Resultados de las características métricas

Anderson (1982) in his Adaptive Control of Thought (ACT) model separates the process of skill acquisition into three main stages: declarative, knowledge compilation and procedural. Declarative knowledge consists of factual information about the processes involved in a given activity. It is explicit information, in that a person can report it. It provides a semantic network of facts about items within a domain, which the learner must put together with general problem-solving strategies to perform tasks within the domain (Groeger, 2000). Productions i.e. condition- action rules are formed on the basis of the outcomes of the application of declarative strategies in different circumstances. These productions are different from declarative knowledge in that the production is committed to a specific goal (Anderson, 1992; Anderson, Matessa, & Lebiere, 1997). In the knowledge compilation stage, sequences of productions that follow each other in solving a particular problem are condensed into a single production. This is known as composition. This process serves to speed up the action process, as there is no longer any need to think though the various steps involved in task production (Anderson, 1982). The outcome of repeated use of the same production is known as proceduralisation. Skilled behaviour is seen as procedural in nature, and is often implicit, revealed in outcomes rather than conscious awareness of the routines themselves (Anderson et al., 1997). In Anderson’s initial model, it was claimed that all knowledge first came into the system in declarative form, but the most recent version of the ACT-R (Adaptive Control of Thought – Rational) claims that most learning involves the acquisition of both types of skill. However, some declarative

knowledge must be acquired before proceduralisation is possible (Anderson & Fincham, 1994). Anderson’s model predicts that learners will benefit from on-going instruction and feedback on performance, by enhancing the declarative knowledge available to the learner and directing the learner’s attention to those aspects of the task that are important. However, the theory does not predict that classroom-based education will be an effective means of learning to drive, and any effect of such education would be expected to diminish as the learner gains experience of actually performing the task (Groeger & Banks, 2007).

As a skill becomes more practised, it becomes automatic and thus, interferes less with a concurrent task, and is less interfered with by a concurrent task (Anderson, 1992). A number of studies have investigated the idea that driving is an automatic task, and therefore that once a driver gains sufficient experience, full proceduralisation will have taken place (McKenna & Crick, 1997). These studies have failed to find any evidence that this is the case. Strayer and colleagues have conducted in-depth analyses of mobile phone use while driving (Strayer & Drews, 2004, 2007). In a series of driving simulator studies they have found that the use of a mobile phone or a hands free kit while driving led to slower response times, greater following distances and a two-fold increase in the number of rear-end collisions. Lansdown, Brook-Carter, and Kersloot (2004) used a driving simulator to examine interference effects on driving. In their experiment, participants were initially asked to drive as they normally would until a number was presented on the left screen, after which they were told to press a particular key to dismiss the number depending on whether it was odd or even. A separate group of participants were told to press a key to dismiss a letter appearing on the right screen depending on whether it was a vowel or a consonant. Both of these tasks occurred simultaneously with the driving task. Results indicated that participants maintained a shorter headway when completing the secondary tasks than a control group who completed no secondary task. Participants also maintained a higher speed in the control task than in the secondary tasks. The introduction of either secondary task led to detriments in performance, but the introduction of both the letter and number task at the same time led to the greatest detriments in performance, indicating that simultaneous interaction with multiple secondary tasks leads to increased mental workload and further degradations in performance than one secondary task alone. These studies provide evidence that even

17/12/2013 Ruth Madigan 86 experienced drivers have not adopted a fully proceduralised/automatic ability to complete all of the tasks involved in driving, as dual task studies show evidence of distraction effects across a number of spectrums.

Groeger and Clegg (2007) have found that a power law relationship exists between amount of driving practice and driving instructors’ comments to pupils, with the frequency of comments declining in a systematic manner across the training. This decrease in comments reflected the number of times the manoeuvres were repeated, not just the time spent behind the wheel, suggesting that some aspects of driving skill are acquired faster than others e.g. simple manoeuvring tasks such as moving off or stopping. This research suggests that although some elements of driving may become proceduralised relatively quickly, other elements of the task will take longer to become automatic, and may never do so. Interestingly, an earlier study showed that the one area of driving training where instructor comments do not decrease in a power function is where comments were concerned with appreciation of risk and likely behaviour of other drivers (Groeger, 2001). However, the terms ‘risk’, ‘danger’ and ‘hazard’ comprised only two per cent of all instruction given, suggesting that learner drivers are not receiving an opportunity to improve their hazard perception skills, and thus are not developing procedural knowledge regarding hazardous elements in the environment.

Although the effects of distraction have rarely been examined in relation to hazard perception skill, a few studies have looked at the effects of a secondary task. McKenna and Crick (1997) argued that if driving was an automatic task, with practice it would become relatively undemanding, allowing the driver to carry out other tasks without interference. They liken hazard perception skill to running a simulation, claiming that drivers are actively involved in constructing and running a predictive model, thus the task is not automatic. To test this, they conducted a study whereby drivers had to listen to a continuous sequence of speech and make appropriate responses while at the same time completing a hazard perception test involving button press responses to hazards presented on a screen. It was found that the secondary task led to interference in the time taken to detect hazards. This supports the notion that the detection of hazards is not an automatic process, and requires declarative processing of the driving situation.