"Earworms" and Involuntary Musical Imagery

An analysis of "Repetition and recency increases involuntary musical imagery of previously unfamiliar songs" by Timothy Byron and Lucinda Fowles

Elaboration 2017 5 Pages

Musicology - Miscellaneous


Involuntary Musical Imagery

An earworm (involuntary musical imagery, or INMI) may be described as the experience of a short tune, which becomes trapped in the mind and is heard repeatedly outside one’s conscious control (Floridou, Williamson, & Müllensiefen, 2012). A study by Liikkanen showed 91.7% of the 12, 420 participants reported INMI at least once a week; 33.2% every day and; 26.1% several times a day (2008). This research essay analyses a published research paper by Timothy Byron and Lucinda Fowles titled Repetition and recency increases involuntary musical imagery of previously unfamiliar songs. In addition to this, an explanation of the experimental design of a primary experiment conducted by the 1105QCM class will be provided. The results and their importance will be discussed in detail with reference to the article by Byron and Fowles. Finally, the strengths and limitations of the experiment will be examined to draw conclusions about the accuracy and reliability of the data. The research aimed to determine the relationship between familiarity, likeability, number of earworms experienced and earworm length. It was found that familiarity had no impact on the number of earworms or earworm length for a particular song, which rejected the initial hypothesis. It was discovered that the likeability of a particular song increased the number of times it occurred as an earworm and increased earworm length, which supported the hypothesis.

Byron and Fowles’ experiment investigated “the effect of repetition, recency, and levels of processing on the induction of involuntary musical imagery in previously unfamiliar songs” (Byron & Fowles, 2013). Participants were separated into four groups and were played an unfamiliar song, which ranked in the Top 10 UK Charts but not in the Top 50 Australian Charts. The low-familiarity group heard the selected song twice while the high-familiarity group had six hearings. The “Semantic Association” group were required to answer general questions about the songs while the “Autobiographical Association” group answered questions about their personal lives. Afterwards, participants completed a form on demographic details and rated the song’s relatability, likeability, catchiness and familiarity on a 7-point Likert scale. For the following 72 hours, participants were sent an SMS message asking them to complete an experiment-sampling form (ESP) in intervals of 2 hours. The study involved 53 Australian, first-year psychology students, though 17 were excluded for failure to comply with the conditions of the experiment. From the remaining 36 participants, there were 24 females and 12 males with a range of musical training and experience. The age range was 17-27 with a mean of 19.64 years. The results showed participants from the high-familiarity group were more likely to experience INMI. Also, when presented with an unfamiliar stimulus, INMI was more likely to occur early on, due to the recency effect. One strength of the study was the control over variables, including volume, gaps between hearings, session duration, and method of instruction. A limitation was the lack of consideration for the likeability of the songs in the statistical analysis, excluding personal connection as the foundation for likeability (Byron & Fowles, 2013). This experiment relates to the 1105QCM experiment as they both explore the impact of likeability and familiarity on INMI.

The class experiment investigated the impact of likeability and familiarity on the prevalence of INMI in students enrolled in 1105QCM. Data was analysed from 72 participating students, which included 48 females and 24 males with a mean age of 22.53. Participants were required to complete an earworm monitoring journal over two consecutive days, documenting all INMI experiences. On the second day, students had to listen to their allocated song at 9:00am six times in a row with no breaks. At the conclusion of the experiment, students were required to complete a demographic questionnaire, which included responses to their familiarity and liking of the song. These conditions ensured results were accurate, as all students had to undergo the same procedure using the same method of documentation within a set time frame. The stimuli utilised in the experiment were four current pop songs so it was unlikely students would have old memories of these songs, and many may would have never heard them before. This was important because familiarity was one of the variables being tested in the experiment. The familiarity of the song was measured through the demographic questionnaire, which asked students how many times they had heard the song before the experiment commenced. The questionnaire asked the participant to rate their liking of the song on a scale of 1-7 after the first playing and final playing, as well as their liking of the genre and how catchy they thought it was to determine likeability (T. Byron, 1011QCM lecture notes, 21 July 2017). It was hypothesised that both familiarity and likeability would increase the likelihood of INMI occurring, based on evidence from previous research (Goodrich, 2017; Jakubowski, 2016; Wierwille, 2011).

The hypothesis suggested that familiarity increases the likelihood of INMI. According to the results, there was a “38% likelihood that the difference in the amount of earworms between people who had heard the song a lot and people who had the song less had occurred as a result of chance” (T. Byron, 1011QCM lecture notes, 28 August 2017). As this figure is greater than 5%, it was not a significant result, meaning familiarity did not impact on the number of earworms likely to occur. Further evidence of the insignificance of familiarity from the experiment is derived from the statistic stating there is a “92.8% chance that the difference in length of earworms between people who had heard the song a lot and people who had not was not a real difference” (T. Byron, 1011QCM lecture notes, 28 August 2017). This figure is less than 95%, meaning familiarity did not have a significant impact on the length of the experiment song earworm. The hypothesis pertaining to likeability, however, was supported by the results, which proved likeability to be significant in the development of INMI and increased earworm length. Statistically speaking, there was a 0.1% chance that the increased occurrence of the INMI for people who documented an above average rating for the likeability of the song was simply due to chance. There was also a “0.1% chance that the difference in length of earworms between people who gave the song higher liking ratings and people who gave the song lower liking ratings was simply due to chance” (T. Byron, 1011QCM lecture notes, 28 August 2017). These statistics are below 5%, which indicates that the likeability of a song significantly impacts on the likelihood of INMI occurring and increases the length of earworms. The previous experiment conducted by Byron and Fowles revealed a significant difference between the low-familiarity and high-familiarity groups where a greater number of earworms occurred in those who had heard the song six times, as opposed to four. This outcome contrasts with the results of the primary experiment but this could be due to the division of the participants in the secondary experiment into groups of familiarity. Further research is required to confirm the link between familiarity and the development of INMI.

There were a few strengths to the experiment including the analysis of the way the variables impacted on INMI and the methodology. The experiment determined the relationship between familiarity, likeability, number of earworms, and earworm length. By considering multiple variables, more knowledge could be gained from the statistics generated. Another strength was the methodology, which effectively controlled the variables not being tested. Some of these included the time of day to listen to the selected song, the song format (audio only), and the due date to submit the journal. Controlling these aspects of the experiment meant all participants followed the same guidelines so there could not be any discrepancies in the results. Some limitations associated with the experiment were the irregularities of people documenting their earworms and the inability of the experiment to identify why some songs appear more likeable than others. The forum did not specify a time interval for participants to follow and was a suggestion rather an instruction. This ambiguity meant some students set alarms while others did not, and the students who set alarms probably used different intervals. A consequence could be inconsistencies in the data due to students who forgot to document their earworms. The experiment often referred to the likeability of songs but didn’t explore what elements made it likeable, such as the musical elements (e.g. structure, chord progression, melodic shape) or emotive elements (e.g. relatability, expression, mood). Also, the questionnaire could have been more specific by asking participants which parts of the song liked in particular. Overall, the experiment effectively determined the relationship between familiarity and likeability on the number of earworms experienced and earworm length, achieving the aim of the study. The conclusion drawn from the results were supported by previous experiments so it is fair to say the experiment provided an accurate summation by building on existing knowledge.

In conclusion, INMI is experienced by 98% of individuals, which is why it is important to understand how the phenomenon works (McCauley, 2014). This research essay proved there is a link between likeability and the prevalence of INMI. The primary experiment confirmed that if a song is well-liked, it will have a higher number of occurrences as an earworm and earworm length will increase. This aspect of the hypothesis was validated, but further research could explain why some songs are more popular than others and therefore more likely to become an earworm. The primary experiment and secondary research by Byron and Fowles had contrasting results in determining the impact of familiarity on INMI. The primary experiment showed no significant impact while the previous research documented a substantial variation between INMI occurrences in low-familiarity and high-familiarity groups. Therefore, more research is required to determine this relationship and decide whether this aspect of the hypothesis was supported.

Reference List

Byron, T., & Fowles, F. (2013). Repetition and recency increases involuntary musical imagery of previously unfamiliar songs. SAGE Journals, 43 (3), 375-389. doi:10.1177/0305735613511506

Floridou, G., Williamson, V., & Müllensiefen, D. (2012). Contracting earworms: The roles of personality and musicality. Proceedings of the 12th International Conference on Music Perception and Cognition and the 8th Triennial Conference of the European Society for the Cognitive Sciences of Music, 516-518. Retrieved from https://www.researchgate.net/profile/Victoria_Williamson/publication/235417631_ConContract_Earworms_The_Roles_of_Personality_and_Musicality/links/0fcfd5116185c5050d000000.pdf

Goodrich, K. (2017). Why songs get stuck in your head: Earworms. Retrieved from https://www.brainscape.com/blog/2015/12/earworm-why-songs-get-stuck-in-your-head/

Jakubowski, K. (2016). Earworms: Why some songs get stuck in our heads. Retrieved from https://www.dur.ac.uk/research/news/thoughtleadership/?itemno=29480

Liikkanen, L. (2008). Music in everymind: Commonality of involuntary musical imagery. In K. Miyazaki et al. (Eds.), Proceedings of the 10th International Conference on Music Perception and Cognition (ICMPC10), 408–412. Retrieved from http://i.org.helsinki.fi/lassial/files/publications/080904-Music_in_everymind_pdf.pdf

McCauley, A. (2014). Earworms: The science of and the statistics behind getting a song stuck in your head. Retrieved from http://www.ecojazz.com/2014/11/14/title/

Wierwille, S. (2017). Earworms: Traits of catchy music. Retrieved from https://steffyroo.wordpress.com/2011/01/17/earworms-traits-of-catchy-music/



ISBN (eBook)
File size
394 KB
Catalog Number
Institution / College
Griffith University – Queensland Conservatorium of Music
INMI Involuntary Musical Imagery Earworm Experiment Research Literature Review



Title: "Earworms" and Involuntary Musical Imagery