Naslov
Masked priming paradigm reveals automatic processes in visual word recognition: An ERP study

Autori
Palmović, Marijan ; Kuvač Kraljević, Jelena ; Hržica, Gordana ; Padovan, Nevena ; Jerbić, Ana Branka

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, ostalo

Izvornik
Book of abstracts 4th Croatian Congres of Neuroscience / / - Zagreb : Hrvatsko društvo za neuroznanost ; Hrvatski institut za istraživanje mozga Medicinskog fakulteta Sveučilišta u Zagrebu, 2013, 66-66

Skup
4th Croatian Congress of Neuroscience

Mjesto i datum
Zagreb, Hrvatska, 20.09.2013. - 21.09.2013

Vrsta sudjelovanja
Poster

Vrsta recenzije
Domaća recenzija

Ključne riječi
ERP ; masked priming ; visual word recognition

Sažetak
Masked priming paradigm has been used to study automatic processes related to visual word recognition since 1980-ties. The strength and the robustness of the priming effect in a masked priming experiment proved to be larger than in semantic or associative priming (Evett & Humphreys, 1981, Forster & Davis, 1984). During the last 15 years the masked priming paradigm has established as a reliable tool to study automatic processes, primarily orthographic, phonological and morphological. In this study we examined early processes in visual word recognition. Recent studies emphasize the importance of occipito-parietal cortex and, in particular, left fusifom gyrus (Dehaene et al., 2001) in extracting orthographic and phonological information unconsciously, during the exposure to the stimulus that precedes the target word by 50- 80 ms and is masked with the "########" symbols before and after its presentation. In a classical lexical decision task we manipulated graphical properties of the masked prime: it consisted of the identical word/pseudoword in the base condition. In two other conditions the masked prime had all the junctions and edges (L and T shaped parts of letters) either deleted or preserved (with the straight lines deleted). In these conditions exactly 50% of pixels of each letter were deleted. All pseudowords were generated by the OLD20 algorithm with the index of distance between words and pseudowords ≥2 (either two letters differ, or two letter positions or a combination of these) with the letter and bigram frequencies kept equal between words and pseudowords. The results clearly dissociate orthographic and lexical retrieval processes both in terms of different latencies and different scalp distribution. While a common N400 effect has been obtained in all pseudoword conditions (with three graphically different primes), the difference in primes shows an early effect in N1P1 (?) complex on the parietal and occipital electrode sites. These results are consistent with the current models of visual word recognition and may be further used in e.g. studying dyslexia.

Izvorni jezik
Engleski

Znanstvena područja
Temeljne medicinske znanosti, Psihologija, Filologija

POVEZANOST RADA