1 of 17

UBC VANLAB

Investigating the

Relationship Between

Musical Sophistication

& Perceived Rhythmic

Complexity

Spencer Dai and Katherine Hsu

Principal Investigator: Dr. Leigh VanHandel

PAGE 01

2 of 17

PAGE 02

3 of 17

Overview

01 Introduction - What are perceived complexity & musical

02 Objectives - What are we interested in?

03 Methodology - Stimuli, surveys, and statistics

04 Results - Stay tuned...

05 Discussion - Major findings & future directions

PAGE 03

sophistication?

4 of 17

Perceived (rhythmic) complexity

What makes a rhythmic pattern in music sound complex?

Our perception can vary with:

Tempo (speed)
Density (number of event onsets in a rhythmic pattern)
Variability (difference in duration of successive events)
Syncopation (offbeat accents)

PAGE 04

INTRODUCTION

5 of 17

INTRODUCTION

UBC VANLAB

Faster

Slower

PAGE 05

Tempo

6 of 17

INTRODUCTION

UBC VANLAB

PAGE 06

7 of 17

Concept used to describe an individual’s experience and ability to engage in music

Measured by scores on a standardized, self-report scale (GMSI)

Musical Sophistication

PAGE 07

INTRODUCTION

8 of 17

PAGE 08

INTRODUCTION

A multifaceted concept used to describe an individual’s ability to engage with music (Müllensiefen et al., 2014)

Sub-scores:

Active Engagement - how much time and money resources spent on music
Perceptual Abilities - accuracy of musical listening skills
Musical Training - amount of formal musical training received
Singing Ability - accuracy of one’s own singing
Emotional Engagement - ability to talk about emotions that music expresses

Goldsmiths Musical Sophistication Index (GMSI)

[From Dr. Van: move this slide earlier]

The GMSI aims to give a holistic overview of musicianship that accounts for 5 aspects that contribute to an individual's ability to engage with music.

the general GMSI score is calculated based on the scorings of these five subscores.

active engagement covers musical engagement behaviours in daily life, using questions such as “I keep track of new music that I come across”, “I often read or search the internet for things related to music”.

Perceptual abilities is a self-assessment of musical abilities, mostly regarding listening skills. Participants were asked to rate how much they agree with statements such as "I can tell when others sing out of tune" and "I usually know when I’m hearing a song for the first time"

Musical training measures the extend of one's musical training and practice by having participants report information on their musical training experiences, including years engaged in musical training and practice time per day

Singing abilities reflects different skills and activities related to singing. for example, whether one is able to sing in harmony when someone is singing a farmiliar tune.

Emotional engagement covers emotional reactivity to music. Statements including " “I am able to talk about the emotions that a piece of music evokes in me” and "Music can evoke my memories of past people and places"

9 of 17

OBJECTIVES

PAGE 09

Previously...

VanHandel (2022, 2023, 2024) found a negative correlation between the overall GMSI score and average complexity rating

Those with higher self-reported levels of musical sophistication tend to rate rhythmic stimuli as less complex

10 of 17

PAGE 10

OBJECTIVES

Goals of the study

Replicate - Confirm the relationship between overall musical sophistication (general score) and perceived complexity ratings

Explore - Look at how different aspects of musical sophistication (sub-scores) affect how we perceive the various characteristics (tempo, density, etc.) that contribute to rhythmic complexity

11 of 17

PAGE 11

METHODOLOGY

Tempo

Milliseconds (ms) Inter-Onset Interval (IOI): amount of time between beats
Smaller number = faster

Perceived complexity

6 point scale
1 = very simple, 6 = very complex

Level of musical sophistication

Goldsmiths Musical Sophistication Index (GMSI)

Measurements

12 of 17

PAGE 12

METHODOLOGY

Experimental design

Participants

456 usable responses

Procedure

Participants randomly assigned to fast/slow condition
Listened to 150 audio stimuli and rated the complexity
45 GMSI questions scattered throughout survey

We recruited through the Human Subjects Pool and ended up with 456 sets of usable data.

Participants were randomly assigned to fast or slow condition. Both groups were asked to complete a survey where they rate the complexity level of stimuli upon hearing them.

[The fast group listened to stimuli at 333ms or 180 bpm and 433 ms or 150 bpm, and the slow group lstened to the same stimuli at 800ms or 75 bpm and 933ms or 64 bpm. Both group listened to the stimuli at 533ms and 666ms, which are 113 bpm and 90 bpm respectively

do we need this section? also someone might ask how the tempi determined

we do not]

Each participant listened to and rated 150 stimuli, presented to them randomly at differet tempi. To keep their attention and avoid fatigue, 45 GMSI questions are scattered thoughout the survey.

13 of 17

RESULTS

UBC VANLAB

PAGE 13

Correlation coefficient: -0.224
p < 0.001 (2-tailed)

Slow tempo only (N = 218)

Correlation coefficient: -0.162
p < 0.05 (2-tailed)

Fast tempo only (N = 238)

General Musical Sophistication Score vs Average Complexity Rating

Combined (N = 456)

Correlation coefficient: -0.192
p < 0.001 (2-tailed)

14 of 17

RESULTS

UBC VANLAB

PAGE 14

GMSI Sub-scores vs Average Complexity Rating

Combined

Musical Training

Correlation: -0.250
p < 0.001 (2-tailed)

Perceptual Abilities

Correlation: -0.184
p < 0.001 (2-tailed)

Singing Ability

Correlation: -0.125
p < 0.01 (2-tailed)

Slow group

Musical Training

Correlation: -0.230
p < 0.001 (2-tailed)

Perceptual Abilities

Correlation: -0.206
p < 0.01 (2-tailed)

Singing Ability

Correlation: -0.198
p < 0.01 (2-tailed)

Fast group

Musical Training

Correlation: -0.263
p < 0.001 (2-tailed)

Perceptual Abilities

Correlation: -0.158
p < 0.05 (2-tailed)

Singing Ability

Correlation: -0.068
p > 0.05 (2-tailed)

15 of 17

Successful replication

Individuals with higher overall GMSI scores tend to rate rhythmic stimuli as less complex

Tempo remains relevant

Overall correlation was stronger and more significant in slow condition compared to fast

Musicianship-based sub-scores driving overall correlation

Musical Training had consistently the strongest and most significant correlation across both tempo conditions
Perceptual Abilities showed the second strongest correlation with more significance in the slow condition than fast

Other sub-scores?

Singing Ability showed a significant correlation in slow but not fast condition
Active Engagement and Emotional Engagement did not show significant results

PAGE 15

DISCUSSION

Summary

Overall, this study successfully replicated thhe results from earlier studies that individuals with higher overall GMSI scores tend to rate the stimuli as less complex.

Moreover, we confirmed the negative correlation between GMSI score and tempo, and that the correlation is stronger in the slow condition.

In terms of the subscores, musical training displayed the strongest and the most significant correlation across both conditions. While perceptual abilities was the second strongest correlation, the impact was much more significant in the slow condition [explain why? Could be due to our capacity to process auditory information naturally being limited at higher rates of input?] .

Similarly, singing ability also showed a significant correlation in slow condition, but it had little impact on the fast condition.

16 of 17

Limitations

Lack generalizability

Rhythmic stimuli are based on western conception of metrical structure, and are limited to one meter
Musical sophistication questions are also developed based on engagement patterns with music typical of Western culture

Future Directions

How individual GMSI questions predict complexity ratings
How tempo affects individual aspects of GMSI

Do they show similar relationships as tempo & general score?

PAGE 16

DISCUSSION

17 of 17

UBC VANLAB

Thank You!

VanLab Research Team:

Dr. Leigh VanHandel

Betty Bao

Parker Keeley

CC Liang

Jasmine Ren

Meera Thomas

Konrad van Heukelom

PAGE 17

ACKNOWLEDGEMENTS

References

Müllensiefen, D., Gingras, B., Musil, J., & Stewart, L. (2014). The musicality of non-musicians: An index for assessing musical sophistication in the general population. PloS One, 9, e89642. https://doi.org/10.1371/journal.pone.0089642.

VanHandel, L. (2024, November 7–10). Not just syncopation: Rhythmic complexity is… complex [Conference presentation]. Society for Music Theory 2024, Jacksonville, FL, United States.