How Music Influences our Emotions, Feelings, and Behaviors | Dr. Amy Belfi | TEDxMissouriS&T

[Music] so this morning we all woke up and we had to make a series of decisions let's say the first thing we do is we go into our closet and have to pick out what shirt wear you want to wear next we go into the bathroom and choose from a variety of different fragrances and once we've gotten ourselves all ready for the day let's say we sit down at the breakfast table and start swiping left or swiping right on a potential date after all that we're ready to go to work we get in our car and we have to figure out what to listen to on the radio and while we're flipping through the stations it might sound a bit like this [Music] okay so all these choices what's where what to smell like who to date what to listen to these are what I would call aesthetic judgments or aesthetic decisions and basically what that means is just deciding what we like or what we find appealing versus what we dislike or find unappealing and over the course of our lifetimes we've built up our own unique aesthetic preferences or tastes to the point that making these types of decisions almost comes like second nature it's really easy for us to do like in the case of the car radio it seemed like we were able to flip through the stations within a few seconds so in my research I'm interested in figuring out exactly how much time it takes before we can tell whether or not we're gonna like a piece of music and really not just how quickly can we decide but do those snap judgments so those quick decisions actually line up with what we would decide given more time so if I had just stayed on that first station a little bit longer what I've eventually stuck with it or did I make the right decision to switch then so in my work I have kind of come to the conclusion that the snap judgments actually are quite accurate and in the next few minutes I'm going to lead you through some of the experiments that have helped me come to that conclusion so let's start with the science the first experiment I did we had participants come into the lab and they listened to 60 pieces of music these were electronic jazz and classical excerpts and the whole point was to try and see how quickly can people make aesthetic judgments of music so we varied the amount of time people heard each clip ranging from 250 milliseconds to 2,000 seconds so here's an example of what 250 milliseconds worth of music might sound like all right it's very short very very short okay so they would hear that and then they're asked how much do you like this piece of music raid it on a 1 to 9 point scale 1 you don't like it 9 you like it a lot ok so let's say we gave that a 2 then they would hear the same piece for 500 milliseconds all right we're at 500 milliseconds all right then we rate that again then they heard it for 750 milliseconds rated again a thousand milliseconds rate that and 2,000 milliseconds and rate that again okay so this is a little bit of an artificial scenario they weren't hearing the same exact piece over and over and over again they heard all 60 of these pieces in a mixed up order so the point of this is to see at what point - their decisions on these time durations line up with what they would decide if they listened to the entire piece of music well we don't have enough time to have everybody listen to full you know 4 minute pieces of music so we use ten-second clips as our long clip and that would sound like this [Music] okay so they heard 10 seconds and then we asked them again how much do you like this piece 1 to 9 point scale and let's say we gave out of 5 so then what we did to score the data is we compared all the earlier trials to the last trial to see when they matched up so in this case the ratings on the 750 1000 and 2000 millisecond trials were 5 which is the same rating that they had given on the 10 second trial so we scored those trials as correct because they matched up with our final rating the first two trials were then scored as incorrect so in the next slide I'm going to plot you the data for all the trials across all the participants here on the y-axis is the percentage correct and on the x-axis represents those different time durations the most important part I think of this graph is the horizontal dotted line which represents what you would expect given chance performance so that's the percentage correct we would expect if people were just randomly putting numbers in so what I was looking for is when do these diagonal lines cross the chance performance line and the three colors represent the three different genres of music so what's shown in this graph is that at 750 milliseconds for all genres people were performing better than chance which is pretty fast it's quite a bit faster than our example of the car radio which we had estimated to be several seconds so people make snap judgments and they're pretty accurate but something I noticed from looking at these data was that there's gonna be a difference between the classical and the electronic in jazz electronic and jazz people were even quicker at judging they were able to do that at the 500 millisecond clip so I thought okay could this be due to familiarity our participants are undergraduate students maybe they're really into EDM and jazz music and they were better able to judge things they were more familiar with so he did another version of this experiment directly testing this question the seemingly we used in this experiment were not classical electronic and jazz they were pop music and half of the pieces of music were familiar these were hit singles and half of the pieces were unfamiliar so the familiar hit single might sound like this [Music] okay so the unfamiliar pieces were songs by the same artists just not their hits so they were the b-sides something like this okay so unless you're a huge LMFAO fan I totally am you probably didn't recognize the second song but you did recognize the first one and we did a pretest to make sure that our familiar pieces were actually rated as being more familiar then we did the same exact experiment is the last one we had people listen to these familiar and unfamiliar pieces rate them at all the different time durations and then see where do their ratings match the rating of the ten-second piece so the next slide is gonna plot the same exact figure just with these new data so what's shown up here is that the dotted line are the familiar pieces the straight line are the unfamiliar pieces and for both groups people were above trance performance at 500 milliseconds so this is similar to what we saw for the jazz and electronic pieces but there was no difference between the familiar and unfamiliar pieces until we got out to the later time duration so at one in two seconds people were more accurate for familiar pieces of music so this is refining the conclusion a bit that people make these snap judgments and they're quite accurate but you do get information added over time so this led to another question how much information is added over time and for how long do you reach a certain point where the information you're getting is no longer going to influence your decision so we did a final version of this experiment looking at longer pieces of music in this task we played the same original electronic jazz and classical Clips was first 60 seconds each and during the time the participants were listening to the music we asked them to rate how much do you like this music right now and they used a rollerball mouse to move a visual slider like up here from the left beating I don't like this part right now so the right meaning I do like this part right now so I'm going to give you a brief example of what this might look like [Music] okay that was only ten seconds the real clips for 60 seconds long but let's say we plot that response with liking on the y-axis that rating and time on the x-axis it might look something like this so in addition to this continuous rating during the music we also ask participants the same old question how much do you like this piece of music overall give me a rating and so let's say in this case they rated it a 5 and then the outcome of this was to classify the pieces as being most liked or least liked based on that overall rating and then to see do these curves diverge at any point between the clips you altima Thalia liked and you ultimately disliked so what I have plotted in this graph here is on the y-axis that continuous rating curve and they're separated for the most like clips those are the clips people rated the highest overall and the least liked clips where the clips people rated the lowest overall and the lines you can just see eyeballing it start to diverge pretty quickly within the first few seconds I was interested in what's the earliest point I can detect a significant difference between these two lines and that having to be three seconds so pretty darn quick in three seconds is not the same as 500 milliseconds but in this case with having to make a motor movement we wouldn't expect it to be that quick so this is going to reinforces the conclusion that people make pretty accurate initial quick judgments of music but also it's evident in this graph that there is information added over time that 10 seconds looks different than 60 seconds but what seems to be the case is that people make a decision and they stick with it so you say I like this song and you just like it more and more the more you're listening to it so as you get back in your car to drive home today at the end of this event hopefully you can feel pretty confident that whatever radio station you end on is the right one for you at least from my research I've kind of come to the conclusion that the old adage of go with your gut ends up being true and hopefully we can take away from this that we should trust our intuitions about what we like and what we dislike Thanks you you

Channel: TEDx Talks

Views: 86,005

Keywords: TEDxTalks, English, Science (hard), Music (topic), Research, Science

Id: jPDKi-i618U

Length: 10min 13sec (613 seconds)

Published: Tue Oct 02 2018