Early vs. Late selection models of attention

‘What we see, hear, feel and remember depends not only on the information entering our senses, but also upon which aspects of this we choose to attend’ (Driver,2001). Due to the limited capacity of humans’ sensory and information- processing systems, people cannot deal with all the information perceived. One would select to pay attention to some information over others, despite whether the selection process is conscious or not. Selective attention is the general term to describe the ability to focus on particular aspects of the environment while ignoring others (Goldstein,2007).

The discussion of selective attention has started as early as in 1910 (Hicks, cited in Edgell,1947). However, it only became a hot debatable topic until Broadbent (1958) proposed the first theoretical model (Broadbent’s filter model) to explain the mechanism behind this mental process. Broadbent’s filter model is also called the Early Selection Model due to his conception that selection happens at an early stage of processing. His theory was influential and quickly led to intensive experiments in this area. It had also inspired hypotheses in explaining the mechanism of selective attention, including Filter-Attenuation Model (Treisman,1964) and Late Selection Model (Deutsch & Deutsch,1963; Norman,1968).

The conflict between the early selection model and late selection model is the most debatable topic as they were both supported by vast amount of research despite their remarkable difference in selective locations. The essay will discuss the major theoretical conflict between these two hypotheses with experiment evidences, and also the theories which had tried to solve the conflict between them.
Early Selection Model (Broadbent’s Filter Model)
The early selection model, also known as the Broadbent’s filter model, is the first theory that attempted to explain the mechanisms of selective attention. The inspiration of this theory could be traced back to the 1950s, in which much of the early attention studies were carried out to solve air traffic controllers’ problem. In the past, controllers received messages from pilots by loudspeakers in the control tower. The mixed voices of many pilots over a single loudspeaker could make controlling task difficult. This effect was first described and called the ‘cocktail party problem’ – As the noisy controller room was like a cocktail party, in which many sounds and noices would come to ears at once. (Cherry,1953) This ‘cocktail party’ problem had unveiled two basic fundamental questions in attention. First is how did people pick up and notice sounds which are relevant to them? In other words, how could efficient selective attention take place; and whether they could receive information from unattended messages. If they could gather information from these unattended messages, how much could they obtain? Experiments, mostly based on the dichotic listening tasks, had been carried out in spite of answering the above questions. In the dichotic listening task, the listener is presented with two different spoken messages, such as a speech or a list of random words simultaneously. Different messages are directed into each ear by means of headphones and tape-recorders. Listeners were instructed to concentrate on one message and ignore the other. In order to confirm that attention would be strictly focused on one message, the listener was instructed to repeat aloud the message immediately after hearing it. This method is also known as shadowing as it requires listener to ‘shadow’ the unattended message and focus on the attended message (Cherry,1953). The speech shadowing techniques was significant for these experiments because it ensured that the subjects were attending to the desired message. By the means of dichotic listening task with shadowing, influential findings have been explored and had provided the initial answers for the above questions:
(1)