Advantages of experimenting on yourself

In the coming months, I’ll be taking results from the solo response project to several conferences, and reviewer feedback has me worried about people dismissing this data because I collected the data from myself. I keep getting distracted by these imaginary confrontations with suspicious researchers so it’s time I lay down some concisely-expressed arguments to appease the hypothetical skeptics.

Problem 1: One subject = bad empirical research

I don’t like this one because the premise is wrong, but I’ve gotten it a lot already, so here goes.    Continue reading “Advantages of experimenting on yourself”

Solo Response recording set up

Before I forget everything, let me get down the setup details of the experiment I ran last summer (2012). Besides selecting the 25 pieces and working out where I was going to run the experiment, there was a lot of other relevant details. The following descriptions are for the purpose of documenting the experiment’s methodology; I hope anyone interested in employing these methods will seek higher authorities for instructions of best practices.

Experiment setup
Stephen McAdams was kind enough to let me borrow some CIRMMT equipment (Thought Technologies’ ProComp Infiniti and a pile of sensors) and occupy some of his lab space for a month. Though a little casual by some standards, I made myself a cubical out of spare sound absorption panels in a large room that was usually unoccupied while I was recording. To get data from the ProComp in a useful format, Bennet Smith helped sort out some of his old scripts that conveniently time stamped the physiological sensor data and packaged it as UDP messages. That left me with getting some system set up run the experiment, provide a behavioural response interface, and save the recorded responses in a reliable fashion.   Continue reading “Solo Response recording set up”

Variables of Valence

While I am hoping to soon start blogging about responses to each stimulus used in the solo response project, it will take a bit longer to get all the signals tidied and a format of analysis set.  In the interim, here is a simple version of the mini talk I presented at the most recent NEMCOG meeting with links to audio for each example.


Variability of Emotional Valence: Inconsistencies in self-report continuous emotion ratings – Finn Upham, New York University

Listeners often report feeling emotions in response to music, whether happy or sad. Empirical work on continuous reports of felt emotion are however often challenged by substantial variation in emotional dynamics reported by different participants. Variability in responses is supposed to be caused in part by differences in individual listeners’ musical expertise, sensitivity and cultural background. Working with multiple responses from a single human subject should then make it easier to explore which other factors contribute to the variation in felt emotions reported during music listening.

This summer, I collect continuous responses from myself (see the solo response project post). The analysis to follow uses the two dimensional felt emotion ratings (Arousal X Valence) as well as information from notes collected during the listening sessions.

Felt emotion from 24 listenings
Summary of felt emotion ratings from 24 listenings to Varud by Sigur Ros. Arousal and Valence ratings under rating change activity analysis and average rating time series.

Continue reading “Variables of Valence”

Studying myself

How else does a grad student get 40+ hours of experimentation time over 20 sessions from one human subject?

I thought someone might have an issue with my idea of using myself, but apparently not. The NYU ethics review office didn’t even want to hear about what responses I’d be recording, or what protocols I’d be following. Maybe they assumed that music is a harmless stimulus, or that I knew I wasn’t allergic to the glue we use to attached electrodes, or that I was very unlikely to sue the school because of research I’d planed myself. I hope they didn’t care because the data collection is happening at another institution, in collaboration with someone who’s standing ethics certificate does cover this type of experiment.

Since no one has tried to stop me, I am now over half way through collecting what should be a very interesting continuous response data set. I’m collecting felt emotion ratings (Valence X Arousal) with an optional third rating scale to report whether I was experiencing emotion from the perspective of a listener or a performer, or in between. I say optional because in practice, I haven’t consistently remembered to evaluate this as well as keep up with the emotion stuff. The physiological data being collected is skin conductance, temperature, blood volume pulse, chest expansion (respiration), and sEMG of the zygomaticus (cheek), the corrugator (eyebrow), and the trapezoid (back of the neck). These signals include information related to emotional arousal and valence, though the how and why can be a bit indirect.

The stimuli are fairly divers as well. Once the recordings are finished, I’ll go through each piece to discuss why it was chosen and how these responses related to what is in the music. But until then, here is the quick list of pieces (track, artist, album), in no particular order:

1. Varúð, Sigur Rós, Valtari
2. Bizness, Tune-Yards, Whokill
3. Visiting Hours, Shane Koyczan, Visiting Hours (spoken word)
4. The Littlest Birds, The Be Good Tanyas, Blue Horse
5. Wavin’ Flag, Young Artists For Haiti, Wavin’ Flag
6. Basket, Dan Mangan, Nice, Nice, Very Nice
7. Movimento Preciso e Meccanico from Chamber Concerto (Ligeti), Alarm Will Sound, a/rhythmia
8. Feel Good Inc., Gorillaz, Demon Days
9. Gong Hotel, Radio Radio, Havre de Grâce
10. String Quartet No. 14 in C-Sharp Minor, Op. 131: II. Allegro molto Vivace et III. Andante moderato, Artemis Quartet, Beethoven: String Quartets, Op. 131, Op. 18-2, Op. 132, Op. 59-3
11. Something to Write Home About, I Am Robot And Proud, Uphill City
12. Dutch, Dessa, A Badly Broken Code
13. Stampede, The Quantic Soul Orchestra, Stampede
14. Sinnerman, Nina Simone, The Definitive Rarities Collection – 50 Classic Cuts
15. The Stand, Mother Mother, Eureka
16. Invention No. 2 in C minor BWV 773, Glenn Gould, Bach: Two and Three Part Inventions and Sinfonias, BWV 772-801
17. 1685/Bach, Nosaj Thing, Drift
18. Thieving Boy, Cleo Laine, Wordsongs
19. Fortuna Imperatrix Mundi: O Fortuna, OSM, Orff: Carmina Burana
20. Lousy Reputation, We Are Scientists, With Love And Squalor
21. Boum!, Charles Trenet, Paris After Dark
22. Le Rosier De Trois Couleurs De Roses, STRADA, Gadje
23. Romance (Debussy), Rachel Talitman and Luc Loubry, The Golden Age of Harp and French bassoon
24. Ne m’oubliez mie/DOMINO (Mo 236), Anonymous 4, Love’s Illusion
25. Portal, Origin, Informis, Infinitas, Inhumanitas

This totals to nearly an hour and a half of music. During each sessions, I take notes between pieces, which adds some time as well. Some are just to warn of sneezes or yawns which affect the physiological data, others mark changes in how I am interpreting the piece and reflections on what I think triggers my responses of one type or another.

The ethics issue is one I am still worrying about because I would like to release this data for other researchers to use. We need more public data sets for evaluating new ways of extracting relevant information from these continuous response series, and to help the discussion on what is and isn’t captured by these measures. In the interest of science, sharing is good. But this data set will necessarily be full of problems, not the least of which being that it comes from me, a member of the research community. I am going to discuss this data processionally, and I can’t honestly refer to the subject as anonymous, or FU. As someone with many ideas and opinions about emotion and continuous responses and music, I am not a naive participant; this data cannot be treated as that of the “average listener”. Nor do I expect people will be all that comfortable digging into a particular person’s emotional experiences. There must be discussions the problems of self study in psychology research, perhaps I can find some ideas for sharing strategies from that literature.