In Anchorage, Alaska, in the darkest winter of winter, the sun rises shortly after 10 am, calls near the horizon for several hours and drops again down just before 16:00 – all in all, just five hours and 28 minutes of daylight. Around this winter solstice (just in time for Christmas), Spotify's listeners in the northernmost latitudes of the world tilt the intensity of their musical choices, choosing a more relaxed and relaxing music.
No, it's not because of the chilled Christmas playlists: in the southern hemisphere it's just the opposite, with a peak of intensity just after the summer solstice in December. A book in Nature Human Behavior this week attracted data from the listening of almost one million Spotify listeners from around the world describing the everyday and seasonal differences in how people listen. Researchers suggest that the results point to a universal human habit that probably sounds familiar: choosing music to match and change your mood.
Morning Music Against Lunch Music
t The way in which mass trends in musical choices can illustrate the rhythms of the musical role in the emotional life of people. So, when finally interns in Spotify, Park uses the opportunity to work with colleagues Spotify and her counselor to dig into the treasury of information that's already available to people's music choices.
recommends listeners' music, announcing that in 2016, the year used in the survey, more than 80% of the songs were personal choices to consumers. Researchers also limit their focus to songs they have played all the time, arguing that users will miss songs that they do not like or have not been in the mood for (they probably decided that Spotify's limitations for missing profile accounts are not It is a big problem to throw away their results significantly.
The researchers took data from listeners in 51 countries, making sure their samples match the demographics of each country, but otherwise they chose by chance. Spotify's music data, they track a variable called "musical intensity," ranging from a highly relaxing (acoustic, instrumental, ambient and flat or low tempo) to highly energetic (rhythm, dancing, loud and bouncing)  These intensity preferences track the daily rhythms more or less exactly as you would expect: songs with a lower intensity in the morning that rise to normal working hours, then remain stable before retiring in the evening and the weekends seem little These results coincided with a previous study, tracking the emotions in the Twitter user's speech, but it differed in one question: the language showed an afternoon decline, but there was no such decline in the choice of music. It is possible, according to the authors, that people choose music that gives them a boost.
The data also show some cultural differences – more energetic music, on average in Latin America, more relaxing music in Asia – and the gender difference in the hemisphere: women listen to less intensive music in the Northern Hemisphere and more intensively in the southern hemisphere
The seasonal intensity
the intensity peaks correspond to the summer solstice in each hemisphere, and these swings were more extreme in more extreme latitudes. Near the equator the changes in intensity were much flatter throughout the year, while the more northerly and southern places (which have larger variations in the length of the day) had bigger changes in musical preferences. The length of the day is better than that of a number of other options.
This may sound like a bright example of how people react to factors such as temperature or sunlight, but there may be a great deal of culture there, too. In a culture that has a long and distinctive autumn season, there will be associations of associations with what it means to fall – tastes, clothes and feelings – that can influence people's choices of music. More dance music during the summer break also makes sense. Obviously, these cultural factors are also influenced by biological and environmental factors, but everything is tied to a knot that is difficult to unravel. Work comes with warnings, of course. On the one hand, researchers relied on IP addresses and user profiles to classify users in their demographic and geographic regions. Users may be lying or using a VPN, so there is likely to be some vagueness in the data. And all the results are purely descriptive – they can not tell us anything about the causes of the models, such as whether people really use music to deliberately influence the mood. Intuition may try to fill in empty spaces, but to confirm these intuitions, different kinds of research will be needed – like lab tests.
Large arrays of data like this also contain a lot of noise that can be confused with patterns, especially if strange or confusing results are rejected and are not reported. And, of course, Spotify's users are a different (and relatively rich) demographic – not necessarily accurate model of all human experience.
Nature Human Behavior 2018. DOI: 10.1038 / s41562-018-0508 -z (For DOIs).