4. Methodology
As this experiment is considered as an improvement to the previous experiments, the other cars that might appear on these pictures were controlled by removing any other cars that appear on these pictures, except the approaching one. Then on the opposite lane, pictures of one or two cars were added to act as distracters. By this method, the road was transformed from an empty road, to either mild or high busy road creating another factor that can be tested (Traffic Density factor). This factor can be acted as more realistic (see figure 4.1).
Figure 4.1 a. Examples of the target pictures that were used as stimulus. There were pictures of a junction and an approaching vehicle is coming toward the junction. The approaching vehicle was either a car or a motorcycle approaching from three different distances away from the junction: Near, Mid, Far. For each pictures, the number of cars appearing in the opposite lane was modified creating three type of road traffic density: empty, low, or high traffic. These are examples of the car pictures.
Figure 4.1 b. Examples of the target pictures that were used as stimulus. There were pictures of a junction and an approaching vehicle is coming toward the junction. The approaching vehicle was either a car or a motorcycle approaching from three different distances away from the junction: Near, Mid, Far. For each pictures, the number of cars appearing in the opposite lane was modified creating three type of road traffic density: empty, low, or high traffic. These are examples of the motorcycles pictures.
Figure 4.2. Examples of the non-target pictures that were used as stimulus. They were the same traffic scenes that were used for target pictures. The only change was the approaching vehicle, which was deleted.
The pictures were presented in a 20” computer monitor using E-Prime® presentation software, and a keyboard was used to collect responses. The distance between the subjects and the monitor was fixed at 90 centimetres.
4.1 Participants
Forty participants from Cardiff were divided into four groups. (2 male, 38 female, an average of 25.6 years of age, an average of 4.2 years of car driving experience)
4.2 Procedure and design
The main idea of this experiment is to see how drivers evaluate the level of danger across the different types of vehicles, and over the different type of location and traffic density of the road. The task that was chosen for this experiment was asking the participant whether they think it was safe to pull out front of the oncoming vehicle or not. The first parameter that can be tested in this study was the frequency of danger evaluation. This parameter represents the number of trial the drivers think it was safe to pull out. The second parameter for this study was the time duration needed to evaluate the picture and make the judgement over the varying type of appearance of the approaching vehicle (Decision Time).
As same as the previous experiments; to make sure that the participants were looking at the fixation cross, the (No-Go) task was added to the experiment.
After the fixation cross, a number between 1-8 appears for 250 milliseconds before the appearance of the target pictures. Then the participant was asked to press the “Space” button if the number appears was an “Odd” number. And if the number was “Even”, the participants asked do follow the original task and look at the junction and see whether there is an oncoming vehicle approaching or not. On each testing session, 20 No-Go trials were added to ensure that the participants were looking at the fixation cross. The data of any participant with less than 70% accuracy on the No-Go task was removed from the analysis for this experiment, because they were considered either did not understand the task, or did not pay sufficient attention during the experiment.
After the appearance of the fixation test, the target picture appears for 250ms. During this time, the participants should make their judgment whither they think it was safe to pull out, or not.
5. Results
A 4X2X3X3 mixed design was used in this experiment to analyse the target pictures. The “experiment group” was the between group’s factor with three levels: control, sound, imaginary and verbal. There were three within-groups factors. The first one was the type of vehicle approaching: Car or Motorcycle. The second factor was the traffic density of the road: Empty, Low, or High. The last within factor was the distance of the approaching vehicle with three levels: Near, Mid, or Far.
The experiment includes four different groups, and several factors that can be tested.
Therefore, the results and discussion sections for this experiment was divided into two sections to simplify examining the results. The first section looked at the behavioural data that includes: frequency of danger evaluation. The second section is the decision time.
5.1 Frequency of danger evaluation
Analysis of variance was carried to test the number of times participants evaluated the pictures as safe condition to pull out in front of the oncoming vehicle.
The analysis did not reveal any significant effect of answers between the groups F ( ) = 0.3408, MSE= , p>0.05, as all groups performed within the range of %- % (graph 1). Although a significant effect was not established, the graph indicates a reduction in performance for the image and verbal groups.
Graph1: count in identifying a vehicle for all distraction groups tested. The numbers are how many times the participant’s answer that it’s not safe in the stimuli.
The analysis revealed a significant effect on the traffic density
factor F ( ) = 4.874, MSe = , p <0.05.
Graph2: The difference in count identifying a vehicle for level of busy.
The analysis revealed a significant effect on the type of vehicle
factor F (1, 51) = 21.792, MSe = 24.395, p < 0.001
Graph3: The difference in count identifying a vehicle for cars & Motorcycles
Regarding the distance factor, the analysis also revealed a significant effect F ( ) = , MSE , p< .001. A post-hoc Tukey test showed a significant decrease in accuracy for the Far distance compared to Mid distance ( % vs. %, p< .001) and compared to near distance (78% vs. 96%, p< .001). On the other hand, the accuracy was almost the same for the mid and near distance (95% vs. 96%, p>0.05).
Graph 4: count in percentage for all three distances (Far, Mid & Near)
5.2 Decision time
Graph 8: Reaction time in ms for all four groups tested.
Graph9: : Reaction time in ms for vehicle for level of busy.
Graph 9: Reaction time in ms for Vehicles (car & Motorcycle)
Graph10: Reaction time in ms for all three distances (Far, Mid & Near)
6. Discussion
6.1Result Interpretations
It was therefore established that motorcycles exceed the threshold for conspicuity, but why are they less detectable and why do drives (ie. participants) sometimes fail to detect them at all? Studying road accidents, many reasons can be listed. One simple reason may be carelessness. A study by Langham (1999) which filmed drivers as they approached T-junctions reported that they spent very little time looking onto the road they were approaching; the mean was about 0.3 to 0.4 seconds. This in turn brings into question the amount of time given to the participants in our study. It may have been possible that 250ms be considered too little time especially for examining the image which contained a motorcycle at the very far end of the junction, thus explaining the relatively high reaction time shown in the study for far vehicles. Rasanen and Summala (2000) which studied the topic at a real-life situation found that drivers tended to look only in one direction as they approached junctions. Accidents were therefore more likely to ensue from collision with motorcycles approaching from the neglected direction. This is more associated with high speed drivers who are more likely to be selective about their observations. Nonetheless, for fear of admitting they hadn’t looked, accidents such as these were as a result still classified as `looked but failed to see`. On the other hand the LFBTS label was also given appropriately to situations where an approaching motorcycle’s view was actually blocked or obscured by street furniture such as mail boxes for example (Becklen & Cervone, 1983). Excluding those obstructed-view accidents, instances where the motorcycle situated at a close proximity to the offending driver still did not ‘see’ the motor bike (Olson, 1989). Wulf et al. (1989) reported that this may be largely explained by ‘cognitive conspicuity’ rather than ‘sensory conspicuity’. Cognitive conspicuity refers to what the driver expected to see. Many drivers, careful drivers, spend time looking out for approaching cars or trucks, because cars and trucks are what they expect to see on a motorway or road. This may explain the significant effect of interaction in this experiment which showed that cars may prevent other drivers from seeing motorcycles at car junctions. Some, especially those with no or low exposure or experience to or with motorcycles, do not expect to see motorcyclists and cyclists and as a result, `information about their presence does not reach conscious awareness` (Hole, 2007; Van Elsande, P., & Faucher-Alberton, 1997). This in turn sheds light onto the three participants in our study that did indeed have an experience of riding a motorbike. Although the effect of their observations provided little to alter the larger group, eliminating them from the subjects or replacing them with less experienced participants may have increased the accuracy of the results.
Hole (2007) argues that drivers approaching a T-junction have to make quick but deliberate decisions or their opportunity to pull out is lost. These quick scans of the scene or ‘shorthand codes’ as they have been termed by Hole and Tyrrell (1995) are based on the past scene perceptions of T-junctions that the drivers have had to cross in the past. Their experience had taught them to mostly look out for trucks, cars and buses and thus motorcycles and cycles have a different `spatial frequency content` and since attentional processes rely heavily on memory and experience and are not optimized for detecting and new objects, the presence of these vehicles is therefore not `registered` (Hole, 2007).
Even if one detects a motorcycle, an experienced driver may also face problems with determining the motorcycle’s speed-to-distance ratio due to the latter’s smaller frontal area, therefore providing less `cues to their speed in terms of radial expansion of their image on the retina` (Hole, 2007). A motorcycle that may be misperceived as far enough will actually arrive earlier than expected and thus collide with an early pull out by the offending driver. All these studies and observations conclude that sensory conspicuity is not sufficient for detection; drivers must also correctly interpret and appraise what they are seeing.
6.2Methodological issues
Before marking the conclusions and future implications of this study, the strengths and possible limitations of the methodologies employed will be briefly discussed. Several ways of studying driving psychology and road accidents exist. Some include extensive reviews of accident statistics, interviews and observational studies. Most psychologists’ method of choice would be the experiment. The experiment enables the researches to isolate cause and effect in a way that no other method can. The experimenter also owns complete control over variables. One can add further variables such as the effect of age and gender. In our studys, although the participants had the same average age, driving experience and general backgrounds (being university students), their sex ratios were uneven with 17.9 % only of the participants being male. Considering the gender differences in brain function and processing as proven by the scientific community, it may have been more advantageous to at least even-out the numbers or maybe separate the genders. Landeur, Armstrong, and Digwood (1980) found that women’s reaction rates are less than men’s by (485 ms to 534 ms respectively). Exposure rates (to motorcycles), discussed earlier, are also interrelated with gender as men drive more than women and are thus more likely to expose themselves to risk. It was mentioned above that the participants mostly formed students from the University of Cardiff which according to Hole (2007), is not representative of the general population. On the other hand, he argues that public members recruited for experiments fail to realize that experimenters are interested in obtaining data that can be averaged with that of other participants and not in their individual performance, thus preferring psychology students over members of the public after all. The sample size of one hundred also is relatively small and a larger sample sized would definitely increase the accuracy of the results and the number of responds in the second experiment.
6.3Attentional spotlight
Eriksen and Hoffman (1973) described how attention is like a spotlight; it highlights selected information and leaves other information `outside the focus in the dark`. So like a spotlight, one’s attention moves to different space regions to `illuminate` anything there. Spotlight theory was also examined by studying the movements of attention. Tsal (1983) studied the movements of attention at three distances from fixation and found that `the time needed to focus attention at a location increased as the eccentricity of the target increased.` This may explain how the participants in this paper’s study found it easier to spot closer objects than further ones. Whether they spotted the object nearest to the fixation cross before viewing the image, regardless of the location of the vehicle, has not been analysed and may be subject to further investigation in future experiments. Eriksen and Murphy (1987) confirmed the fact that indeed `more time is needed to identify more distant objects because of poorer visual processing in the periphery of the visual field.` However, tracking one’s attention as explained above requires tracking eye movements, which was not the purpose of this study. Studies which employ eye tracking devices in their methods may therefore take on such issues.
Conclusion
Attention is a vital aspect of visual processing. These studies attempted to observe the difference in the attention given by drivers to cars and motorcycles. In general, it was found that cars were more readily spotted than motorcycles when alone or combined in one scene. Distance did indeed play a factor with closer vehicles being spotted more easily than ones situated far away. Because the visual system can only give high analysis processing for only a small portion of a scene at any given moment, attention needs to play a role in complementing the visual processing system. It was shown that attention is not only dependent on sensory factors, but also largely on cognitive factors. Cognitive factors in turn are dependent on several environmental, social and exposure factors. Those related to motorcycle conspicuity have been discussed at depth. Further research is needed to examine the role of motorcycle conspicuity in road traffic accidents, ways to enhance drivers’ familiarity with motorbikes and more advanced methods of research that are able to translate experimental conditions to real-life situations.
References:
Becklen, R., & Cervone, D. (1983). Selective looking and the noticing of unexpected events. Memory and Cognition, 11, 601-608.
Brown, I. D. (2002). A review of the `looked but failed to see` accident causation factors. Behavioural Research in Road Safety XI. Department for Transport, Local Government and the Regions, London, UK.
Cairney, P., & Catchpole, J. (1995). Patterns of perceptual failure at intersections of arterial roads and local streets. In A.G. Gale (Ed.), Vision in vehicles VI. Amsterdam: Elsevier.
Crundall, D., Humphrey K., & Clarke D. (2008). Perception and appraisal of approaching motorcycles in junctions. Transportation Research Part F, 11, 159-167.
Eriksen, C. W., & Hoffman, J. E. (1973). The extent of processing of noise elements during selective encoding from visual displays. Perception & Psychophysics, 14, 155-160.
Eriksen, C. W., & Murphy, T. D. (1987). Movement of attentional focus across the visual field: A critical look at the evidence. Perception and Psychophysics, 40, 225-240.
Fulton, E. J., Kirkby, C., & Stroud, P. G. (1980). Daytime motorcycle conspicuity (Suppl. Rpt. No. 625). Crowthorne, Berkshire: Transport and Road Research Laboratory.
Gredebäck, G., Johnson, S., & Von Hofsten, C. (2010). Eye Tracking in Infancy Research. Developmental Neuropsychology, 35, 1–19.
Henderson, J. M., & Hollingworth, A. (1998). Eye movements during scene viewing: An overview. In G. Underwood (Ed.), Eye guidance in reading and scene perception (pp. 269-294). Amsterdam: Elsevier.
Hesketh, B., & Godley, S.T. (2002). A comparison of time estimations in driving with target only in motion, self-only in motion, and self-and-target in motion. Ecological Psychology, 14(3), 111-125.
Hole, G. (2007). The Psychology of Driving. New Jersey: Lawrence Erlbaum Associates, Inc.
Hole, G. J., & Tyrrell, L. (1995). The influence of perceptual “set” on the detection of motorcyclists using daytime headlights. Ergonomics, 38(7), 1326-1341.
Landauer, A. A., Armstrong, S., & Digwood, J. (1980). Sex differences in choice reaction time. British Journal of Psychology, 71, 551-555.
Langham, M.P. (1999). An investigation of the role of vehicle conspicuity in the “looked but failed to see” error in driving. Unpublished PhD thesis, University of Sussex.
Olson, P. L. (1989). Motorcycle conspicuity revisited. Human factors, 31(2), 141-146.
Rasanen, M., & Summala, H. (2000) Car driver’s adjustments to cyclists at roundabouts. Transportation Human Factors, 2(1), 1-17.
Rayner, K. (Ed.) (1992). Eye movements and visual cognition: Scene perception and reading. New York: Springer-Verlag.
Recaarte, M. A., & Nunes, L. M. (2000). Effects of verbal and spatial-imagery tasks on eye fixations while driving. Journal of Experimental Psychology: Applied, 6(1), 31-43.
Tsal, Y. (1983). Movement of attention across the visual field. Journal of Experimental Psychology: Human Perception and Performance, 9, 523-530.
Van Elsande, P., & Faucher-Alberton, L. (1997). When expectations become certainties: A potential adverse effect of experience. In T. Rothengatter & E. Carbonell Vaya (Eds.) Traffic and transport psychology: theory and application, Amsterdam: Pergamon.
Van Loon, E. M., Khashawi, F., & Underwood, G. (2010). Visual strategies used for time-to-arrival judgments in driving. Perception advance online publication, doi:10.1068-6067.
Van, R., & Van, G. (2007). Eye Movements: A Window on Mind and Brain. Amsterdam: Elsevier.
Williams, M. J., & Hoffman, E. R. (1977). The influence of motorcycle visibility on traffic accidents. Melbourne, Australia: University of Melbourne, Department of Engineering.
Wulf, G., Hancock, P. A., & Rahimi, M. (1989). Motorcycle conspicuity: an evaluation and synthesis of influential factors. Journal of Safety Research, 20, 153-176.
