Consistency norms for 37,677 english words

Abstract

Consistency reflects the mapping between spelling and sound. That is, a word is feedforward consistent if its pronunciation matches that of similarly spelled words, and feedback consistent if its spelling matches that of similar pronounced words. For a quasi-regular language such as English, the study of consistency effects on lexical processing has been limited by the lack of readily accessible norms. In order to improve current methodological resources, feedforward (spelling-to-sound) and feedback (sound-to-spelling) consistency measures for 37,677 English words were computed. The consistency measures developed here are operationalized at the composite level for multisyllabic words, and at different sub-syllabic segments (onset, nucleus, coda, oncleus, and rime) for both monosyllabic and multisyllabic words. These measures constitute the largest database of English consistency norms to be developed, and will be a valuable resource for researchers to explore the effects of consistency on lexical processes, such as word recognition and spelling. The norms are available as supplementary material with this paper. Keywords Consistency . Feedforward . Feedback . Word norms . Word recognition The relationship between spelling and sound is a key factor in determining skilled reading, and has been the focus of a number of influential models of word recognition (e.g., Coltheart,

Distinct neural sources underlying visual word form processing as revealed by steady state visual evoked potentials (SSVEP)

“…All words were fully feedforward consistent based on rime according to the database provided by Ziegler et al 98 . When averaging across consistency values for each word’s onset, nucleus, and coda in the database provided by Chee et al 99 , words had an average token feedforward consistency of 0.79. All in all, there were 30 exemplars of each type of stimulus, for 120 exemplars total.…”

Mediated phonological–semantic priming in spoken word production: Evidence for cascaded processing from picture–word interference

“…Distractors across categorically related and mediated phonological-semantic-related conditions were matched closely on a range of lexical properties using the English Lexicon Project normative database (Balota et al, 2007), including log frequency (SUBTLEX; Warriner et al, 2013), the number of letters, phonemes, and syllables, concreteness, age of acquisition (AoA; Kuperman et al, 2012), semantic diversity (Hoffman et al, 2013), orthographic (Yarkoni et al, 2008), and phonological (Suárez et al, 2011). Levenshtein distances (OLD20 and PLD20), spelling-to-sound (feedforward), and sound-tospelling (feedback) consistency measures for first syllable and composite (for multisyllabic words) onsets (Chee et al, 2020; see Table 1). The auditory distractors were recorded by a female native English speaker in an anechoic chamber, and edited and normalised using Audacity (Audacity Team, 2019) and the DC offset removed.…”

Quantifying the regularities between orthography and semantics and their impact on group- and individual-level behavior.

“…Needless to say, each of these two methodological decisions is justifiable. Thus, there is no a priori reason why one would use a type- or a token-based calculation for computing OSC, and it is likely that either mode of calculation provides important information on O-S associations that the other does not (see, e.g., Chee et al, 2020, for a discussion in the context of O-P regularities). Similarly, the decision to include information about the word itself in the calculation of OSC can be defended on a theoretical ground (for example, if a stem X is more frequent than a set of semantically-unrelated neighbors that include this stem—say Xy and Xz —in most cases where the orthographic form X is encountered the same meaning is present, and the original OSC definition captures that).…”