Building a Phoneme Inventory Through Blocked Ears: The Effects of Prior Otitis Media with Effusion on Children’s Phoneme Discrimination
Fri, April 9, 11:45am to 12:45pm EDT (11:45am to 12:45pm EDT), VirtualAbstract
Most children learn spoken language through auditory input. However, what happens if access to the auditory signal is inconsistent? This is the reality for children who periodically experience mild-to-moderate hearing loss due to Otitis Media with Effusion (OME). OME is an inflammation and fluid buildup (effusion) in the middle ear without bacterial or viral infection (Cai & McPherson, 2017). OME is a highly common pediatric illness affecting around 80% of children before the age of four (Zielhuis et al., 1990; Klein, 1980). Thus, prevalence of OME is highest during the years where children attune to their native language (Kuhl, 2004) and, subsequently, construct a phoneme inventory. Despite this fact, little is known about long-term consequences of OME on phonological development. Therefore, the goal of this study is to investigate phoneme categorization by toddlers with an OME history. We hypothesize that children who experienced temporary hearing loss secondary to OME exhibit difficulties in native phoneme perception in later stages of development. We specifically believe that insufficient information relevant to the perception and categorization of speech sounds leads to the formation of “super-categories” - two phonemes that are merged into one because they are only distinguishable at frequencies that children with OME cannot perceive.
We will test three-to-four-year-old German toddlers with (N ≥ 40, control group) and without (N ≥ 40, clinical group) an OME history. The clinical group will consist of children with persistent OME fulfilling the criteria for tympanostomy tube insertion to drain the fluid from the middle ear cavity. Children will participate in the study after the surgery. Their ability to categorize different types of native language consonant contrasts (i.e., /t/ vs. /k/, and /d/ vs. /g/) will be assessed in a child-friendly XAB task (Brasiliero, 2009; Giezen et al., 2010). On each trial, children will watch and hear three cartoon characters pronouncing three nonwords: the stimulus X followed by the stimuli A and B, one of which will match X. Children will have to decide whether X was more like A or B and point at the corresponding character.
IRB approval has been granted. Children will be recruited from a local audiology practice, where they will also be tested. COVID-19 is currently limiting the number of potential participants, as due to social distancing fewer children get OME. However, we are ready to commence data collection and are optimistic that we can finalize data collection in time for the meeting.
We will analyze children’s categorization accuracy as the dependent measure in a mixed-effects logistic regression with group and contrast type as fixed factors and subject and item as random factors. We expect to find a higher accuracy in the control group compared to the clinical group. Our results will shed light on how OME affects phonological development. As the impact of OME on language acquisition is still debated (Casby, 2001; Roberts et al., 2004), these findings can clarify whether children with OME face developmental challenges beyond their medical condition.