S1W8-Read

Question 1

Heterographs

Answer 1

Words that are spelled differently, mean different things but sound the same

E.g. maid/made.

Question 2

Homographs/homphones

Answer 2

Words that are spelled the same but have two meanings and sometimes two different pronunciations.

E.g. reading book/Reading station.

Question 3

Lexical decision task

Answer 3

Participants asked to decide whether a string of letters is a word or not.

Measure: reaction time.

Question 4

Naming task

Answer 4

Participants asked to say printed words out loud as quickly as possible.

Measure: reaction time,

Question 5

Semantic priming effect

Answer 5

• Using a lexical decision task
Reaction time for a word was shorter when the previous word was semantically related

E.g. doctor/nurse; vs. doctor/house.

Question 6

Dual-Route Cascaded Model (DRC) definition

Answer 6

The processes involved in reading words and nonwords differ from each other (dual route hypothesis).

The lexical and non-lexical routes work in parallel.

Computational model (7,981 words).

Cascaded so activation passes through each level before processing at first level is complete.

Simulates word recognition and reading aloud.

Question 7

Non lexical route DRC

Answer 7

Orthographic analysis (letters identified and grouped together).

Grapheme-phoneme conversion (converting groups of letters into sounds).

Used for reading regular words and regular nonwords.

Irregular words and irregular nonwords are regularised due to strict grapheme/phoneme rules.

Question 8

Lexical Route DRC

Answer 8

Orthographic analysis (letters grouped together).

Orthograpgic output lexicon (store of all familar words).

Phonological output lexicon (sound pattern created and word is read aloud).

Used for reading familiar words (regular and irregular).

Frequent errors reading unfamiliar or nonwords as they aren’t contained in the store.

Question 9

Lexicon and semantic route DRC

Answer 9

Not yet implemented.

Obtains meaning for the familar words in a semantic system.

Processing between semantic system and orthographic input lexicon is bidirectional (explains semantic priming effect).

Question 10

Acquired Surface Dyslexia

Answer 10

The lexical route is damaged.

Forces people to use non-lexical route.

Patients can read regular words

Patients can read non-words

Patients have problems with exception or irregular words.

Question 11

Acquired Phonological Dyslexia

Answer 11

Can read familiar words (regular and irregular).

Problems reading unfamiliar words and non-words.

This means patients have specific problems with grapheme-phoneme conversion (damage to non-lexical route).

Question 12

Acquired Deep Dyslexia

Answer 12

Problem reading unfamiliar words and non-words.

Also produce semantic errors.

Damage to lexical and semantic route.

Question 13

Neuroimaging DRC- Taylor, Rastle & Davis (2012)

Answer 13

Meta-analysis of 36 neuroimaging studies.

Mainly left-hemisphere activation

Orthographic analysis: occipito-temporal cortex

Lexical and/or semantic processing: anterior fusiform, middle temporal gyrus

Spelling–sound conversion: inferior parietal cortex

Phonological output resolution: inferior frontal gyrus

Question 14

DRC performance

Answer 14

Read 99% of the 7,981 words.

Read 98.9% of 7,000 one-syllable non-words.

Question 15

DRC strengths

Answer 15

Accounts for surface and phonological dyslexia.

Simulates frequency effect in naming (high frequency words named faster than low-frequency).

Explains lexical-decision performance.

Question 16

DRC weaknesses

Answer 16

Only deals with mono-syllabic words.

Unable to learn.

Fails to treat words and nonwords differently.

Only applies to alphabetic languages.

Poor at naming times (39.4% nonword naming times, 4.5% word naming times).

Question 17

Distributed Connectionist Approach

Answer 17

Reading of regular, irregular and non-words through the same route/processes.

Highly interactive during learning.

ASSUMPTIONS:

Pronunciation of words affected by pronunciation of similar words (Neighbourhood Effect)

E.g. hut and but will affect the pronunciation of nut.

High-frequency words have stronger influence than low-frequency words.

Question 18

Plaut et al (1996) Connectionist Model

Answer 18

Learns to read words accurately as connections develop between grapheme-phoneme units.

The model is based on backpropagation (output is compared to the expected correct response).

Question 19

Performance of Connectionist Model (Plaut)

Answer 19

Trained on 2,998 words.

Irregular words took longer to name than regular words.

Low-frequency words took longer to name than high-frequency words.

Read over 90% of words correctly, which is similar to proficient adult readers.

Question 20

Follow up study of Connectionist Model (Plaut)

Answer 20

Lesions made to simulate surface dyslexia.

Behaved similar to patients with surface dyslexia.

Good on regular high and low-frequency words and non-words.

Poor on irregular high-frequency words

Worst on irregular low-frequency words.

Question 21

Strengths of Connectionist Model (Plaut)

Answer 21

Evidence supports that orthographic, semantic, and phonological systems are used in parallel.

Reading non-words is influenced by knowledge of reading words.

Includes learning mechanism.

Question 22

Weaknesses of Connectionist Model (Plaut)

Answer 22

Only monosyllabic words.

Not as accurate as DRC.

Did not simulate phonological dyslexia.