Methodologies

By tracking where listeners look as the speech signal unfolds over time, we can learn a lot about how listeners process the speech signal and access lexical representations. This paradigm works well with all ages, from young infants to adults.

This classic infant testing paradigm allows us to test how language experience influences infants’ ability to tell apart various auditory stimuli (e.g., speech sounds, voices, words, or languages).

This paradigm works well with 6- to 12-month-old infants, who can sit up and control their head movements. We have used this paradigm to test infants’ recognition of words in speech, as well as toddlers’ development of grammatical sensitivities.

The CLASS Lab is most well-known for their work on speech perception, but we also run production studies. Some paradigms we have used include altered auditory feedback, phonetic convergence, and elicited production.

Corpus studies allow researchers to better understand children’s language input. We often use corpus studies to complement our experimental work. We have recently added LENA to our toolbox and look forward to using it more in the future.

Some questions are best addressed by giving children explicit choices. For example, infants can be presented with puppets who speak in different ways. Infants’ tendency to subsequently grab certain puppets can tell us a lot about their perception of talking styles.

Most of our work focuses on speech and language development in infants and toddlers, but we also work with older children. Once children can operate a mouse or use a touch screen, we can test them using simple video games we design in the lab.

We often collect standardized measures to complement our experimental work (e.g., the MacArthur-Bates Communicative Development Inventory (MCDI), the Structured Photographic Expressive Language Test (SPELT), and the Comprehensive Test of Phonological Processing (CTOPP)).

The artificial languages we use typically consist of a small collection of nonsense words with carefully controlled phonological characteristics. By testing listeners’ ability to learn patterns embedded in these toy languages, we can test hypotheses about how real language acquisition works. 

Although we are primarily a developmental lab, we also work with adults. In addition to eye tracking (mentioned above), some other common paradigms we use with adults include talker recognition line-ups and speech transcription in noise.

During the pandemic, we developed a host of online studies powered by platforms such as Gorilla and Qualtrics. These studies have been so successful that we have decided to maintain this line of work indefinitely. For more information about our online studies, click here.