface dyslexia
but surface have better rapid naming and orthographic coding
briefly state surface dyslexia
deficit to lexical route
no problem reading regular non-words
problem reading irregular words
briefly state phonological dyslexia
deficit to non-lexical route
problem reading non-word
impairment of non-lexical (GPC) route
why is evidence of dyslexia subtype debated (following review of 5 studies Sprenger-Charolles & Serniclaes)
results more in line with hypothesis that phonological deficit at core of developmental dyslexia, than with idea that clear dissociation exists between surface, phonological profile
what compensatory mechanism may children with dyslexia show
relies on semantic processing
if struggling using nonlexical route to decide word may still access some phoneme in word resulting in activation of list of candidate words that could fit partial decoding
then use context to disambiguate situation
outline research into impact of context for lexical access
ppts start repeating specific words from recorded sentence 200ms after word onset, but only if sentence makes sense
language system predict what word could come up, activated in lexicon
Frith & Snowling (1983) tested childrens ability to correctly read out loud sentences ending in homograph
what did they find
homograph = spelled same, pronounced differently
before he made a speech, gave bow
dyslexics more likely correctly pronounce bow than control, shows more accurate prediction of what word comes next
whats a homograph
spelled same, pronounced differently
Nation & Snowling (1998) studied reading words in context, what did they find
read following word “aunt” out loud
I went shopping with my mum, and ___
dyslexic use context to process both regular, irregular words more than normal
normal uses context more for irregular compared to regular word reading, which makes sense as irregular are harder to read due to unable to use nonlexical route
how can you test semantic priming effects for predictive processing
if individual has stronger predictive processing and links between semantic concept = should have larger semantic priming effects (calculate via subtracting reaction time for related from unrelated conditions)
how do dyslexic readers use context in priming tasks (Van De Kleijj, 2019)
showed picture prime of car, word target of bicycle
dyslexics able to use context to greater extents
how can automatic word processing be tested (RAN) what is found with dyslexics
name images, letters, digits as fast as can
dyslexic and weak reader slower to name than control
dyslexics name less objects, less correct words than control but doesn’t mean cannot do task, just with less fluency, speed
how do dyslexics perform on phonological awareness (Nation, 2019 learning to read longitudinal data set)
poorer performance on measure of phonological awareness at 7 than 5 (on both phoneme isolation, and non word repetition)
when do phonological awareness deficit appear in dyslexia
not in kindergarten but after 1st year of reading, disappear at end of primary schools
outline recording eye movement as a research method
provide unobtrusive, detailed record of attention-related processes, but deciding what processing occurs during each fixation is issue
outline naming tasks
says printed word out loud rapidly as can
emphasises links between orthography and semantics
outline lexical decision tasks
rapidly decide if string of letters form word
what are saccade-regressions
eyes moving backward in text
what is perceptual span in saccades, what is capacity, affected by, use
effective field of view = 3-4 letters left of fixation, 15 to right
affected by text difficulty
size of perceptual span mean info from parafovea (area surrounding fovea) used during reading
explain E-Z reader model
assume mind and eyes tightly coupled, so pattern of eye movement give info on reader’s processing strategies
attend to 2 word in 1 fixation, single processing model (only 1 word processed at time)
what does E-Z reader model assume
we check familiarity of word thats currently fixated, completion of check signals initiation for eye movements
then engage in 2nd stage of lexical access to access semantic and phonological form
is faster when word easily processed
completion of lexical access shifts attention
give support for E-Z reader model
common words fixated less time
words following rare fixated longer as receive less parafoveal processing when words rare
identifies factors like word frequency, predictability as determining eye fixations in reading
give limitations of E-Z reader model
downplay role of higher-level processes (info being integrated in sentence, using schemas)
more parallel processing than acknowledged
what is parallel processing model SWIFT (contrasts with E-Z reader model)
assume duration of eye fixation in reading influenced by parallel processing of previous, next word alongside current one
in word recognition, outline interactive model
word superiority effect mean performances better when letter string forms word
assume top-down process from word level to letter level activates letters
orthographic neighbours facilitates word recognition when less frequent, and inhibit if more frequent
give limitations of interactive model of word recognition
no account of role of meaning in visual word recognition, phonological processing often involved in word recognition but not in model
give strengths of interactive model of word recognition
applications to visual word processing, account for word superiority effect, effect of top-down
how does semantic priming cause word to be recognised quicker
due to context automatically activating stored representations of all words related to it, or controlled processing of us expecting related words
define sentential context effects
using context to predict next word so context affects early processing of it (top down processing)
define deep dyslexia
problem reading unfamiliar words and inability to read non-words
