A new approach to measuring absolute pitch on a psychometric theory of isolated pitch perception: Is it disentangling specific groups or capturing a continuous ability?

Autor(es)

Germano, Nayana Di Giuseppe ^[1] ; Cogo-Moreira, Hugo ^[2] ; Coutinho-Lourenco, Fausto ^[3] ; Bortz, Graziela ^[4]

Afiliação do(s) autor(es)

^[1] Univ Fed Santa Maria, Dept Mus, Santa Maria, RS - Brazil

^[2] Univ Hong Kong, Sch Publ Hlth, Hong Kong - Peoples R China

^[3] Univ Fed Sao Paulo, Dept Psychobiol, Sao Paulo, SP - Brazil

^[4] Univ Estadual Paulista, Dept Mus, Sao Paulo, SP - Brazil

• Tipo de documento: Artigo Científico
• Fonte: PLoS One; v. 16, n. 2 FEB 22 2021.
• Citações Web of Science: 0

Resumo

Absolute Pitch (AP) is commonly defined as a rare ability that allows an individual to identify any pitch by name. Most researchers use classificatory tests for AP which tracks the number of isolated correct answers. However, each researcher chooses their own procedure for what should be considered correct or incorrect in measuring this ability. Consequently, it is impossible to evaluate comparatively how the stimuli and criteria classify individuals in the same way. We thus adopted a psychometric perspective, approaching AP as a latent trait. Via the Latent Variable Model, we evaluated the consistency and validity for a measure to test for AP ability. A total of 783 undergraduate music students participated in the test. The test battery comprised 10 isolated pitches. All collected data were analyzed with two different rating criteria (perfect and imperfect) under three Latent Variable Model approaches: continuous (Item Response Theory with two and three parameters), categorical (Latent Class Analysis), and the Hybrid model. According to model fit information indices, the perfect approach (only exact pitch responses as correct) measurement model had a better fit under the trait (continuous) specification. This contradicts the usual assumption of a division between AP and non-AP possessors. Alternatively, the categorical solution for the two classes demonstrated the best solution for the imperfect approach (exact pitch responses and semitone deviations considered as correct). (AU)

• Processo FAPESP: 16/08377-4 - Ouvido Absoluto e Ouvido Relativo: Um Estudo Comparativo das Habilidades Cognitivas
• Beneficiário: Nayana Di Giuseppe Germano
• Modalidade de apoio: Bolsas no Brasil - Doutorado
• Processo FAPESP: 19/02133-4 - Os efeitos sobre estruturas cerebrais, habilidades sociais e cognitivas em crianças expostas ao Programa Guri Santa Marcelina na Grande São Paulo: um estudo quase-experimental
• Beneficiário: Graziela Bortz
• Modalidade de apoio: Auxílio à Pesquisa - Pesquisa em Políticas Públicas