LECTURE: JANUARY 31ST, 2013
TOPIC: CHAPTER 4ATTENTION
◦ Process of concentrating on specific features of the environment or certain thoughts or
➔ Information Processing Model
◦ Attention is the next step after perception.
◦ Perceive something and then decide where to store what you perceive.
◦ Focus on one specific object
◦ Focus on text
◦ Focus on more than one thing
◦ Focus on text and TV
◦ Physically move eyes to look at something
◦ Not physically moving eyes to look but can see in peripheral what is going on.
◦ Ability to focus on one message and ignore all others
◦ Filtering out some info and promoting other info for further processing
◦ James (1959): attention and perception
▪ Auditory Stimuli
▪ Can only attend to a fraction of what is in our environment or else if you tried to pay
attention to everything, one would get overwhelmed.
▪ Therefore, we only attend to aspects that are important or meaningful to us.
▪ Everyone has a certain threshold for how much one can pay attention to. It is pretty
limited for most people.
▪ Selectively attending to something has to be timed; a planned out process.
▪ Can be considered overt and covert attention.
▪ Early researches such as William James
• Selective attention helps us focus on what we need to attend to and helps us refrain
from what we do not need to attend to.
• Weakness: we are able to selectively attend to what we want to is what James says
but we also perceive things that we are not attending to as well.
➔ Research Method: Dichotic Listening Task (no demonstration)
◦ Cherry was the first to do this
◦ One message is presented to the left ear and a different message is being sent to the right
◦ Participant is asked to repeat the sentence that he/she is suppose to be attending to. In the
picture on the power point the person is asked to repeat the left ear message. Repeating this
message is called shadowing.
◦ Unattended ear → selective attention would suggest that the person would not process it or
understand the content. ◦ Can we filter out the info in one ear and only attend and hear the message one is asked to
➔ Research Method: Dichotic Listening Task
▪ Participants could not report the content of the message in unattended ear BUT:
• Knew that there was a message
• Knew that gender of the speaker
▪ However unattended ear is being processed at some level
• Change in gender is noticed
• Change to a tone is noticed
➔ Models of SelectiveAttention
◦ Early Selection Theory Model
▪ Broadbent's filter model
• Attention filter happens early on.
◦ Intermediate Selection Model
▪ Tresiman's attenuation theory
• Attention filter happens in the middle.
◦ Late Selection Model
▪ E.g. McKay (1973)
• Attention filter happens later
◦ Attention filter, filters what info is important.
◦ These three theories differ when this filter process happens.
◦ Codes for both the physical characteristics as well as the content of the message.
➔ **ON EXAM**
➔ Broadbent's Filter Model
◦ FIGURE 4.3 P. 84 and FIGURE 4.2
◦ Early Selection Model
◦ Sensory Memory
▪ Holds all incoming information for a fraction of a second
▪ Transfers all info to next stage
• Restricts the amount and flow through the filter.
• Messages come through sensory memory and then info is transferred to next stage
which is the filter. The filter focuses on physical aspects and message.
• Dichotic Listening Task → only attended message goes through filter.
• Detector → process it and determine what other higher level characteristics of the
message are involved. Short term memory then receives the output of detector and
holds onto the info for 10-15 seconds deciding whether to transfer it to long term
memory or discard it.
• Info in unattended message should not be accessible to our unconscious at all.
• Filter can detect physical characteristics and then decides to send through attended
• This model goes against the dichotic listening task. Seen as a great theory to explain
some attention abilities with certain types of tasks and stimuli but not all selections
▪ Participant's report hearing own name • This according to others should not even enter our consciousness.
• Moray (participants were to shadow the message in one ear and ignore the message
in the other. Many still detected hearing their name in the unattended ear. This goes
against Broadbent and his idea that the filter is to only let through one message
based on physical characteristics. Instead, the name was analyzed enough to detect
• Cocktail Phenomenon
◦ At a party, you are hanging out and then you hear your name and so you
disengage from your original conversation to attend to that stimuli.
◦ This goes against Broadbent because if you are not attending to something you
should not even be perceiving it.
▪ Participants can shadow
▪ Wedderburn (1960)
• FIGURE 4.4 ON P. 85
• Attended ear: Dear Seven Jane
• Unattended ear: NineAunt Six
• Info moves from left to right so your left ear hears Dear, your right ear hearsAunt
and your left ear hears Jane instead of Dear Seven Jane.
• People heard Dear Aunt Jane instead. People switch their attention from one ear to
another.Also, because it is a more meaningful message to talk about an aunt than a
number. Example: DearAunt Jane or Dear Nine Jane.
▪ Effects of practice
• With practice people can start to attend to both messages by repeating it and such.
➔ Tresiman's Attenuation Theory
◦ FIGURE 4.5 P. 86 and 4.6
◦ Intermediate-selection model
▪ Attended message can be separated from unattached message early in the info-
▪ Selection can also occur later
▪ 'Leaky-filter model' → leaks out certain aspects of UNATTENDED MESSAGE
• Attenuator analyzes 3 things for incoming messages: physical characteristics (voice
such as tone, volume), language (how the message groups into syllables), and the
meaning (how the phrases are grouped into something meaningful).
• In terms of the dichotic listening task, once the attended and unattended message
have been identified, both messages are let through the filter or attenuator.
Difference: it determines the strength of the message let through. So the attended
message is let through with great strength and the unattended message still gets
through but at a much weaker level.
• After the filter it the message goes to the dictionary unit which contains that have
thresholds for being activated. Common words have lower thresholds and
uncommon words have higher thresholds. Example: name → important word an low
threshold. Example: Rutabaga → high threshold because it is uncommon. Example:
Boat → is in the middle for a threshold because it is somewhat familiar but not as
attention grabbing as one's own name being called in a grocery store. BUT..if you
were a cook and used rutabaga all the time you may form a lower threshold
• Top-down processing * • Broadbent → bottom-up processing*
➔ Late Selection Model
◦ Selection of stimuli does not occur until after info has been analyzed for meaning
◦ McKay (1973)
▪ In attending ear, participants heard ambiguous sentences (Ex. They were throwing
stones at the bank).
▪ In unattended ear, participants heard either
▪ Choose which was closest to the meaning of attended to message
▪ The meaning of the biasing word affected participant's choice
▪ Participants were unaware of the presentation of the biasing words
• Participants not really aware of biasing words (biased words in addition to
ambiguous sentence → affected how they formed the meaning of the sentence).
• Not attending stimuli until you form a meaning.
• Meaning affected their judgment of the sentence.
• Late selection → late filter
• Incoming info is processed through meaning and then goes through selection
process. Select info after it has been processed for meaning.
➔ Factors That Impact SelectiveAttention
◦ FIGURE 4.7 P. 87
◦ Cognitive Resources
▪ Our capacity
◦ Cognitive Load
▪ High load → less resources left after
▪ Low load → more resources
◦ Task-irrelvant Stimuli
▪ When powerful, it affects your ability to selectively attend because the task-irrelvant
stimuli is so strong
➔ Factors That Impact SelectiveAttention
◦ FIGURE 4.8 P. 88
◦ Cognitive Resources → capacity to load info
▪ Flanker-Compatability Test
• Flankers → distractors
• Press z if Aor B is the target and M if C or D is presented.
• (a) The flankers are the B's. Response to flanker and target are the exact same →
• (b) Incompatible → response for flanker and target are different
• (c) Neutral → got no instructions on what to do if you see a letter other than what
◦ Compatible → fastest response
◦ Incompatible → slower response because CAC → C andAhave different
responses in terms of pressing a key.
◦ Neutral → intermediate response to target
◦ “B” took up most resources. ◦ Can participants focus their attention on detecting the target so that the identity
of the distractor will not affect their performance?
◦ Cognitive Load
◦ FIGURE 4.9 P. 89
◦ (how much info you are getting presented and how much you can put based on the load one
has). Exam → high cognitive load. Higher the load less resources and vice versa.
▪ Press a certain key when X was a target and a certain key when N was a target
▪ The 0's are considered the load
▪ N was the flanker. People told to ignore it.
▪ Low Load condition → ignoring N, target is X → easier to see X on low load versus the
▪ Reaction time is longer for incompatible distractors.
▪ Low Load condition → incompatible flanker (response for X and N different. Low load
still helped the process → can see X easily).
▪ Extra cognitive re