Brunel University Cognitive Psychology Lecture Notes PDF

Document Details

Uploaded by Deleted User

Brunel University London

André Szameitat

Tags

cognitive psychology attention models of attention lecture notes

Summary

These are lecture notes from a cognitive psychology course at Brunel University London. The document covers topics including attention, models of attention, automaticity, and the Stroop effect. The notes include examples, diagrams, and relevant research.

Full Transcript

Cognitive Psychology Dr André Szameitat [email protected] Office hours: Wed 11.30-12.30; Thu 1.30-2.30 Room GB263 Module PY2025 Lectures on YouTube: http://tinyurl.com/o6cqz7n or search for...

Cognitive Psychology Dr André Szameitat [email protected] Office hours: Wed 11.30-12.30; Thu 1.30-2.30 Room GB263 Module PY2025 Lectures on YouTube: http://tinyurl.com/o6cqz7n or search for “Andre Szameitat” Quick feedback questionnaire on the lecture: http://bit.ly/1dAbHo3 group: https://www.facebook.com/groups/py2025.2015/ Today’s Schedule André Szameitat Psychology Slide 2 Introduction to Attention Models of Attention Automaticity Demonstration: Stroop-Interference Introduction to Attention André Szameitat Psychology Slide 3 Introduction André Szameitat Psychology Slide 4 Definition – “Everybody knows what attention is.“ (William James in “Principles of Psychology”, 1890) – “Everybody knows what attention is. It is the taking possession of the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought. Focalisation Focalisation, concentration concentration, of consciousness are of its essence.[…] It implies withdrawal from some things in order to deal effectively with others […]” (William James in “Principles of Psychology”, 1890) Introduction André Szameitat Psychology Slide 5 Definition – Attention is the cognitive process of selectively concentrating on one aspect of the external or internal environment while ignoring other aspects. Experimental research on attention started in the 1950s Introduction André Szameitat Psychology Slide 6 Attention in Perception – If we look around us we experience an amazing richness of detail. – But we do not notice the detail of objects unless attention is directed to them. Demonstration: – Close your eyes. – What colour do your neighbours‘ jumpers have? The chairs in the lecture theatre? My shirt? Introduction André Szameitat Psychology Slide 7 Another demonstration: Change blindness – See http://www2.psych.ubc.ca/~rensink/flicker/download/index.html Introduction André Szameitat Psychology Slide 8 Another demonstration: Change blindness – Airplane Introduction André Szameitat Psychology Slide 9 Another demonstration: Change blindness – Dinner Introduction André Szameitat Psychology Slide 10 Another demonstration: Change blindness – Money Introduction André Szameitat Psychology Slide 11 Change blindness – „Flicker paradigm“ (Rensinck et al., 1997) – Even big changes to a picture take a while to notice Introduction André Szameitat Psychology Slide 12 Change blindness – Flicker paradigm – Explanation The „bottom-up“ perceptual processing is optimised towards continuous input. In continuous input, changes are easily detected (usually changes are due to motion). Motion is highly salient and pulls our attention towards it. The brief blank screen between the two pictures disrupt the basic bottom-up processing. No continuous motion can be detected. Changes have to be searched for by deliberate attentional „top-down“ control, searching location by location… Introduction André Szameitat Psychology Slide 13 Change blindness – Works outside the laboratory. – You can actually swap conversation partners! See: http://youtu.be/FWSxSQsspiQ Introduction André Szameitat Psychology Slide 14 And a final demonstration, how attention may influence our perception. – You will see two teams of players, one wearing white shirts, the other black shirts. – Your task is to count how many times the players in white shirts pass the ball among each other. – Since all players mingle all the time, you really have to pay attention. Simons & Chabris (1999) See: http://www.youtube.com/watch?v=IGQmdoK_ZfY http://viscog.beckman.illinois.edu/grafs/demos/15.html http://www.youtube.com/watch?v=vJG698U2Mvo Introduction André Szameitat Psychology Slide 15 The „invisible gorilla“ is an example of inattentional blindness – Involvement of attention in other tasks may let us miss obvious events. Introduction André Szameitat Psychology Slide 16 There are many examples in movies. Ace Ventura. Background changes. Introduction André Szameitat Psychology Slide 17 There are many examples in movies. Harry Potter. Harry changes side of table. Introduction André Szameitat Psychology Slide 18 There are many examples in movies. Speed. Sometimes the bus has ads on it and sometimes not Introduction André Szameitat Psychology Slide 19 There are many examples in movies. The usual suspects. Plane lost 2 jet engines. Introduction André Szameitat Psychology Slide 20 There are many examples in movies. Password Swordfish. Position of headrest changes. Introduction André Szameitat Psychology Slide 21 The influence of attention on perception has implications – Eye witness testimony Witnesses may miss changes in the identity of the suspect. – Driving a car in traffic Elderly are slower in change detection – Human Machine Interaction Missing of important signals E.g. radar operators Introduction – Summary André Szameitat Psychology Slide 22 Attention allows us to focus on a particular stimulus while at the same time disregarding other stimuli. Attention affects perception – Change blindness – Inattentional blindness Attention also affects many other processes, e.g. – Learning – Working Memory – Vigilance Generally, attention is a central topic in Psychology Introduction André Szameitat Psychology Slide 23 Any questions? Models of Attention André Szameitat Psychology Slide 24 Models of Attention André Szameitat Psychology Slide 25 Attention allows us to focus on a particular stimulus while at the same time disregarding other stimuli. Key research questions – How do people choose which stimuli to attend from the environment and which to ignore? – At which level is information processed when we decide to pay attention to a stimulus? Physical characteristics (tonality, loudness, brightness…)? Not only physical characteristics but also meaning? – What happens to the stimuli that are not attended? Models of Attention André Szameitat Psychology Slide 26 Cocktail Party Phenomenon – In a party, you have no problem concentrating on your conversation despite all other conversations – You also have little difficulty switching your focus from your conversation to another Models of Attention André Szameitat Psychology Slide 27 Cherry (1953) wanted to investigate the Cocktail Party phenomenon scientifically Dichotic listening paradigm – A different message is played to each ear. – When voices had the same physical properties, participants could not separate the messages using meaning only. Message 1 Message 2 Models of Attention André Szameitat Psychology Slide 28 How much is retained from the unattended message? Shadowing task – A different message is played to each ear. – Overtly repeat message played to one ear. - Overt repetition requires strong focusing of attention on the attended input. - Consequently, no attention on the ignored input. - Was anything of the ignored remembered or recognized? Models of Attention André Szameitat Psychology Slide 29 How much is retained from the unattended message? – Mostly physical properties (e.g. tonality) were noticed – Virtually no noticing of Meaning of the message Change in language (English to German) Change in gender (male to female) Message in reverse speech – It seems that unattended messages are filtered very early in the sensory processing, because virtually no „higher-level“ information is retained. Models of Attention André Szameitat Psychology Slide 30 Filter model (Broadbent, 1958) Ears Eyes [… channels for every sensory modality …] Donald Broadbent 1926 - 1993 Models of Attention André Szameitat Psychology Slide 31 Filter model (Broadbent, 1958) Which channel is attended to? – E.g. Vision, Audition, …? – Guided by attention (voluntarily, goal-directed, top- down) – May also be guided by reflexive attention, i.e. a salient stimulus automatically captures attention (involuntary, bottom-up) Models of Attention André Szameitat Psychology Slide 32 Brief detour: The concept of „Salience“ (or „saliency“) Definition: The state or quality by which an item (e.g. object or person) stands out relative to its neighbours. Salience typically arises from the contrasts between items and their neighbourhood. Models of Attention André Szameitat Psychology Slide 33 Brief detour: The concept of „Salience“ (or „saliency“) Examples AAAAAAA AAAAAAA AAAAAAA AAAAAAA AAAAAAA AAAAAAA Salience by color. Note that the same color (e.g. red) can result in a stimulus being salient (left) or not salient (right) – depending on the contrast (i.e. difference) to its neighbours. Models of Attention André Szameitat Psychology Slide 34 Filter model (Broadbent, 1958) Which channel is attended to? – Guided either top-down or bottom-up Which information is attended to? – Which conversation do we follow? – To which point in our view do we look? – Guided bottom-up by basic physical features (e.g. pitch, loudness, colour, direction,…) Models of Attention André Szameitat Psychology Slide 35 Filter model (Broadbent, 1958) The stimuli passing the filter become the focus of attention The unattended stimuli are completely disregarded It is an „all-or-nothing“ filter Models of Attention André Szameitat Psychology Slide 36 … last week I went golfing with Edward, and … Me: Edward is in conservation with Betty. Then John drops Edwards name. It is very likely that Edward will notice this and will orient his attention towards John‘s conversation. The early-selection filter model by Broadbent cannot account for this finding. Models of Attention André Szameitat Psychology Slide 37 Attenuator model (Treisman, 1964) – Is based on Broadbent‘s filter model – Different mechanism for filtering: Not by an all-or-nothing filter (Broadbent) But by a gradual working „attenuator“ Anne Treisman (with her husband, Daniel Kahneman, Psychologist and 2002 Nobel Prize Winner) Models of Attention André Szameitat Psychology Slide 38 When is an attenuated message Hierarchy of Analyzers analysed? Messages are analysed hierarchically When its activation exceeds a (simple to complex) threshold. If task demands are high, little Threshold affected by: Context and processing of attenuated message. priming; Subjective importance; If task demands are low, attenuated Degree of Attenuation messages may be fully processed. Models of Attention André Szameitat Psychology Slide 39 Attenuator model (Treisman, 1964) – Is an Early-Selection model (like Broadbent‘s model). – Allows for initially unattended information to become attended (and by this aware). – All other information is decaying in early sensory memory. – Unattended information can have no effect on behaviour. Behaviour is only affected by information we become aware of, i.e. by attended information. Models of Attention André Szameitat Psychology Slide 40 Late-selection models (Deutsch & Deutsch, 1963; Norman 1968) – All information is fully analysed for meaning. – Then, based on physical properties and meaning, selection by a filter takes place. Evaluation – Advantage: Since all information is processed, it is unlikely that important information is missed. – Disadvantage: Very resource demanding. Models of Attention André Szameitat Psychology Slide 41 Late-selection model Broadbent‘s filter model (Early selection) Treisman‘s attenuator model (Early selection) Late selection model Models of Attention André Szameitat Psychology Slide 42 Late-selection models (Deutsch & Deutsch, 1963; Norman 1968) – How to test whether they are true? Key assumption: All information is fully processed. – Including information which is unattended and unaware. – Thus, unattended and unaware information should have the power to influence behaviour (which is impossible according to all early-selection models) Models of Attention André Szameitat Psychology Slide 43 How can one test whether unaware information affects behaviour? James Vicary (1957): Presented subliminal messages in a movie for 0.3 ms (every 5 s) He reported that sales of Coke increased by 18%, of Popcorn by 58% Although the results are most probably faked, they initiated the discussion on subliminal persuasion [subliminal = below threshold of perception] Models of Attention André Szameitat Psychology Slide 44 How can one test whether unaware information affects behaviour? – Subliminal priming (Dehaene et al., 1998) The masks make Nevertheless, the prime Target affects response times, i.e. the prime Mask 2 behavior. invisible. Prime The prime is Mask 1 unaware The fact that the prime information. affects behaviour irrespective of notation shows: It is no perceptual priming. A – Alphanumeric (“6”) Its semantic meaning V – Verbal (“SIX”) has been processed. Models of Attention André Szameitat Psychology Slide 45 Subliminal priming works well – Many variations have confirmed subliminal priming James Vicary’s “Coca Cola”-ads did not work What is the difference? – In subliminal priming paradigms (like Dehaene’s) participants prepared themselves to respond as quickly as possible to a number, and the prime was a number as well. the influenced behaviour (i.e., the response) occurred very soon after the prime. Models of Attention André Szameitat Psychology Slide 46 Subliminal priming What is the difference? If the prime does not relate to the current behavioural goals, it is ineffective. Left Drink Coke In these examples, there would most likely be no effect on behaviour at all. Models of Attention André Szameitat Psychology Slide 47 There is evidence for early selection models – Information of unattended channels is not really retained There is evidence for late selection models – Subliminal stimuli are processed for their meaning What model is correct? – Maybe both? – Maybe sometimes selection is early, and sometimes late? – But what determines whether selection is early or late? Models of Attention André Szameitat Psychology Slide 48 Theory of perceptual load (Lavie; 1995, 2000) – „load theory“ Difficult tasks – E.g., find a specific DVD among other DVDs Neely Lavie UCL London – Require a lot of attentional resources – Selection is early to make resources available Easy tasks – E.g., find a DVD among books – Require only little attentional resources – Selection is late since resources are available Models of Attention André Szameitat Psychology Slide 49 Two choice response task. Target letters: “x” – left finger “z” – right finger Distractor letters: “X”, “Z”, or any other letter. Task irrelevant. Is to be ignored. Distractor is processed (Difficult task) Low load: (and, thus, disctracts) (Early selection) Only one potential only in the easy task. target letter. Easy task. High load: (Easy task) Find target letter (Late selection) among 5 other letters. Difficult task. Models of Attention – Summary André Szameitat Psychology Slide 50 Early selection model (Broadbent, 1958) – Unattended information is filtered out completely early in processing Attenuator model (Treisman, 1964) Unattended information is attenuated early in processing Late-selection model (Deutsch & Deutsch, 1963) – All information is processed, it is filtered out only late in processing Theory of perceptual load (Lavie, 1995) – Selection is early in difficult task, and late in easy tasks Models of Attention André Szameitat Psychology Slide 51 Any questions? 5 Minutes Break André Szameitat Psychology Slide 52 00:00 00:30 01:00 01:30 02:00 02:30 03:00 03:30 04:00 04:30 05:00 Automaticity André Szameitat Psychology Slide 53 Automaticity André Szameitat Psychology Slide 54 Controlled processing Automatic processing Limited capacity Unlimited capacity Requires attention Does not require attention Rather slow Rather fast Effortful Effortless Is aware, conscious Outside awareness Controllable Uncontrollable Flexible: Can be used Inflexible: Cannot be easily flexibly in changing changed to changing circumstances circumstances Some theories classify processing strictly as either controlled or automatic, while other theories regard it as a gradual transition between the two. Automaticity André Szameitat Psychology Slide 55 Controlled or Automatic? Walking Automatic Walking Controlled Automaticity André Szameitat Psychology Slide 56 Controlled or Automatic? Driving Driving lessons Automatic Controlled Automaticity André Szameitat Psychology Slide 57 Schneider & Shiffrin, 1977 Task: Does the A B + X R display set contain any stimulus from C D B F the Memory Set? Memory Set Fixation Display Set Automaticity André Szameitat Psychology Slide 58 Schneider & Shiffrin, 1977 Consistent mapping: A B + 1 5 Only letters in Memory Set, only C D B 7 numbers as distractors in Display Set Memory Set Fixation Display Set Target detection due to automatic differentiation between letters and numbers. Varied mapping: 3 B + 1 F Letters and numbers C 9 B 7 can be in both sets Memory Set Fixation Display Set Here, search requires a controlled comparison of each item of the Display Set with each item of the Memory Set. Automaticity André Szameitat Psychology Slide 59 Schneider & Shiffrin, 1977 Controlled processing Automatic processing Automaticity André Szameitat Psychology Slide 60 How does an activity become automatic? – Practice! – But how? Logan‘s (1988) instance theory of automaticity – Unpracticed: General algorithm used to solve problems (controlled process) – Practiced: Memory retrieval of past solutions (automatic process) Automaticity André Szameitat Psychology Slide 61 Logan‘s (1988) instance theory of automaticity Example: Multiplication 5 x 4 – Initially: Sum the „5“ four times (5+5 = 10; +5 = 15; +5 = 20). This is slow and effortful – a controlled process. – After practice, we have learned that 5 x 4 = 20. When we encounter this problem, we can simply retrieve this solution from memory. This is fast and effortless – an automatic process. Automaticity André Szameitat Psychology Slide 62 Logan‘s (1988) instance theory of automaticity – „Novice performance is limited by lack of knowledge and not lack of resources…“ – „Only the knowledge base changes with practice.“ – The more knowledge is available to solve a task, the less attentional resources are required. Automaticity André Szameitat Psychology Slide 63 Automatic processes are inflexible – Shiffrin and Schneider (1977) – Trained a consistent mapping: Letters B – L always Memory Set Letters Q – Z always distractors in Display Set After 2000 trials E B + R T training the two sets of letters (B-L and Q-Z) C G B Q were automatically separated (just like Memory Set Fixation Display Set letters and numbers in the previous experiment). Automaticity André Szameitat Psychology Slide 64 Automatic processes are inflexible – Shiffrin and Schneider (1977) – After training, they swapped the sets: Letters Q – Z always Memory Set Letters B – L always distractors in Display Set It took more than R T + B G 1000 trials to “undo” the automatization. Z Q R K Only then performance returned Memory Set Fixation Display Set to levels at the start of the experiment. Automaticity André Szameitat Psychology Slide 65 Dangers of Automaticity Tourists in the UK are in danger “Highway hypnosis”: People of missing cars because they look drive without any awareness. in the wrong direction. Automaticity – Summary André Szameitat Psychology Slide 66 Automatic processing is effortless, fast, and outside of awareness. Thus, automaticity greatly reduces demands on cognitive control and attention – Suppose you would need to explicitly attend to every action, like walking, driving, drinking/eating, writing,… The most profound disadvantage of automatic processes is their inflexibility. Automaticity can be established by practice. Automaticity André Szameitat Psychology Slide 67 Any questions? Demonstration André Szameitat Psychology Slide 68 Stroop André Szameitat Psychology Slide 69 Stroop task Red Green Blue Task 1: Name Color Task 2: Name Color of of patch ink of the word Finding: Naming colour of word ink is much slower than naming colour of patches. „Stroop interference“ John Ridley Stroop circa 1933 Stroop André Szameitat Psychology Slide 70 Stroop interference Red Green Blue – Reading is automatic. – Thus, incompatible word information interferes with colour naming task. – Controlled processes are required to override this interference caused by automaticity. Support – Slow readers have less Stroop interference. – Colour-words in second language: Less Stroop interference. Stroop André Szameitat Psychology Slide 71 Demonstration – Please team up with your neighbour. – One person reads aloud the colours. – The other persons times it (e.g. stopwatch on mobile) – Note the times and see which list takes longer. – Then swap roles (the one who timed reads now, and the other takes the time) Stroop – Summary André Szameitat Psychology Slide 72 Stroop interference – Demonstrates the power of automaticity – Demonstrates that we nevertheless can overcome such strong automatic tendencies! We actually can do the task, we just get slower. Automatic responses can be overcome by Executive functions; Executive control processes E.g. Supervisory Attentional System (Norman & Shallice, 1986). This will be covered in a later lecture… Stroop André Szameitat Psychology Slide 73 Any questions? Next week: Short Term Memory Lectures on YouTube: http://tinyurl.com/o6cqz7n or search for “Andre Szameitat” Quick feedback questionnaire on the lecture: http://bit.ly/1dAbHo3 group: https://www.facebook.com/groups/py2025.2015/ Dr André Szameitat [email protected] Office hours: Wed 11.30-12.30; Thu 1.30-2.30 Room GB263 Module PY2025

Use Quizgecko on...
Browser
Browser