http://bps-research-digest.blogspot.com/2012/08/music-we-like-is-more-distracting-than.html
fulltext:
http://dl.dropbox.com/u/85192141/2012-perham.pdf
> In lab studies, people who listen to music they like, generally perform better at mental tasks afterwards, an effect that's been attributed to boosts in mood and arousal. But what about the effect of background music that plays on during a task - more akin what we do in real life? This is actually less studied. The traditional mood-arousal literature would predict it to be beneficial too, especially if the music is to the listener's taste.
>
> However, there's another line of research, known as the "Irrelevant Sound Effect", that's all about the way background sounds can interfere with our short-term memory for ordered lists, which would be a bad thing for many work-related tasks. These studies show that the distraction is greater when the sound is more acoustically varied - just like your typical pop song. Based on this, Nick Perham and Martinne Sykora made a counter-intuitive prediction - background music that you like will be more detrimental to working memory than unappealing music, so long as the liked music has more acoustical variation than the disliked music.
>
> Twenty-five undergrads completed several serial recall tasks. They were presented with strings of eight consonants and had to repeat them back from memory in the correct order. Performance was best in the quiet condition, but the key finding was that particiants' performance was worse when they completed the memory task with a song they liked playing over headphones (Infernal's "From Paris to Berlin"), compared with a song they disliked (songs such as "Acid Bath" from the grind core metal band Repulsion). In case you're wondering, participants who liked Repulsion were excluded from the study.
Particularly relevant to my suspicions:
> A further intriguing detail from the study is the participants' lack of insight into the degree of distraction associated with each type of music. Asked to judge their own performance, they determined correctly that their memory was more accurate in the quiet condition, but they didn't realise that their performance was poorest whilst listening to the music they liked.
--
gwern
http://www.gwern.net