Emotion and bias

?
  • Created by: freya_bc
  • Created on: 09-01-17 15:30

Types of cognitive bias

BIASES- systematic distortions in processing of info and world about us 

ATTENTIONAL BIAS- sleection attention to emotionally relared stim presented at same time as neutral ones e.g. Stroop task 

INTERPRETATIVE BIAS- tendency to interpret situ/ambiguous stim in negative/threatening way

1 of 12

The Emotional Stroop

Normal stroop- shown names of colours in congruent/incongruent ink and asked to report colour of ink, slower on incongruent trials 
Emotional stroop- shown both emotional and neutral words in different coloured inks. name colour 

Anxiety-related attentional bias
Normal people with high/low trait anx (not clinical)
Stroop task examines attention to word meanings by looking at how word meaning interferes with colour naming- emotional vs neutral words
Measure trait anx- Spielberger State-Trait Anx Inventory (STAI) 

R: high trait anx show larger interference effects on emotional stroop than low trait anx

C: Emotional meaning of word captures attention away from rel stim attribute (ink colour) Slowing effect occurs for emotional words even when words all in black ink 

2 of 12

Van Honk et al., (2001)

Instead of emotional words a variant of task uses coloured faces expressing emotion

P: present angry and neutral coloured faces (red, green, blue, yellow), ignore face itself and name colour of face 

R: colour naming latencies slower for angry faces. Angry faces capture attention away from primary task (colour naming) so slower
Happens even when present expression briefly and mask it with neutral face 

C: studies like this suggest an attentional bias towards angry faces 

3 of 12

MacLeod, Matthews, Tata (1986)

Emotional (threat-related) and neutral info presented side by side to anx patients (GAD) and control (no anx problems) 

See where attention is deployed early on in sequence of events. Location of threat info controlled- some trials have no dot (catch trials)
Press space bar as soon as you see the dot 

Examine speed of responses when dot occupies location previously occupied by neutral vs emotional stim 

R: POSITIVE BIAS (faster for neutral compared to threatening) HOWEVER reverse is true for anxious pateints, patinets slower for neutral words bc threatening stim capture and hold attention- prevent attention disengagement
Anxious patients allocate attention to threat words 

C: attentional bias for threat 

4 of 12

MacLeod and Matthews (1988)

Tested high and low trait anxious groups of studnets using DP 1 and 12 weeks before important exam. Manipulated state anx- more likely to be highly anxious in exam time 

R: 1 week before: only high trait anxious students showed bias towards threat-relared stimuli 12 weeks before: neither group showed any bias 

C: high trait anxious students only show bias when they are stressed (both high trait LT/dispositional) and high state (temp)

INTERPRETATIVE BIAS- Eysenck, MacLeod, Matthews (1987)
Homonym task- present words auditorily, homophones- pain/pane- semantic interpretation, difficult/ambiguous, pp write down words 
Task on high/low trait anx pps. High trait anx > more threat related spellings e.g. die 

x results suggest anxious people show interpretative bias? response bias? Maybe both interpretations available but high anxiety pp chose to only write down negative one 

5 of 12

Richards and French (1992)

Homographs- dif meaning but look same, instead (dual meanings/same orthography) e.g. batter, punch, stalk 

PRIMING LEXICAL DECISON TASK  + > prime word > target word/non-word > response
If prime and word are relared in meaning, faster response (typical lexical priming)

Batter-Assault should be faster/stronger priming than batter-pancake if threat-related interpretation is made for the homograph prime 

R: greater priming effect for target words related in meaning to the negative interpreation of the grime for high anxiety pp 

C: interpretative bias hsown by high anx pp/interpretation related info that is biased in high trait anx people 

6 of 12

Detecting emotional expressions

Faces shown to have some priority for detection compared with other non-face objects but does the expression displayed influence attention deployment?

BASIC VISUAL SEARCH PARADIGM- cluttered scenes with distractors and targets e.g. where's wally. Task to say whether target is absent or present DV: RT/ accuracy, IV: vary set size of display, properties of target, vary distractors/relationship between target/distractors, serial self-terminating search 

TARGET POP-OUT- detection times for target do not vary as function of set size then search is assumed to be parallel= target pop-out= automatic process not requiring attention, responses are quick.

RTs vary as function of set size then search= serial, self-terminates once target found=controlled process requiring attention 

Target not present, search each item until we reach last item and then respond "no"
CONJUNCTION SEARCH- target= orange square, some orange things arent squares and some of the squares arent orange. If display size doesnt alter RT then parallel search automatic (slope less than 10ms) if more than= serial controlled search 

7 of 12

Detection of threatening faces

OHMAN (1999) 
Evolutionary adaptive to detect threatening faces quickly/automatically
Visual threat-relared srim e.g. angry faces, should be detected faster than non-threatening
Is attention captured by some emotional expressions more than others? Do angry faces show pop out effects in crowds of distractor faces?

FACES IN THE CROWD EFFECT

If detection of threat is fast and automatic, we might expect threatening objects to be detected rapidly in cluttered scenes 

8 of 12

Hansen and Hansen (1988) Exp 1

3x3 grid, 9 DIFFERENT faces

1/2 trials all same emotion, 1/2 had one discrepant emotion. 
Faces show angry/happy/netural. Pps respond 'same' or 'different'

Same condition: all faces are angry, happy or neutral
Different condition: one discrepant face in crowd
Examined RT/error rates for face detection 

R: Pattern of attention suggests there may be an ANGER SUPERIORITY EFFECT- spot angry face more easily than neutral/happy one 

x difficult to explain how finding neutral face in happy crowd is relatively face 

9 of 12

Hansen and Hansen (1988) Exp 2

Presented crowd of 4 faces all containing a discrepant face briefly then masked with scrambled letters 

Angry face in happy crowd
Happy face in angry crowd 

All faces were same idv, discrepant face was in one of the four quadrant locations, pp had to say where the discrepant face was 

R: pp took less time to locate angry face in happy distractors than vice versa

x BUT these experiments don't show whether certain emotional expressions 'pop-out' of crowds as the search sets within an exp were always the same 

10 of 12

Hansen and Hansen (1988) Exp 3

Vary number of faces in set size- 4 or 9

pp search a 2x2 or 3x3 matric of different faces

Faces with angry/happy/neutral expressions as before, pps respond 'same' or 'different'

R: angry faces detected relatively rapidly in happy and neutral crowds 
Search times for detecting an angry face in a crowd of happy distractors not greatly influenced by set sie cimpared with searching for happy face in angry crowd 
Suggests 'pop-out'
Search for happy face in angry crowd shows a set size effect 

INTERPRETING THE 3 EXPERIMENTS
shows angry/threatening faces 'pop-out' in crowds of happy/neutral faces

x not really clear what reasons are for advantage at detecting the angry face

Angry face could capture attention or it could be that searching through the happy distractor faces is faster than searching through angry/neutral ones

11 of 12

Purcell et al., (1996)

Couldn't replicate Hansen and Hansen's (1988) findings 

Suggested that detection of angry faces was result of visual artefact of stim

Dark patches on the angry faces, confounding the interpreation of data 

Nothing to do with emotion 

12 of 12

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Emotion and bias- interpretative bias etc resources »