Naslov
The effect of explicit and implicit category learning on classification accuracy and confidence

Autori
Žauhar, Valnea ; Bajšanski, Igor ; Domijan, Dražen

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

Izvornik
DUCOG Metacognition and reasoning: Program and abstracts / Csaba Pleh, Nevena Padovan, Dora Kampis, Barbabas Szaszi, Aba Szollosi ; bence Palfi - Dubrovnik, 2015, 58-59

Skup
VII. Dubrovnik conference on cognitive science: Metacognition and reasoning

Mjesto i datum
Dubrovnik, Hrvatska, 21.05.2015. - 23.05.2015

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
sustavi kategorijalnog učenja ; točnost klasifikacije ; procjene sigurnosti
(category learning systems ; classification accuracy ; confidence judgments)

Sažetak
The aim of study was to investigate how activation of explicit and implicit categorization systems during learning affects classification accuracy and confidence judgments in the later test phase. One group of participants was informed about the correct classification rule at the beginning of the learning phase (rule condition). Another group was never informed about the rule (no-rule condition). In the test phase, old and new items were presented. New items differed from the old items with respect to transfer type (good or bad) and item consistency (consistent or inconsistent). Good transfer items belonged to the same category as their old twin item, while bad transfer items belonged to the opposite category. Consistent items were highly similar to their old twins, while inconsistent items were less similar. In the rule condition, the results showed that neither transfer type nor item consistency affected classification accuracy or confidence judgments. In the no-rule condition, the results showed that transfer type affected accuracy, but not confidence. On the other hand, item consistency affected confidence but not accuracy. The results suggest that explicit category learning system is activated in the rule condition, while implicit category learning system is activated in the non-rule condition.

Izvorni jezik
Engleski

Znanstvena područja
Psihologija, Kognitivna znanost (prirodne, tehničke, biomedicina i zdravstvo, društvene i humanističke znanosti)

POVEZANOST RADA