Word Reading Flashcards
why is reading important?
important cognitive skill when most information is text based
__% of the UK are functionally illiterate
16
__% of 15y do not attain minimum proficiency levels
17
What is the economic and social cost of low literacy
higher likelihood of major depression, substance abuse and poorer physical health
Whats a possible reason as to why England now has a rise in reading score levels
because reading instructions now are informed by insights from reading researchers rather than politicians
what is the first stage of the reading process?
visual word recognition
involves going from letters to the meaning of a word
what are mental lexicons?
-the systematic organisation of words in our brain, approx. 60-70,000
- dedicated to strong lexical info
What affects word recognition speed?
-Frequency
- Age of acquisition effect
frequency effect
HF words are recognised faster than LF words, providing evidence of a general information-retrieval mechanism
Frequency effect study
Hulme et al 1997
- found an advantage when recalling high frequency words in short memory tasks than lower ones
Problems with the frequency effect
- not really a clean measure of lexical access
-less frequent words and less familiar, meaningful and more non-word like
-if more like non words it is more difficult to say it is a word (post lexical factors)
age-of-acquisition effect
-words learned at a younger age are recognised faster
- can overlap with frequency but exceptions like ‘kite’ where kids would be faster than adults at recognition
- because adults dont use it as often
Orthography
- letter positioning
- affects word recognition
Boundary change paradigm
- technique means when people read the sentence there is one target word that changes (2-3ms) after crossing the boundary
- left of the word= jumbled up letter
-closer to the invisible boundary= word change into real word
-test to see what info was extracted compared to controls
-most people dont notice this
Transposed vs substituted letters (Rayner et al. 2006)
-found substituted letters had more disruption than transposed letters
-transposed letters still costly compared to controls
-The beginning and end letters of the words are more important in relation to recognition (transposed)
- Also a difference in activation for consonants and vowels
what are graphemes?
letters and letter groups that correspond to one sound (phoneme)
What did Rey et al, (2000) find about graphemes
People were better in recognising the A in brash than in board
- This is because the ‘A’ is part of the grapheme ‘OA’
- In order to recognise the A in board, we have to take apart the O and A to find the A
- Shows people use graphemes rather than letters in order to get to a word
how are graphemes processed?
as perceptual reading units, because multi-letter graphemes take longer to process than single-letter graphemes
what are morphemes?
the smallest meaningful unit of language, which might include prefixes and suffixes
- can be a single word (deck) or prefix (un)
Compound decomposition
-we take compounds apart when reading almost immediately
-the frequency of the individual morphemes constitutes will have an effect
- e.g. so farm/house would be faster grave/yard
Whole compound frequency, when does this take place and what does it suggest
- after the effects of lexeme constituent frequency
- So, when we read compounds, we initially break these up then put them back together
What is a pseudo suffix
Affixes that look like real ones but aren’t
- e.g. ‘ing’ is an affix ‘runnING’
- But in a word like swing, ‘ing’ is not an affix even though it looks like one
psuedo suffix in word recognition/decomposition
suffix and psuedo suffix are extracted/decomposed in units early in word recognition
semantic effect
-words that are semantically richer are recognised faster
- more semantic neighbours –> higher imagability/more concrete
How does information flow in recognizing words
- there is a feedback from words (level) back to to letter (level)
- letters in words are detected better then letters in non words
Word superiority effect
we are better at detecting letters in a word rather than letters on their own/non word
models of word recognition
- foster’s search model
- morton’s logogen model
- R&M interactive activation and competition model
foster’s search model (simple model)
- recognise letters
- access some initial unit e.g. first letter,meaning, grapheme which chooses a related bin (‘e.g. C’)
- once in a bin, entries are searched serially for a match, ordered by frequency
- master files are used to get to the meaning of a word
what effect can the fosters model account for
the frequency effect but not the word superiority effect
Features of morton’s logogen model
-parallel access/activation of all word detectors
-no word to letter feedback (one way, features> letters> words)
morton’s logogen model
- sends information about letters to all word detectors (logogens) at once
- each logogen has a threshold based on frequency and prior exposure
- threshold decides how much information is needed for the word to get activated (e.g. cat wont need much) - when activation passes the threshold, the word is fired and recognised
Comparison of morton logogens model to forsters search model
- COW
- with ML model, all letters will activate as this model is PARALLEL
- in the serial model, only C could be detected first
morton’s logogen model: what happens when not enough information has accumulated?
the letter/word is not recognised
HF are recognised faster as their recognition thresholds are lower
What does the new logogen model suggest
- logogens for different modalities
- expands from visual and auditory logogens
- e.g. reading, witting logogens
Features of the R&M IAC model
-parallel access/activation of all word detectors
-word to letter feedback
R&M interactive activation and competition model
- we analyse and activate all letters relating to word (parallel search)
- if we partially recognize a word we will have feedback back to the letter level
3.in order to recognise a word, we must inhibit other words that may be right, more info we get= more words inhibited
High frequency words in the IAC model
- higher resting levels of activation, however all words have the same threshold to cross for recognition
- differs from logogens model
how does feedback help in the IAC model
- letter perception is faster/stronger for letters in active words because of feedback
IAC model eval
+ word superirority effect
when a letter is presented in a word, all the relevant detectors are activated, which makes it easier to recognize
- accounts for pseudowords (Ripamonti)
2018)
regular vs irregular words
- regular= pronunciation is predictable based on spelling (e.g. sat)
- irregular= pronunciation is not predictable based on spelling (e.g. yacht )
dual route model
involves a nonlexical route for novel words and a direct lexical route for known words
nonlexical route
- early visual processing and letter recognition
- grapheme-phoneme conversion system (GPC) to turn letters into sound
- phonological encoding
- speech
What can the non lexical route be used for
-pronouncing new regular words as well as existing words
What can the direct be used for
-irregular words as the GPC rules may result in mispronunciation
- also frequent words
lexical route
- early visual processing and letter recognition
- orthographic input (dictionary for written words)
- semantic and phonological output lexicon to turn words straight to sound without converting individual letters to sounds
- phonological encoding
- speech
following damage to the lexical system…
-patients will have difficulty reading exception words, but no issue with non-words or novel words
-this is because reading will always occur via GPC- causing mispronunciation to irregular words
- Impairment of irregular words, making them regular (regularisation)
- pint to rhymes with ‘mint’
surface dyslexia …
patients struggled to recognise exception words, but had no issue recognising regular words and non words
this is because their nonlexical route was impaired
acquired dyslexia from stroke or brain injury
what happens when reading must occur via the lexical route?
- high performance in regular word reading
- LF word reading worse than HF
- lexicalisation errors with nonwords, as these activated the closest lexical word
lexicalisation errors
- non words activate closest word in the lexical route
- plage > page
Phonological dyslexia
- good regular and irregular word reading
- very poor non word reading, causing lexicalisation errors
- acquired dyslexia from stroke/brain injury
deep dyslexia…
severe from of phonological dyslexia
damage to non-lexical route and the semantic system within the lexical route
results in semantic paralexia (naming incorrect words which are related in meaning)
evidence that not all data from neurotypical participants is the same
- not all nonwords are processed the same (exception neighbours can influence RT)
- regular consistent words are processed better/faster than inconsistent, e.g., wade= consistent (jade) , wave= inconsistent (have)
why should the DR model be changed into the DR cascaded model?
-if following the GPC route, nonwords should not be affected by word neighbours but they are
- shows not all words are processed the same
DR cascaded model
- both routes race against each other and the fastest wins
- conflict resolution for both outputs occur at the phonological encoding stage
Hallmarks of developmental dyslexia
- poor reading, especially with pseudo words
-poor spelling - poor comprehension
- not related to intelligence
Prevalence of dyslexia
10% worldwide
- more apparent in languages with inconsistant orthographies
Etiology of dyslexia
- moderate heritability of 0.5
- twice as common in males
- multifactorial cause (evironmental risk+ genetics)
Intervention for dyslexia
explicit instruction in phonics , phoneme awareness and reading comprehension
What are the 4 main theories of developmental dyslexia
- Phonological deficit theory
- Double deficit theory
- Magnocellular deficit
- Cerebellar deficit
Phonological deficit theory of dyslexia
- poor phonological awareness (ability to manipulate sound)
- poor verbal short term memory
- slow lexical retrieval
Double deficit theory of dyslexia
- extension of the phonological deficit theory
- there is a phonological deficit but also naming speed deficit
- most dyslexic children have a double deficit
- deficit needs to be identified to know what type of intervention is needed
Magnocellular deficit
- dyslexics have a reduced sensitivity in the neural pathways of the visual system
- magnocells are smaller
- more visual stress
Cerebellar deficit
- right cerebellum have reduced activation with dyslexics
- temporal processing deficit so poorer naming speed
