Word recognition

Identification of a word form under a specified input modality. Accuracy, speed and access to meaning are distinguishable measurements.

How often a word occurs in a defined language sample or learning history. Corpus frequency and an individual child's actual exposure need not coincide.

Identification of printed word forms. Familiar-word accuracy, unfamiliar-word decoding, fluent access and semantic interpretation should not be treated as interchangeable outcomes.

Acquisition of information about written forms and their relations. Retention of a particular spelling and generalisation to untrained combinations are different outcomes.

Variation in surrounding letters while a target mapping remains stable. A first-grade training experiment found transfer benefits from varied consonants around vowel mappings, motivating tests of how example distributions support abstraction.^[ref]

Participants learn controlled written forms, pronunciations and sometimes meanings, followed by trained-item and transfer tests. Manipulating frequency and consistency can test sensitivity to input structure without uniquely identifying a learning algorithm.^[ref]

The proposal that phonological recoding of unfamiliar written forms creates opportunities to acquire word-specific orthographic representations. Delayed discrimination from homophonic alternatives tests more than immediate pronunciation.^[ref]

Efficient and appropriately coordinated reading, assessed through specified accuracy, speed and connected-text demands. Accurate isolated-word performance need not establish fluent passage reading.