The Italian Crowdsourcing Project: Visual word recognition times for 130,495 Italian words.

Area of Science:

• Cognitive Science
• Psycholinguistics
• Computational Linguistics

Background:

• Italian language research lacks extensive language processing datasets.
• Existing resources are insufficient for large-scale psycholinguistic studies.

Purpose of the Study:

• To introduce the Italian Crowdsourcing Project (ICP) dataset, the largest of its kind for Italian word recognition.
• To validate the ICP dataset using established psycholinguistic measures.
• To explore unique Italian linguistic phenomena using the ICP data.

Main Methods:

• Collected word recognition times and accuracy from over 156,000 native Italian speakers via an online task.
• Validated the dataset by correlating reaction times with lab-based lexical decision latencies.
• Replicated known psycholinguistic effects (e.g., word frequency, length) and introduced word prevalence for Italian.

Main Results:

• The ICP dataset contains responses for 130,465 unique Italian words.
• ICP reaction times showed a strong correlation (r=0.78) with traditional lexical decision data.
• Key psycholinguistic variables and word prevalence effects were successfully replicated.

Conclusions:

• The ICP dataset is a validated, large-scale resource for Italian language processing research.
• The dataset enables novel investigations into Italian-specific linguistic features, such as inflectional entropy and clitic effects.
• ICP data will significantly advance the study of the Italian lexicon and psycholinguistics.