Multi-tasking on the Information Super Highway: Why Using a Cell Phone Can Make You Drive Like You’re Drunk David Strayer Department of Psychology RMPA: April 13, 2007

Research Questions  Does conversing on a cell phone interfere with driving?  What are the sources of the interference?  Peripheral interference (dialing, holding the phone)  Attentional interference (cell phone conversation)  How significant is the interference?

Observational Study (Residential 4-way Intersections)  Odds ratio for failing to stop:  0.27 for drivers not using a cell phone  2.93 for drivers using a cell phone Traffic Violation No Traffic Violation On Cell Not on Cell

Epidemiological Studies (Case Crossover Design)  Redelmeier & Tibshirani (1997) NEJM  699 driver involved in a non-injury automobile accident  4-fold increase in risk of accident when using cell phone  McEvoy et al., (2005) BMJ  456 drivers requiring hospital attendance after automobile accident  4-fold increased likelihood of crashing when using a cell phone

Driving Simulator Studies (Car Following Paradigm)  Drivers conversing on a cell phone were 5 times more likely to be involved in a traffic accident Traffic Accident No Accident Single-Task Dual-Task

High-Fidelity Driving Simulator

Simulator-Based Studies  Car-following paradigm  Follow periodically braking pace car  Required timely and appropriate reactions  Hands-free cell phone (positioned in advance)  Conditions  Single vs. dual-task  Low vs. moderate density *  Measures  Reaction time  Following distance  Rear-end collisions Low Mod. SingleDual

Reaction Time

Following Distance

Rear-end Collisions

Summary (Experiment 1)  Cell-phone driver’s  Slower reaction times  Drivers compensate by increasing following distance  Increase in rear-end accidents  Cell-phone interference  Even when manual contributions are eliminated  Naturalistic conversations

Why Do Cell Phones Cause Interference?  From earlier studies, no interference from:  Radio broadcasts (audio input)  Books on tape & recorded conversations (audio/verbal input)  Simple shadowing (audio/verbal input, verbal output)  Implies active engagement in conversation necessary  Impairments from both hand-held and hands-free units  Implies central / cognitive locus  Inattention-blindness (James, Neisser, Simons)

Why Do Cell Phones Cause Interference?  From earlier studies, no interference from:  Radio broadcasts (audio input)  Books on tape & recorded conversations (audio/verbal input)  Simple shadowing (audio/verbal input, verbal output)  Implies active engagement in conversation necessary  Impairments from both hand-held and hands-free units  Implies central / cognitive locus  Inattention-blindness (James, Neisser, Simons)

Why Do Cell Phones Cause Interference?  From earlier studies, no interference from:  Radio broadcasts (audio input)  Books on tape & recorded conversations (audio/verbal input)  Simple shadowing (audio/verbal input, verbal output)  Implies active engagement in conversation necessary  Impairments from both hand-held and hands-free units  Implies central / cognitive locus  Inattention-blindness (James, Neisser, Simons)

Experiment 2: Inattention-Blindness  Test for evidence of cell-phone induced inattention blindness  High-fidelity driving simulator  Hands-free cell phone  Naturalistic conversation with confederate  Eye tracker  Two phases to the study:  Phase 1: Single & dual-task driving  Phase 2: Recognition memory tests for objects encountered while driving

Recognition Memory

Recognition Memory Given Fixation

Experiment 2a: Summary  50% drop in recognition memory from single to dual-task, consistent with inattention blindness interpretation  What about items more relevant to safe driving?  Do drivers divert attention from processing items of low task relevance (e.g., billboards), but protect high task relevance items (e.g., pedestrians)?

Experiment 2b: Effects of Traffic Relevance  Phase I: Single & dual-task driving  Interstate driving (with traffic)  Hands-free cell phone, naturalistic conversations  Unique items placed in single & dual-task scenarios  Phase II: Surprise 2AFC recognition memory test  Single-task items (driving only)  Dual-task items (driving & phone)  Control items (not seen while driving)

Driving Safety Relevance Ratings

2AFC Recognition Memory Given Fixation (Corrected for Guessing) Dual-task interference not modulated by driving safety relevance

Experiment 2: Summary  Cell phone conversations create inattention blindness for traffic related events/scenes  Cell phone drivers look but fail to see up to half of the information in the driving environment  No evidence that cell phone drivers protect more traffic relevant information

Experiment 3: Encoding or Retrieval Deficits?  Encoding deficits  Reduced attention to perceptual inputs  Clear implications for traffic safety  Retrieval deficits  Failure to retrieve prior episodes  Less clear implications for traffic safety  Event-related brain potentials recorded to traffic brake lights  Single-task  Dual-task

Traffic-related Brain Activity

Experiment 3: Summary  Brain waves suppressed by cell phone conversations  Cell phone conversations impair encoding of information necessary for the safe operation of a motor vehicle

Cell Phone vs. Passenger Conversations  Conditions  Single task / dual task  Conversing on cell phone  Conversing with passenger  Design  Task (2) x Condition (2) Single task Cell Passenger

Lane Keeping Errors

Successful Navigation

Traffic References

Summary (Experiment 4)  Cell-phone conversations  More navigation errors  Fewer references to traffic  Passenger conversations  Collaborative problem solving  Shared situation awareness  Passenger actively supports the driver

Experiment 5: How Significant is the Interference?  Cell-phone vs. drunk-driver  Redelmeier and Tibshirani (1997) reported epidemiological evidence suggesting that “the relative risk [of being in a traffic accident while using a cell-phone] is similar to the hazard associated with driving with a blood alcohol level at the legal limit” (p. 465).

Cell-phone Driver vs. Drunk Driver  Car-following paradigm  Follow periodically braking pace car  Required timely and appropriate reactions  Conditions  Single-task driving  Cell-phone driving *  Intoxicated driving (BAC= 0.08 wt/vol)  * Hands-free = Hand-held

Reaction Time

Following Distance

Rear-end Collisions

Summary (Experiment 5)  Compared to drunk driver, cell-phone driver’s reactions  Slower reaction times  Longer to recover lost speed following braking  Drivers compensate by increasing following distance  Increase in rear-end accidents  When controlling for time on task and driving conditions, cell-phone drivers’ performance is worse than that of the drunk driver

You Cannot Practice Away The Interference Cooper & Strayer et al., (2007) HFES Accident Frequency for City and Highway Driving Day 1 Training Day 4 Training Day 4 Transfer Single Task Dual Task

Teen Drivers More at Risk Chisholm et al., (2006) HFES Accident Percentage – pedestrian and pullout vehicle events Single-TaskDual-Task Novice (< 6 months) 3.3%6.4%4.85% Experienced (10 years) 0.0%2.5%1.25% 1.65%4.45%

Research Questions  Does conversing on a cell phone interfere with driving?  Yes  What are the sources of the interference?  Peripheral interference (dialing, text messaging)  Attentional interference (inattention blindness)  How significant is the interference?  Worse than listening to radio/books on tape  Worse than in-vehicle conversations  More impairing than driving while intoxicated at legal limit