HCI assignment: Study on smartphones and driving safety
Title: The effect of music familiarity on drivers attention.
Most drivers listen to music in their car, often from a smartphone that is connected to the car via Bluetooth or a cable. My own observations suggest that when drivers listen to music that is unfamiliar, they tend to either look down at their phone or the central display in the car to skip or like the song. Both of these options divert the drivers’ attention away from the road. This study is intended to understand the relationship between the familiarity of music being played in the car and the drivers’ attention and behavior on the road.
Research question: How does the familiarity of the music playing in the car affects drivers’ attention to the road.
Variables:
- Independent variable: The familiarity of the music playing in the car via the drivers’ phone (unfamiliar vs. familiar)
- Dependent variable: Drivers’ attention to the road
Hypothesis: When drivers listen to unfamiliar music, they are more likely to become distracted and behave recklessly while driving.
My study will be designed in the following ways:
- Methods:
- I will recruit about 20 participants from diverse backgrounds and of different ages.
- I will start with 20 participants, but may increase this number as needed in order to achieve statistically meaningful results.
- Equipment
- A driving simulator to model a controlled driving environment.
- An eye tracking device to record the number of times the driver looks away from the road to change or like a song
- 2 Music playlists:
- The drivers’ liked songs
- A playlist with songs from a mix of different genres and time periods
- Procedure:
- Record baseline data
- The driver will drive in the car in silence to establish a baseline attention level.
- Experimental method for measuring data
- I will randomly assign participants to 2 different test groups. Each participant in the first group will drive while listening to a playlist of their liked songs. Each participant in the second group will drive while listening to a playlist that has been created with songs from a mix of different genres and time periods.
- Drivers in both groups will drive in the same conditions: on the interstate with stop-and-go traffic for 5 minutes, on the interstate with no traffic for another 5 minutes, in a quiet neighborhood, and through a city.
- Record baseline data
Limitations:
- Some driver participants in group 2 may still recognize some of the songs that play in the playlist of songs of different genres and time periods. This may interfere with the results.
- The driving simulator may not depict completely realistic conditions.
Nice design. I appreciate the acknowledged limitations. The driving task or instructions given may also affect things, e.g., “Get there as fast as you can” vs. “Count how many trees you see en route” or “Enjoy a leisurely drive.” You could ask for people’s music preferences beforehand, perhaps having them rank a list of diverse songs as familiar – unfamiliar on a Likert scale.
You mention starting with 20 participants but increasing as needed for significance. That’s actually one of the no-nos in experimental research. You might ask ChatGPT or Claude “If I’m running an experiment, why is it bad to keep running participants until I have the significance I want?” or just google p-hacking. (Instead, you do a power analysis beforehand.)
I am assuming the baseline is how many times someone looks away from the road. Anecdotally, I get distracted when it is not my playlist.