Predictability and Music: Why You Never Like a Song on First Listen

Reference: Madison, G., & Schiölde, G. (2017). Repeated listening increases the liking for music regardless of its complexity: Implications for the appreciation and aesthetics of music. Frontiers in Neuroscience, 11, 147.

Guest post by Caroline Ibarra 

The release of Taylor Swift’s new album (“Midnights”) prompted the question in my mind: why do I never like music the first time I hear it? Every time a new song or album is released, even when it’s from an artist I adore, I usually feel indifferent or even dislike the song the first time I hear it. I always push through, because inevitably my first impression will be proven wrong. 

In a couple days, I’m bound to be listening to the song on repeat. 

I discussed this phenomenon with my brother who told me there was actual science behind it. Of course, I had to do my own research and find out more.

Researchers Guy Madison and Gunilla Schiöde conducted an experiment proving that repeated listens increases the liking of music. They found that liking increased with repeated listens across all levels of musical complexity, concluding that familiarity is the most important variable in determining musical liking. Repeated listening can increase the enjoyment of almost any piece of music. 

The Experiment 

They began their experiment by consulting professional musicians. The musicians would rate the complexity of the music examples that would be used in the study. The researchers chose 80 examples to use in their experiment. 

A group of fifteen people were recruited to take part in the experiment. The group had a variety of favorite genres and music listening habits. This variability in the sample allows us to apply these results to the general population. If they had chosen a group of 15 people who all loved pop, the results of the experiment would only apply to people who love pop. In addition, any consistent results found in the experiment could be attributed to this shared trait. 

Each person was given a CD containing the music examples in a randomized order. The participants were told to listen to one music sample a day, except for one day of the week on which they would rate the examples.

Participants rated the extent to which they agreed with a group of statements about the music examples. During the first rating session, participants were asked to rate how familiar the musical examples were (e.g., “I listen to similar music”). Later on, this statement would be replaced with “I have recently listened to similar music”. The other statements used to rate examples were: “I like this music example”, “This music example is odd”, and “This music example is dull”. 

The Hypotheses and Results

Madison and Schiöde’s hypotheses center around the concept of the Wundt curve. The Wundt curve (as can be seen in the figure below) is a bell-shaped curve that illustrates moderate increases in a stimulus (e.g., music) will have pleasant effects, whereas increases at higher levels will have unpleasant effects. 

The researchers had three main hypotheses, which are illustrated in the figure below. 

  1. Liking of the musical examples would exhibit a Wundt curve relationship with the number of presentations.
  2. Liking of the musical examples will exhibit a Wundt curve relationship with the complexity of the examples. 
  3. Repeated listening should cause an increase in liking of musical examples with greater complexity up to a certain number of presentations.
The hypothesized Wundt Curve that Madison and Schiöde predicted would result from the experiment.
Adapted from the original article.

The study was designed with optimal conditions for a Wundt curve to manifest itself, but it never did. Instead, the researchers found a monotonic increase in liking for both higher complexity and number of presentations. A monotonic function is one that is entirely nonincreasing or entirely nondecreasing. That is, as X increased, so did Y.

Example of a nondecreasing monotonic function.

In previous studies, liking began to decrease after multiple presentations. In the current study, such a decrease in liking was never observed, even when participants listened to the same example over 25 times (far more than in previous studies).

What can explain these discrepant results? 

  1.  A key difference in Madison and Schiöde’s study is the freedom they gave their participants. They were in control of how and when they listened to their musical examples, so perhaps it felt like they were just listening to music rather than participating in an experiment. 
  2. Previous studies had chosen music samples that were odd to a group of American participants (Jänke et al., 2015). Madison and Schiöde made sure to choose music that closely represented popular music at the time. 
  3. The same short music example was heard several times in the same session (Burke and Gridley, 1990). The present study was designed so that each example was played only once per session. 

Limitations and Directions for Future Research

With every experiment, there are limitations. Thus, every research project could be improved and should be read with some level of skepticism. Two key points may have affected this study’s results:

  1. The musical examples do not represent a random sample of all the published music within a certain time period. The music examples were specifically curated for this experiment. If the researchers had chosen their music completely randomly, they may have found different results. 
  2. The participants did not constitute a representative sample of all music listeners. Each participant had to be willing to spend 40 minutes of their day listening to music examples. This condition attracted people with an affinity for music, which may have made them more open to unfamiliar music. Someone who rarely listens to music may have rated the examples differently. 

Future researchers should repeat this experiment, but with a broader selection of music.It would be interesting to see those results compared to the results of this experiment. In addition, they should try to attract participants who rarely listen to music as well as music lovers. 

Familiarity Determines Musical Liking

The biggest conclusion from this study is that familiarity is the most important variable in determining musical liking. Under natural conditions, repeated listening can increase the liking of almost any piece of music. 

The results of this experiment may be due to the mere exposure effect. This concept in psychology claims people develop preferences for things they are more familiar with. The effect is most likely to occur when there is no pre-existing negative attitude toward the material. The participants of this experiment all enjoyed music, making them perfect candidates to be influenced by the mere exposure effect. 

Next time you find yourself initially disappointed with an album, just try giving it some more time. You may just need some more exposure before reaching your final opinion.

Caroline recently graduated from the University of Central Florida with a degree in biology. She currently works as a lab technician for the University of Florida where she assists with lab and field work for projects related to agricultural sustainability. She hopes to have a career as a science writer in the future.

Check out more of her written pieces here!

Additional References

Alexandrov O. 2007, Monotonicity example1, accessed 16 January 2023,

Burke, M.J., and Gridely, M.C. (1990). Musical preferences as a function of stimulus complexity and listener’s sophistication. Percept. Mot. Skills 71, 687-690. 

Jänke, L., et al. (2015). Time course of EEG oscillations during repeated listening of a well-known aria. Front. Hum. Neurosci. 9:402. 

Nickerson C., “What is the Mere Exposure Effect.” SimplyPsychology, Accessed 16 January 2023

“Wundt Curve.” APA Dictionary of Psychology,