When do we find it difficult to do more than one thing at a time? What theories have been proposed to account for such difficulties?
In some situations humans can do two things at the same time quite easily, for example writing an essay and listening to an album. In other cases attempting to do two tasks simultaneously will lead to serious disruption of both tasks, for example writing an essay and talking on the phone. Evidence suggests that the amount of disruption experienced when we try to do two things at once depends upon the similarity and difficulty of the tasks, as well as the amount of practise an individual has of them. Theories that attempt to account for such difficulties have to specify the level at which a limitation occurs, i.e. input (perception), output (action), or some intermediate stage of information processing. Relevant theories include Single Channel theories (E.g. Broadbent, 1958), Central Capacity based theories such as Kahnman (1973), Modality Specific theories (E.g. Wickens, 1984), and Baddeley and Hitch’s (1974) Working Memory theory, which involves both a Central Executive and Modality Specific subsystems.
One of the most important factors in determining how well two tasks can be performed at the same time is their similarity. Similar tasks are said to rely on the same ‘processing resource’, resulting in disrupted performance. This can involve both input (visual versus auditory) and output (verbal versus manual) modalities. Allport et al (1972) found that when people did a shadowing task (repeating back auditorily presented passages) whilst simultaneously learning auditorily presented words their recognition memory for the words was weak. Conversely, if they had to learn visually presented pictures, whilst shadowing, then subsequent recognition memory was good. This suggests that two tasks can be adequately completed if they rely on different input modalities. However it would be interesting to include a further condition involving visually presented words. If different input modalities result in adequate dual-task performance the recognition memory for visually presented words would be the same as for visually presented pictures. Alternatively visually presented words may be converted into auditory representations for processing. If this is the case then recognition memory for visually presented words would be as poor as for auditorily presented words.
Dual-task performance also depends upon how similar the outputs/responses of the tasks are. McLeod (1977) conducted an experiment involving responding to a tone either vocally or manually whilst simultaneously performing a continuous tracking task. Performance on the tracking task was worse when the response to the tone was similar to the tracking task (manual) than when the response to the tone was different (vocal).
A second factor that has been identified as effecting dual task ability is the difficulty of the tasks involved. Though it seems intuitively true that more complex tasks will hamper dual-task ability, it is difficult to assess the difficulty of a particular task, relative to a particular subject. One way of testing the effect of task difficulty on performance involves using varying levels of redundancy in shadowing tasks. Eysenck and Keane (2000) provide the example of Sullivan (1976) as a test of the effects of redundancy on dual task performance. The concept of redundancy involves the predictability of information. Highly redundant material allows the subject to predict sequences of words, letters or notes, whilst less redundant material will be comparatively unpredictable. Sullivan’s (1976) work involved subjects completing an auditory shadowing task, whilst trying to identify target words in a second, un-shadowed message. By varying the redundancy of the shadowing task, it was found that a more redundant message enabled greater accuracy in identifying the non-shadowed targets. These results support the idea that increasing difficulty of a particular task has a negative effect on our ability to perform other tasks simultaneously.
The third factor effecting dual-task performance is how well practiced the tasks are. For example, novice musicians find it difficult playing their instruments and singing simultaneously, whereas experienced musicians can perform both easily. Spelke, Hirst and Neisser (1976) conducted experiments involving subjects writing to dictation while simultaneously reading passages. The quality of their writing and their understanding of the reading task was then assessed. At first the performance of these two tasks was poor, but after six weeks of practice performance at both was improved, so that the reading tasks were understood as well as when performed as an individual task. Spelke et al (1976) observed that whilst the reading tasks were understood, the subjects were relatively unaware of the meanings in the dictation tasks, but this also improved with further practice until they were able to identify the semantic categories of dictated words, without effecting their performance on the reading tasks. This demonstrates how experience can dramatically improve the performance of dual-tasks.
The evidence presented above shows that people find it difficult to do more than one thing at a time when tasks are similar, difficult, and unpractised. These are the difficulties that theories of dual task performance must account for. Early theories of attentional mechanisms postulated a single channel through which all information is processed. This type of theory views deficits in dual-task processing as being due to a ‘bottleneck’ in the single channel.
Evidence in support of a ‘bottleneck’ in attention processing comes from the work of Welford (1952) on the psychological refractory period. Work on this effect involves presenting subjects with two tasks, both requiring responses to stimuli, in quick succession. Where the stimulus for the second task occurs during the processing of the first task, there is a measurable delay in response time for the second. This delay in response time is the psychological refractory period. The PRP occurs, according to Welford, because our ability to decide upon responses can deal only with one task at a time. The second task, stimulated during response processing for the first, is forced to queue until the first is completed.
Welford (1952) along with Broadbent (1958), used these findings to suggest a single channel theory of attention, whereby attentional tasks are dealt with in a serial manner, one after each other, with a filter selecting from competing tasks for response. These single channel theories propose to explain our difficulty in performing dual-tasks in terms of a bottleneck occurring at some point in the attention process. When we perform two actions together, therefore, our attention is alternating between tasks, acting only in a serial manner.