2. Word Recognition Flashcards
What are the main questions concerning word recognition and what studies support them?
- Eye movement
(Reichle et al., 2003) - Does knowledge about words help letter recognition?
(Reicher, 1969), (McClelland and Rummelhart, 1981) - What factors affect word recognition?
(Balota et al., 2014), (Brysbaert and Cortese, 2011) - Serial or parallel?
(Forster, 1976), (McClelland and Rummelhart, 1981) - How do we encode info about letter position?
(Grainger and Whitney, 2004), (Davis, 2010) - Does sound play a role in recognition?
(Van Orden, 1987), (Lesch and Pollesk, 1998) - How soon do we start to activate meaning?
(Rodd, 2007)
How do we move our eyes during reading?
(Reichle et al., 2003)
• saccades of 8 letter spaces
• considerable variation in timing and 10-15% movements are regressions
• Long words nearly always fixated, short 25% of time
Does our knowledge about words help letter recognition?
Word superiority effect:
(Reicher, 1969)
• letters easier to identify in a real word- tachistoscopic presentation
(McClelland and Rummelhart, 1981)
• Bias to identify partially obscured letters such that string forms word
What factors affect word recognition?
(Balota et al., 2014)
• MegaStudy approach- used multiple regression to identify effects of different variables on speed of response
• Frequency, length, sound-to-spelling consistency, imageability
• Age of acquisition remains controversial (Brysbaert and Cortese, 2011)
Recent experience:
Pruning and interference- helpful or unhelpful context
Is word recognition serial or parallel?
(Forster, 1976)
• Serial Search Model- two stage- forms then content
(McClelland and Rumelhart, 1981)
• Interactive Activation and Competition Model (IAC)
• Interconnected networks of neurones- connectionist view
General consensus of parallel as serial is implausible neurologically
How do we code information about letter position?
(McClelland and Rumelhart, 1981)
IAC- slot based coding
(Grainger and Whitney, 2004)
• Subsets of letters give priming effect
• Changing order destroys priming effect
• 92% overlap in garden and gadren priming, 25% in galten
-> encoding based on ordered letter pairs (open bigrams)- relative positions of letters
Led to development of:
Spatial Coding Model of visual word identification- letter position coded dynamically
(Davis, 2010)
How soon do we start to activate the meaning of a word?
(Rodd, 2007)
• Experimental evidence for parallel activation of meanings
• Leotard effect- semantic categorisation task
• Slower responses to experimental items
• However, May be due to priming effects/expectation
Does the sound of a word play a role in recognition?
(Van Orden, 1987)
• Category decision task- high error rates for homophones with different spellings pair pear
• phonological mediation May play higher role for low-frequency words and poor readers