How was it possible for those individuals to deal with all the incoming information and decide on information that was relevant to them? Broadbent propose that there needed to be a clear difference between inputs for selection to take place. Subsequently this led Broadbent to design a dichotic listening experiment to investigate the processes involved in switching attention.
In an artificial manipulation of multiple sounds, participants were tested on how they responded to sounds being played simultaneously in the ear via head phones. In this instant each ear would receive a different message, however when participants were asked to report the content of the message in the unattended ear, this was met with great difficulty and almost nothing of the message could be recalled, Broadbent (1954, as cited in Naish, 2010).
In his explanation of these findings, Broadbent (1954, as cited in Naish, 2010) proposed that a bottleneck exist within the attentional system, consequently this is produced when large amounts of information enters the brain. Filtration of that information occurs based on its relevance before it is processed in the conscious part of the brain. Furthermore according to this model the brain has a certain criteria for filtering information and this is based on the physical properties of the sensory information, for instance voice (who is speaking), location (which ear is the information being spoken in). According to Broadbent the meaning of the messages is not taken into account by the filter until after the message has been selected. Furthermore since there is only a limited capacity for information processing the un-selected inputs remain briefly in the sensory buffer store, failure to undergo processing results in the rapid decay of these stimuli. Broadbent assumed that the filter rejected the non-shadowed or unattended message at an early stage in processing.