M4, T1, word recognition model and tests Flashcards
Word recognition model
*look up image
Word recog model - visual orthographic analysis
- Each letter = one unit
A=a - Font or case irrelevant
- Identifies letters
- Codes letter position
- May perceptually groups letter together
-> CHURCH = CH/UR/CH - Used for words and nonwords
Word recog model - orthographic input lexicon
- Visual word recognition units
- Store of familiar written words
- Lexicon of printed words
- Used to differentiate between familiar and unfamiliar words
- Word activation occurs over time
Word recog model - Grapheme to phoneme conversion
Letter to sound rules
-> Grapheme to phoneme correspondence rules
Read words and nonwords by translating graphemes into phonemes
letters -> sounds
Cannot use successfully for irregular words e.g. have, done
Word recog model - semantic system
- Semantic memory
- Meaning of spoken word, written word and objects known to person
- How concepts are represented in this system?
- One idea each unit semantic system = feature
- Comprehension of a word or object involves activation each unit that corresponds to the of feature(s) of a word/object
- Activation occurs over time
Word recog model - orthographic output lexicon
- Mental dictionary of all the words you can spell or write
- Lexicon for spelling/writing of words
- Lexical activation rises over time
- Intended spelling goes to Orthographic Output Buffer
Orthographic output buffer
- Holding buffer for temporary storage of material from Orthographic Output Lexicon
- Holds sequences of letters until word production is complete
- Needed as writing or spelling of a word may require repeated search Orthographic Output Lexicon for each letter -> slow process
- Word length effects operate here
Word recog model - phonological output lexicon
- Lexical entries of person’s own vocabulary
- Store of spoken form of familiar words
- Activation rises over time
- Activations starts from time person intends to say word
- Higher frequency words “recognised” earlier than lower frequency words
Word recog model - phonological output buffer
- Temporary storage for phonemes to be uttered
- Easier than a phoneme-by-phoneme production of speech
- Slips of the tongue occur here
- Word length effects operate here
Word recog model - phoneme to grapheme conversion
- Sound to letter rules
- Grapheme: Letter or sequence of letters
- Phoneme: sounds of letter(s)
- Use PGC rules translate spoken word or nonword to spelling
-> Yacht – yot
-> Phone – foan or fone - Spelling of unfamiliar words or nonwords
Word recog model - visual orthographic analysis - tests
Test: Cross case matching
- Mark the capital letter that matches the small letter
-> f
-> T F
Test: Word or Nonword matching
- Do these two words (nonwords) have the same letters in them?
-> Tower-TOWER
-> lmeon- DMEON
Word recog model - orthographic input - tests
Test: Written lexical decision
- Is this a real word?
-> KED
-> DOLL
Test: Read aloud exception words
- Read these words to me
-> YACHT
-> PINT
Word recog model - orthographic input lexicon - tests
Test: Understand written words (define, match to picture)
-> Point to the picture of a GOAT (among other images)
Word recog model - grapheme-phoneme conversion - tests
Test: Read aloud nonwords
- Read aloud these nonsense words
-> KED
-> NAR
Test: Decide if a printed nonword sounds like a real word
- Point to the made-up word below that sounds like the word FOX
-> Phocks -> Phoalks
Word recog model - semantic system - tests 1 and 2
Test : Match pictures and written words
- Which picture matches the word SPIDER? (choose from four images)
Test : Name to written definition
- Write down the name of the animal known as the king of the jungle
- Tell me the name of the animal known as the king of the jungle
