0014: Cars & Phone experiment
Day 31.
If I were to design a phone and driving experiment, I would test the GPS mapping functionalities. Stephen mentioned a school shooting and direction detection experiment here at Iowa State and I think something like that would be excellent for testing phones while driving.
Question
Does GPS type (audio vs visual vs combination) affect distracted driving level?
Variables
Independent: The type of GPS and the type of driving environment
Dependent: The level of distracted driving
Design
We will have two lists of routes: a list of cities with high, but similar, traffic styles. we will also have a list of long distance routes with complex directions, long stretches of roads, and similar traffic and stops.
We will select subjects from both city and rural driving areas at home in all states to get a large cross section of data. We will be testing the subject’s data against themselves.
Subjects will be presented with lists of cities and driving routes. Subjects will then mark which cities they have visited and which routes they have driven over 51% of*.
Each subject, based on the lists they have provided us, will be asked to drive 8 different routes total, three in cities they have never visited and three long routes that are unfamiliar to them. One will be a city route that they are familiar with to gauge driving style and another will be a long route that they are also familiar with to gauge rural driving style.
The first route driven for both long distance and city will be a combination of audio played through the car speakers and a visual map displayed by a phone mounted on the dash. Both groups will use a proprietary software for directions (similar to google or apple maps). Route two will be an audio only provided route and route three will be a visual only provided route.
Data Collection & Measurements
We will be measuring how distracted drivers are while driving. The variables we will be collecting are:
- time eyes left the road
- number of times of hard breaking
- number of accidents which occurred during the test
- time while in hard acceleration
- number of time evasive maneuvers had to be conducted
- number of missed turns
- number of illegal turns and lane changes
- number of times they were pulled over by officers
- how long their speed was greater than 5 over the speed limit in favorable conditions
- how long their speed was under 5 under the speed limit in favorable conditions
- how far off predicted arrival time they were
Based off of these measurements we will determine the relative levels of both distracted and aggressive driving, comparing their 6 GPS routes to their regular routes.
*we will decide 51% of routes driven by presenting them with a map of all the routes we will be conducting experiments on marked on the map. Participants will then be asked to draw a line through the areas where they have traveled or driven through. We will then measure the distance of each length of subject drawn line to calculate percentage along each route. If there are several routes that are below 51% we will select the three routes with the lowest percentage totals.
Hypothesis
Visual based GPS driving will cause more hazardous driving conditions while audio-based GPS will cause longer route times.
Alee, nice study idea! I like the way you operationalized the DV of distracted driving into a nice list of behavioral markers that are actually countable. You’ve got more IVs in there than just GPS style, it looks like: rural vs city, familiar vs. unfamiliar, long vs short routes. Some researchers might call those moderating or mediating variables, depending on how you think they influence the results. I’d probably suggest starting off simple and adding variables like that on later.