Rachel Rosalsky

Rachel Rosalsky
Mentor: Dr. Scott Griffiths
College of Public Health and Health Professions
"I got involved in research because I am naturally curious and wanted to dive deeper into the mechanisms of the auditory system and speech perception."


Communication Sciences and Disorders



Research Interests

  • Speech Perception
  • Phonetics
  • Neuroscience

Academic Awards

  • UF University Scholars Program
  • Rita O'Connell Scholarship for People with Disabilities


  • National Student Speech, Language, Hearing Association


  • Shands Rehab at Magnolia Parke

Hobbies and Interests

  • Drawing
  • Painting
  • Learning

Research Description

Speech Perception Errors in Individuals with Varying Auditory Processing Skills
Patterns of perceptual errors in identifying speech stimuli in the presence of noise have been researched in normal individuals. Early work by Miller and Nicely (1954) presented normal individuals 16 consonants before the vowel /a/ with varying Speech-in-Noise ratios at a bandwidth of 250Hz to 6500Hz. It was found via Confusion Matrices that even at a negative SN ratio voice and nasality of speech sounds were clearly distinguishable while duration, affrication, and place were difficult for participants to identify . However, even though poor discrimination of speech in background noise is a key characteristic of Auditory Processing Disorder (APD), there is little research in the aspects of speech that individuals with an APD find difficult to identify in noise. Identification of particularly salient speech cues for individuals with an Auditory Processing Disorder may eventually be used to create a phoneme-specific APD test that is more dependable in identifying the disorder and more useful in guiding intervention. The proposed study will investigate the relationship between auditory processing skills and perceptual speech sound identification errors. In the proposed study, 32 participants will first undergo a hearing assessment consisting of pure tone audiometry, tympanometry, acoustic reflex thresholds and acoustic reflex decay. The individuals with normal hearing sensitivity will then undergo a battery of auditory processing tests, four subtests: Filtered Words, Auditory Figure-Ground, Competing Words, and Competing Sentences. Participants will also complete the Gaps-in-Noise tst. Fourteen bisyllabic Vowel-Consonant-Vowel (VCV) pre-recorded audio files with differing English consonants will then be played individually and monaurally to the participants ears, and perceptual errors will be logged. Data from the consonant confusion matrices will be analyzed using Sequential information analysis (SINFA). Statistically significant differences in the transmission of speech cues such as voicing, nasality, affrication, and place are expected between individuals with high auditory processing skills and those with low auditory processing skills. Identification errors per speech cue of consonants will be logged as a function of auditory processing skill.