Attention

• Focused attention, also known as selective attention, involves attending to one stimulus in the presence of one or more others.
• Divided attention involves attending to more than one stimulus at the same time.
• Cherry (1953) found that a number of factors were important in managing to listen to one person in the presence of many people talking: the sex of the speaker, their location, and their intensity.
• Cherry found it very difficult to use meaning as a way of attending to a single speaker.
• Segregation by pitch is a phenomenon where people tend to group together sounds of similar pitch.
• Cherry (1953) also carried out studies in which participants had to repeat what they heard in one of their ears, while another message was played to the other ear, the dichotic listening task.
• Cherry found that there was very little processing of higher-level information (e.g., meaning, language) coming from unattended sources and very little memory.
• Gross changes in physical attributes of the unattended stimulus, like location, type of stimulus (speech vs music), loudness, etc., could be reported.
• Broadbent (1958) used a dichotic listening task where participants heard three digits in one ear interleaved with three digits in the other ear. Left-ear 4 9 2 Right-ear 7 6 8.
• When participants recalled the digits, they reported them by ear, starting with the digits from one ear.
• Broadbent's filter theory of attention suggests that two stimuli presented simultaneously both access a sensory buffer, one of the stimuli is allowed through on the basis of physical characteristics while the other is kept in the store for later processing, and the filter prevents overloading of subsequent stages.
• A schematic of Broadbent's filter theory includes a sensory register, a selective filter, a short-term memory, and further processing.
• Broadbent's theory suggests a block on the unattended stimulus, and the extent of the block depends on the similarity between the two stimuli.
• Allport, Antonis & Reynolds (1972) found that if both stimuli are verbal, the degree of processing of the unattended stimulus, as assessed using a memory measure, is low.
• Attending to a verbal stimulus does not block memory for simultaneously presented pictures.
• Von Wright, Anderson & Stenman (1975) associated a word with an electric shock, such as the word “dog”, and then presented that word in an unattended channel, like the ear.
• All conditions produced some increased skin response suggesting that people had accessed the meaning of the unattended stimulus.
• The response to these related words in the unattended channel was less than if they were presented in the attended channel.
• Lachter et al. (2004) suggested that there may be slippage (quick attention shifts to the unattended channel) rather than leakage from that channel as had previously been assumed.
• There is evidence for both early selection and late selection, with some studies suggesting that unattended stimuli can be processed up to the level of meaning without participants being aware of their presence.
• Treisman (1960) and Gray & Wedderburn (1960) found that a word from an unattended stimulus could intrude into participants' reports if it was a good fit with the attended stimulus.
• Participants with low working memory capacity were more likely to hear their own name in an unattended channel compared to those with high working memory capacity (Conway, Cowan & Bunting, 2001).
• Deutsch & Deutsch (1963) suggested that all stimuli are fully analyzed, with only the most important stimulus controlling the response.
• Moore & Egeth (1997) found that participants were influenced by an unattended pattern in their judgment of line length, even though they claimed not to have noticed it.
• Treisman (1964) proposed that the attentional filter attenuates the processing of unattended stimuli without blocking it entirely, and that unattended information can be processed up to a level depending on attentional demand.
• Stimuli that are particularly meaningful, like one's own name, can catch attention even when they are in unattended stimuli (Moray, 1959).
• Rees, Russell, Frith & Driver (1999) found that attended letter strings gave different activity patterns on an fMRI scan depending on whether they formed real words or not, suggesting earlier selection.
• The perceptual load framework, as proposed by Lavie & Tsal (1994), assumes that attention is limited in capacity, the amount of attention applied to any one task is related to its perceptual demands, one cannot allocate less than the total amount of capacity at any time, and surplus attentional capacity "spills over" onto what would otherwise be unattended stimuli.
• The perceptual load framework suggests that results will indicate early selection when the attended task is perceptually demanding, no surplus attention spills over to unattended material, and late selection when the attended task isn't so demanding, with surplus attention spilling over to unattended material.
• Lavie (1995) carried out an experiment in which participants had to respond to a letter in one of six positions that was either an x or a z, there were either high-load or low-load letters in the other five positions, and a large distractor letter was presented that was either incompatible or neutral.
• The distracting effect of the incompatible letter was only evident in the low load condition – this was because some attention had spilled over to the distracting letter.
• Visual attention can be focussed without eye movement and, like a spotlight, it makes unattended locations look “darker”.
• Visual attention can change the size of the attended area, like a zoom lens.
• La Berge (1983) found some evidence supporting the view that visual attention can function like a zoom lens.
• La Berge (1983) results show that participants responded equally fast to the probe whatever its position when the word was being categorized (wide zoom) but fast to the middle position and slower to the other positions, when the middle letter was being categorized (narrow zoom).
• Awh & Pashler (2000) found that participants were slow to report digits when one appeared between two cued squares, even though it should be in the zoom area.
• Juola, Bowhuis, Cooper & Warner (1991) presented a target letter for identification in one of three concentric rings, and were given a cue which was pretty good at indicating which ring the target would be in.
• Participants did best when the target was in the ring indicated by the cue.
• Object-based attention has often been demonstrated with people with a disorder called unilateral neglect.
• Neglect patients appear to ignore all stimuli on the left side of space, eating from the right side of their plate, drawing only the right half of objects, and reading only the right-hand part of words.