Cultural 2

Kelly et al (2007)

3, 6 and 9 month old Caucasian babies. Showed them african, asian, middle eastern and caucasian faces. Between subjects - each baby only viewed one race of face. Habituation method (when the baby gets bored of looking at the face and only looks at it 50% of the time). baby is then presented with two faces, one they are habituated to and one novel; they should in theory look longer at the novel face because its new. 3 month old babies all looked significantly more at the novel faces than the ones they had seen before. 6 month old babies looked at the novel image more in chinese and caucasian faces but not african or middle eastern faces. 9 month old babies only looked at the novel stimuli more in caucasian faces.

Support for universality of emotions in free labelling studies:

Boucher & Carlson (1980)

Li et al (1998)

Compared white basketball 'fans' and 'novices' Recognition of black faces Basketball fans were better at recognising black faces Presumed due to greater exposure (majority of pro basketball players are black)

Interracial contact

Contact and exposure to other races is limited which affects recognition.

Display rules

Culture dictates how appropriate it is to express emotions in particular contexts - these are learned in development. They include amplification, deamplification, masking and neutralisation. If this is the case we would predict an in-group advantage for emotion identification.

nature nurture - expressing emotions

Current consensus is that there are definite biological underpinnings but that social factors do still play a role.

Blais et al (2008)

Eye tracking on white western and east asian RPs looking at white and asian faces. Western RPs better at recognising white faces, asian RPs better at recognising asian faces. They found fixations on features of the faces were very similar in both eastern and western participants

What is an explanation for why Blais et al (2008) found what they did?

Fixations doesn't necessarily link to the way we process the face - we might be looking at the same thing but 'seeing' different things. Also westerners tend to view things locally whereas Asians view things globally. Also in Asian cultures direct eye contact is rude therefore this could result in gaze avoidance. Also, the eye movements used by different cultures are suited to recognising faces of their own races. (e.g. in Tan et al (2012) This is contradicted by Jack et al (2009).

Other race effect

Hancock and Rhodes (2010) accuracy is tested on the ability to discriminate between old and new faces. The task at encoding usually varies, people often get asked to rate attractiveness. The memory test is often unexpected.

Engelmann & Pogosyan (2013)

How can ingroup advantage develop? Representational model: Culture dictates behavioural practise, which is affected by both attentional biases and mental representations (which interact with each other), both of which lead to cognitive style which determines the interpretation of the affect.

What are emojis evidence of?

In japenese emojis, most of the emotive detail is in the eyes whereas in western emojis the detail is in the mouth.

Relative homogeneity effect

Ingroup/outgroup effect - its more cognitively efficient to focus on things more relative to you, therefore perhaps people just dont attend enough to faces of other races (the 'outgroup')

Which study contradicts ideas from Blais et al 2008?

Jack et al (2009) suggests that east asians spend more time looking at the eyes when trying to determine emotion therefore eye contact cant be rude.

Attentional Bias and Ingroup Advantage in Emotion Recognition

Jack et al (2009) white and east asian faces and participants, viewed eye movements when viewing faces showing 4 emotions; surprise, fear, disgust and anger. White participants spent time looking at all facial features (eyes nose and mouth) whereas eastern participants spent most of their time looking at the eyes (and a little the nose). Perhaps this is due to adaptations to identify their own cultures emotions best?

Tan et al (2012)

Looked at white, asian and african faces. Accuracy was good for the asian and white faces, but not so good for african faces (due to the contact effect). In terms of eye tracking, a lot of focus was on the nose but also the eyes; this is evidence that they have adapted to process the faces they encounter in everyday life (white and asian - white = eyes and asian = nose)

Goldstein 1979

Measured whether different races actually do look more similar than others.

Thibault, Bourgeois & Hess (2006)

People who 'identify' with cats were better at identifying the cats emotion

Ekman, Sorenson & friesen (1969)

Presented RPs with faces expressing one of 6 emotions (posed by actors and selected on basis of theoretically derived facial movement plans (FACS). Used people from USA and south america and japan. Found similar levels of agreement for emotions presented in every country - suggesting emotion is universal.

Safdar et al (2009)

RPs from canada, usa and japan. asked how appropriate it is to express certain emotions. all 3 said its appropriate to express happiness and surprise (canadians the most). all 3 said its less appropriate to express negative emotions e.g. fear, disgust.

Evidence for the configural-featural hypothesis

Rhodes et al (1989) & Blais et al (2008)

Physiognomic homogeneity

Some races really do have faces that are more similar to each other. Studies that have tried to directly measure this have found no differences between cultures (Goldstein 1979). The time it takes to make judgement on a face (is it familiar or new) doesn't differ across race of face or perceiver (Goldstein & Chance 1976)

Yuki et al (2007)

Studied emojis where the eyes and mouths emotions were incongruent (e.g. happy eyes, sad mouth) and asked RPs to say how happy they think the emoji was from 0-9. USA and japanese RPs. Japanese based their ratings on the emojis eyes, USA based it on the emojis mouths.

Goldstein & Chance 1976

The time it takes to make judgement on a face (is it familiar or new) doesn't differ across race of face or perceiver

What causes ORE?

Two theories are the configural feature hypothesis and the relative homogeneity effect.

Wright et al (2003)

Use black and white south african participants, and white british participants. Both white groups showed the 'other race' effect. The black RPs found no other race effect. This could be explained because in south africa a lot of the population is white therefore everyone there is used to seeing differences. They found that people who reported more contact with white people were better at recognising white faces.

Ellis & Shepherd (1975)

Verbal descriptions of own and other race; caucasians describe upper features more (e.g. eyes hair) and africans described lower features (eg mouth and jaw) again suggesting cultures look at faces differently.

Ackerman (2006)

White participants. Manipulated two things; the race of the face and the emotion presented. (anger = more attention paid to the face = predicted better recognition). Found that white RPs recognised the white faces better when neutral, but when the faces were angry they were better at recognising the black faces.

What are some real world applications of research into culture and facial recognition?

Worldwide communication, and eyewitness testimony.

Matsumoto 1990

american participants tend to rate expressions as more intense than japanese participants. however there is some ambiguity over what participants are actually rating; external display (how intense the expression is portrayed) or internal experience (how intensely the person is feeling the emotion)

In-group advantage

can we judge the emotions of our own group better than others? Albas et al (1976) and Markham & Wang (1996) suggest we can, however Boucher & Carlson (1980) suggest we cant

Hearn (1984)

evidence of variability in the way different cultures portray certain emotions

Darwin

expressing emotions is an evolutionary trait which suggests they are universal.

Hills and Pake (2013)

if you can modify where people look at the face, you might be bale to prevent the ORE. Face perception training using a fixation cross - high (near eyes) or low (near mouth) when fixation was higher up, black RPs were better at recognising white faces than black. when cross was low they were better at recognising black faces. vice versa for white faces. Eye tracking confirmed this. Shows that by changing the fixation cross you change where people look on the face therefore preventing the ORE.

Elfenbein & Ambady (2002)

meta-analysis on ingroup advantage of judging emotions. looked at cross cultural exposure - increased contact with those in the other group means we can recognise their emotions better. another factor is minority vs majority status - there doesn't seem to be an in group advantage when you are a minority group judging a majority group. another factor was whether the emotions were genuine or not; they found this had little effect on how well people were able to identify the emotion. They also looked at the specific emotion involved - disgust and fear tended to show the biggest ingroup advantage. They also looked at identification - where you identify with a group you are less likely to show an ingroup advantage. This is backed up by Thibault, Bourgeois & Hess (2006) - people who 'identify' with cats were better at identifying the cats emotion

Rhodes et al (1989)

people find it harder to recognise their own races face when upside down because they process it configually, however its easier to recognise other races faces upside down because they were processed featurally.

Configural feature hypothesis

this is the idea that well known stimuli are processed in a configural way; this means a hollistic (global processing) way. less familiar faces are processed in a featural way - this means detail based, part orientated and less emphasis on the whole global image.

Free-Labelling

where you allow people to generate their own labels for the emotions being shown. this is difficult to code because there are so many synonyms for each emotion - how do you decide what is close enough to the real definition. Studies typically ask for categorical judgement followed by rating of intensity on likert scale.