Evaluation, Measurement, and Rater Quality in Music Performance
My interest in music performance assessment pertains to the development of valid and reliable measurement tools using rater-mediated assessment data. More specifically, my research utilizes the application of modern measurement theory models (e.g., Rasch Measurement Theory) to evaluate rater quality, ratings scale structure, differential facet functioning, and rater accuracy in various assessment contexts. In the field of music education, research in the areas of measure construction, rater consensus estimates, and rater consistency estimates often utilize classical test theory (CTT) models. Known for its weak theoretical assumptions, CTT most often uses raw scores in the data analysis process and therefore has demonstrated three fundamental weaknesses: (a) the sample and test dependency of estimated person parameters and item parameters; (b) lack of precision in evaluating raters’ use of rating scale categories; and (c) inability to analyze additional sources of variation outside of item difficulty and person ability. Most notably, under this paradigm, raw scores gleaned from rater-mediated assessments (such as large group music performance assessments or solo and ensemble assessments) are associated with characteristics of raters and not necessarily with the performances themselves, thereby limiting the ability to develop valid and reliable measures and to make informed inferences related to examinee ability and item difficulty that extend beyond the context of a single assessment situation.
Application of the the Rasch family of models can help improve objectivity by utilizing probabilistic distributions of responses as a logistic function of person and item parameters in order to define a latent trait. Raw scores are transformed to a log-odds scale using a logistic transformation, thereby mapping hierarchies of difficulty for each relevant item, and each examinee’s discrete item responses onto a single logit (log-odds) scale. The strict set of requirements used to demonstrate good model data fit to the family of Rasch models implies invariant measurement. These requirements can be extended to assessment situations that use rater-mediated assessment data, thereby demonstrating that a set of items being used can measure a single construct (i.e., latent trait), the local independence of items, and sample-independent estimations of person and item parameters (i.e., rater-invariant measurement).
Cognition, Action, and Perception (CAP) in Improvised, Groove-based Music Performance
My interest in auditory cognition, perception and action pertains to joint music performance, including how human cognition and interaction within an ensemble affects ensemble perception of acoustical events and actual acoustical events in music. My research agenda includes investigation into and measurement of temporal rhythmic structure (analysis and synthesis in studies of music performance). In particular, I am interested in the effects of sensorimotor synchronization, motor control, music ensemble coordination/interaction, and kinematics on musical expression.
Music performance is most often a joint action that includes the synchronization of multiple ensemble performers connected by a common goal of producing a coherent musical entity. Until recently, research in joint action has focused on interpersonal synchrony and temporal coordination within a social environment. Latest trends in the study of joint action, however, have embraced the motor cognition (i.e., embodied cognition) paradigm as a means for investigating mental processing. The focus on action underscores motor cognition inquiry and is proving to be a fruitful method for analyzing motor patterns and sensory effects related to music making in a joint context. My research agenda aims to develop a computational model that organizes both the auditory modality (i.e., expressive devices) as well as visual modality (i.e., human kinetics) in improvised jazz and other groove-based performances. More specifically, the model includes (a) the investigation of anticipation, adaptation, and attention through the scope of auditory and motor imagery analyses; (b) emergence of leader/follower relations in timing and movement and its correlation to entrainment; (c) interpersonalism verses intrapersonalism; (d) auditory scene analysis and attending in improvised musical performance; and (e) an analysis-by-synthesis approach to evaluating microstructure. Currently, I am working collaboratively on this model with an international team of cognitive scientists (International Laboratory for Brain, Music, and Sound Research) and music acoustic experts (Institute of Music Acoustics [Wiener Klangstil]) to extract empirical measures of group movement, coordination, and musical expression using strain gauge resistors, wavelet-based time series analysis, and state-of-the-art algorithms from musical information retrieval. We are working to develop a unique model under the following four premises:
1. The current body of research literature related to joint action and auditory cognition in music performance examines motor cognition in the context of tonal, pre-20th century Western European art music. In this context, analysis of ensemble expression is often at the tactus (i.e., tempo-pulse) level. Arguably, in jazz, groove-based music, and other non-Western music, the subtleties of musical expression and nuance is often found at the subtactus level and occurs on a microstructural timescale.
2. The notion of “groove,” or perceptual awareness of rhythmic parameters that often induce kinesthetic movement (i.e., head bobbing and foot tapping) is not paralleled in Western music and therefore does not fit any previously proposed models pertaining to cognition and musical structure.
3. Previously proposed models of joint action do not consider in-the-moment decision making and social interaction in an improvised context. Existing models include performance practice of prepared, pre-20th century Western European art music.
4. Considerations of the cyclic structure of groove-based music and emphasis on rhythmic organization (verses melodic organization in Western music) offer concerns that call for new paradigms and lines of inquiry.