LexOPS: An R package and user interface for the controlled generation of word stimuli

Abstract

LexOPS is an R package and user interface designed to facilitate the generation of word stimuli for use in research. Notably, the tool permits the generation of suitably controlled word lists for any user-specified factorial design and can be adapted for use with any language. It features an intuitive graphical user interface, including the visualization of both the distributions within and relationships among variables of interest. An inbuilt database of English words is also provided, including a range of lexical variables commonly used in psycholinguistic research. This article introduces LexOPS, outlining the features of the package and detailing the sources of the inbuilt dataset. We also report a validation analysis, showing that, in comparison to stimuli of existing studies, stimuli optimized with LexOPS generally demonstrate greater constraint and consistency in variable manipulation and control. Current instructions for installing and using LexOPS are available at https://JackEdTaylor.github.io/LexOPSdocs/.

SCOPE: The South Carolina psycholinguistic metabase

“…Instead of relying on a small sample of carefully chosen words that may be idiosyncratic in some fashion, these studies enable development and testing of potentially more robust word processing models on a larger scale. At the same time, the number of studies that measure word properties by collecting ratings on psycholinguistic variables for thousands of words has also increased rapidly (Taylor et al, 2020), providing rich and robust measures of many lexical properties (e.g., Brysbaert et al, 2014;Lynott et al, 2020;Pexman et al, 2019;Scott et al, 2019).…”

Learning about the meanings of ambiguous words: evidence from a word-meaning priming paradigm with short narratives

“…This choice was made to ensure that our results would not be affected by the number of potential meanings of our ambiguous primes ( Yip, 2015 ). Measures of word length, frequency, familiarity and age of acquisition in Table 2 came from the LexOps shiny app ( Taylor, Beith & Sereno, 2020 ). In lieu of more formal measures of the primes’ predictability or surprisal within the narrative contexts, we also report a measure of semantic similarity between the primes and the remainder of their associated narrative, based on latent semantic analysis (LSA) using http://lsa.colorado.edu/ based on Dennis (2007) .…”

Is time an embodied property of concepts?

“…The stimuli were divided into two lists, each with 325 words. The two lists were balanced for word length, word frequency, and concreteness using the R package LexOPS (Taylor et al, 2020). The lists were administered to 240 undergraduate students; each participant saw just one list, and rated the words in that list on one dimension (time, space, or confusability).…”