Decoding French

In French, several grapheme-to-phoneme correspondences (GPCs) are inconsistent and their specific impact on decoding performance has been little studied. This is the case for the grapheme s that corresponds to /z/ between two vowels and to /s/ in other cases; for the grapheme e that mostly corresponds to /ɛ/ when followed by two consonants and to /ə/ when it is not the case; for the grapheme g that corresponds to /g/ when followed by a, o, u, r, l and to /j/ when followed by e, i or y; for the sequences of letters on, in, and an that respectively corresponds to /ɔ̃/, /ɛ̃/, /ɑ̃/ when followed by a consonant and that are separated (/o-n/, /i-n/ and /a-n/) when followed by a vowel or by the duplicated nasal consonant; and for the final consonants d, p, s, t that are mostly silent at the end of words but audible in a few cases.fr To measure the extent to which such inconsistent graphemes cause decoding challenges to expert readers, 27 adult participants were asked to read aloud 60 pseudowords containing them and 60 matched control pseudowords without them. The "participants" file reports the percentage of accuracy and the mean RTs in ms for each scenario (target and matched groups of pseudowords); the "abbreviation list" file reports the complete description of each colomn of this participants file; and the "pseudowords" file reports the percentage of accuracy and the mean RTs in ms for each (target and matched) pseudoword