Classical Music and Productivity: Evidence from a Randomized Control Trial in a Call Center

How to cite

Proto, Eugenio; Oswald, Andrew. 2016. "Classical Music and Productivity: Evidence from a Randomized Control Trial in a Call Center." AEA RCT Registry. AEARCTR-0001146. https://doi.org/10.1257/rct.1146-6.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Eugenio Proto* , Andrew Oswald

* Primary investigator

Registry ID
AEARCTR-0001146
DOI
https://doi.org/10.1257/rct.1146-6.0
First registered
Last updated
Status
Completed
Trial dates
2016-04-10 – 2016-05-07
Countries
United Kingdom of Great Britain and Northern Ireland (South Yorkshire)
JEL codes
J01, J8, I31
Registry record
http://www.socialscienceregistry.org/trials/1146

Abstract

This study will examine the effect of music on workers' productivity. The setting is a large call-centre business in the United Kingdom, which
has agreed to cooperate with our research team. This study will use a randomized trial to find out whether, in a real workplace, productivity
is higher when workers hear background classical music. The control condition will be hearing no music. Music will be played at randomly
assigned times during the day. The identities of the employees will change through time, but the playing of the music will be orthogonal to
those identities. To reduce possible problems from the multiple-comparisons problem (also known as the multiple-hypothesis testing problem), only one productivity variable will be studied in this project, and only one kind of music will be played. To gain the necessary statistical power, the experiment will take place over approximately 2 weeks.

Intervention

This study will use a randomized trial to find out whether, in a real workplace, productivity is higher when workers hear background classical music. The control condition will be hearing no music. We will also collect data a week before and a week after the intervention. Music will be played at randomly assigned times during the day, for the entire period between 18th April and 1st of May. The working day is from 7am-11pm (16 hours). The music will be played throughout the day withing interval of 2 hours each. Each two hours-slot will be randomly selected ex-ante using Stata. The music will be selected by the expert of classic FM in a way to induce happiness and avoiding sadness. The identities of the employees will change through time, but the playing of the music will be orthogonal to those identities. In total we will get data from 100 employees, not informed about the experiment.

Experimental design

This study will use a randomized trial to find out whether, in a real workplace, productivity is higher when workers hear background classical music. The control condition will be hearing no music. We will also collect data a week before and a week after the intervention. Music will be played at randomly assigned times during the day, for the entire period between 18th April and 1st of May. The working day is from 7am-11pm (16 hours). The music will be played throughout the day withing interval of 2 hours each. Each two hours-slot will be randomly selected ex-ante using Stata. The music will be selected by the expert of classic FM in a way to induce happiness and avoiding sadness. The identities of the employees will change through time, but the playing of the music will be orthogonal to those identities. Employees are not informed about the experiment

Design details

Randomization method
Randomization done in office, ex-ante by a computer
Randomization unit
Slot of two hours during the working day (from 7am-11pm everyday for approximately 2 weeks) when the music will be played and when the music is not played.
Clusters
112 time slots of two hours each.
Observations
100 subjects in an unbalanced panel of 112 time slot.
Arms
112 slots, about half are control slot. With 100 employees in total. The identities of the employees will change through time, but the playing of the music will be orthogonal to those identities.
Primary outcomes
Our main measure of productivity: -ACW (after call work) is the total in seconds within a time slot- the average number of seconds it takes to deal with the query (admin wise etc) post phone call -AHT (average handling time) is in seconds within a time slot- the average number of seconds it takes to deal with the phone call

Cited by

No citation data available for this registration yet.

Keywords