Effects of Word Frequency and Neighborhood Density on Auditory Lexical Decision Time: An Online Replication of Goh et al (2009)

Humans seemingly effortlessly understand speech. As people listen, incoming auditory information is matched to an internal mental dictionary in order to successfully recognize words. This mental dictionary is called the lexicon. Research on the structure of the lexicon and how the lexicon is organized are core topics in the field of speech perception. Using an auditory version of the lexical decision task (ALDT, Luce & Pisoni, 1998), Goh, Suarez, Yap, and Hui Tan (2009) aimed to determine the structure of the lexicon. In the ALDT, participants classify speech utterances as words or non-words as quickly as possible. The speed at which one makes a lexical decision upon hearing aword reflects the extent to which theword is accessible in the lexicon. Goh et al. (2009) found in young adults that phonological neighborhood density (how many words sound similar to a given word) and word frequency (how often a word is used in its given language) can help predict how quickly a word is recognized. Specifically, they found that words that are more frequent in the English language were recognized faster than low frequency words. Their results also revealed an interaction where word frequency effects were larger for words from sparse rather than dense phonological neighborhoods. These results suggest that words are organized in our lexicon both according to their phonological similarity as well as how frequent they appear in a given language, and that these qualities interact. In the current study, we aim to replicate the findings of Goh et al (2009) and examine whether they would replicate on an online platform. Online data collection is increasingly used and is becoming the future of psychological studies because of its ease in accessibility and its ability to collect data from different parts of the country and world. Yet, studies involving auditory processes are not widely tested using online platforms because the experimenter cannot control the participants' auditory environment. Slote and Strand (2016) studied the ALDT online in order to see how well data collection of auditory stimuli in a word recognition task can perform online in comparison to in-lab data collection. In this study, they found a significant correlation in both accuracy and reaction time responses between in-lab and online participants. These results show that there is strong evidence that online data can be an accurate form of data collection for auditory stimuli in ALDT. Participants for the current experiment (ages 18-24) were recruited using Prolific (www.prolific.co). The participants completed an online version of ALDT, created using Gorilla (https://gorilla.sc/). The ALDT experiment used the same set of words as Goh et al (2009). The results of the experiment will be discussed in terms of the replicability of Goh et al's (2009) results and the extent to which online platforms, such as Gorilla, are beneficial for investigations of lexical structure.