MALD in the wild
This project was completed in collaboration with Pearl Lorentzen, Matthew C. Kelley, and Benjamin V. Tucker and with the help of other past members of the Alberta Phonetics Laboratory.
Most experiments are conducted in laboratory conditions. This is because researchers want to make sure that the results they obtain are not a consequence of some external factor that was not predicted and perhaps even exerted its influence unnoticed. One downside of collecting psycholinguistic data exclusively in the laboratory is that it requires that participants come to the lab, which is time-consuming. Many participants are therefore selected by convenience and share characteristics that may distinguish them from other humans that do not participate in studies nearly as often (see, as an example, the construct of WEIRD – Western, Educated, Industrialized, Rich, and Democratic – participants).
The main branch of the Massive Auditory Lexical Decision project contains data from university students younger than 30. This project, internally called TWOSE_MALD, aimed to collect auditory lexical decision data (for information about what that is see this other project page) from a broader population while also collect data in an uncontrolled and noisy environment that much better matches the circumstances in which we listen to speech in our daily lives. The experiment took place over many weeks at the Telus World of Science in Edmonton which offers plenty of districtions: noise from other visitors (especially one’s own children), the quite loud IMAX theatre, exhibits in the hallways, and the researcher-favorite wandering Tyrannosaurus Rex mascot. The experimental session was shortened and the experiment was administered on tablets. More than a thousand participants were tested, competing for the coveted opportunity to be listed as the day’s top performers.
The most important findings are that participants performed on average slower and with more errors than the participants in the standard MALD experiment, but that the overall effects of staple psycholinguistic predictors of performance in the auditory lexical decision task (such as word frequency) remained the same. We also tested a much wider variety of participants. For example, as the figure above shows, we can track how participant performance in terms of their accuracy changes with age, ranging from 4 to 86. Participants become substantially better until they reach the age of 16-20. Interestingly, the participants do not show signs of getting worse as they get old, but that may be partly because the elderly participants that spend their day at the Telus World of Science likely have retained much of their vitality.
We published a brief report on the results. The study is also briefly summarized and the data is made available in a subsequent encyclopedia publication. The encyclopedia chapter also includes other information about MALD and included Yoichi Mukai and Scott James Perry as coauthors as well.
