Attention refers to the process of cognition by which the human brain decides on the information that is significant enough to attend to and which should be ignored, William James (1890, cited in Driver, 2001). Since the 1950s' a number of research has been carried out into the field of attention, for many; a lot of interest has been centred around understanding the mechanism within the brain that enables us to register, attend to and ignore inputs from our external environment. Most of the early researchers attempting to understand the process have generated assumptions that the brain is similar to a computer processor, which meant that the information that humans could attend to is quite limited, so a process of selection takes place. However one central and ongoing debate has been whether selection of information takes place at an early or late stage before processing. But before launching into a detail analysis of the varying theoretical perspectives, perhaps it will be important to note from the outset that both early and late selective models of attention originally spans from research into auditory attention. In attempting to understand how the brain is able to process the vast amount of information, researchers used a psychological test referred to as the dichotic listening task to investigate selective attention. This essay would therefore seek to compare and contrast; Broadbent (1958), Treisman (1964) and Deutsch & Deutsch (1963) which stands as some of the most influential theories of selective attention. It will consider how well these models explain the mental processes involve in auditory attention and further aid our understandings of how we can selectively attend to information within our environments. .
Broadbent (1954, cited in Naish, 2010) was one of the first researchers to investigate auditory attention using dichotic listening experiments. Following the Second World War, Broadbent attempted to understand the problems encountered by Radar operators when trying to communicate with different pilots over a single system.