"Look, these two girls are having fiin playing marbles together. Oh, look another girl has come along with a shiny ball. Now, look, these two girls are going off to play ball, leaving this girl by herself."
—How do you think she feels now?
—Choose a face that shows me what her face looks like now that she feels that way -What's that face?
—Why does she feel that way?
2.2.3. The coding system
For each participant, responses to the Emotion-Recognition Task are transcribed and then coded. Variables which were identified for inclusion in this coding system are directly related to the original aims of the task discussed previously: correctness of identification for each facial expression and total number accurate on the spontaneous labelling phase of the task, range of emotions and faces used on the overall task, embellishment and appropriateness of
justification, coherence, 'matching' of family-character emotions, and appropriate use of multiple emotions. Each variable and procedure for coding will be described below.
2.2.3.a. Phase one: Spontaneous labelling o f emotion faces
Each child is asked to label spontaneously the nine primary emotion faces in the first phase of the task. Participant's responses when attempting to identify the emotion faces used on the task are scored either 1 (correct), 2 (incorrect), 3 ("I don't know"), or 10 (an emotion outside of the 9). "Correct" in this case means that the child labels an emotion face accurately according to what the face is intended to display by the London Parent-Child Project, which based its opinions on similar tasks (Ekman, 1972) as well as consultation with a professional
cartoonist. '’Incorrect" means that the child does not give an appropriate response, but gives one o f the other eight 'basic' labels. "I don't know" as a response is where the participant gives no relevant response when asked to label the emotion face. "An emotion outside of the 9" means that the participant gives a response which is unmatched with the intended expression, and uses another emotion word which is outside the basic emotions portrayed by the 9 faces presented (for example: a child labels a given face "tired," "hungry," "silly," "proud," "confused," "crazy," etc). Answers scored 3 and 10 in the present study were later collapsed into the
"incorrect" category for further analysis. Listed below are the terms that are coded as "correct" for each facial expression:
1. afraid (scared, frightened)
2. angry ('mad' is accepted for American samples)
3. sad
4. disgusted (yucky)
5. plain/neutral (ordinary, feeling nothing)
6. happy and angry/mischievous (sneaky, cheeky, naughty, 'bad' is accepted for American samples)
7. surprised and sad
8. happy
9. surprised (shocked)
Finally, the total correctly labelled is also calculated, to give each participant an overall score on this phase of the task.
2.2.3.b. Phase two: Discussing the emotive scenes Ran2e o f emotion and emotion-faces
In phase two of the task, the participant is asked how each cartoon figure is feeling in the particular situation, and asked to assign a face to the cartoon figure from the nine given in phase one, as well as justify the emotion discussed by answering the question, "Why d o e s___________feel that way?" Emotion and emotion-face answers are registered from transcripts, and assigned numbers 1
through 9 for the primary emotion faces, 10 where participant discusses an emotion outside o f the 9, and 11 where participant discusses assigning a multiple emotion (or two emotion faces), to a character, which is outside of the multiple emotions included in the list of 9. For the present category, the number o f different emotions and emotion faces discussed on the overall task for each participant is counted, after being registered as above from transcripts, to assess the range of the individual's emotion vocabulary and the range of facial expressions utilised to illustrate their discussions.
Justification
Justifications are considered in terms o f differences in embellishment and plausibility of response. Firstly, a justification is given a score of 1 if it is irrelevant to the emotion given or to the scene, or if no response is given, by the participant remaining silent or saying, "I don't know." Other examples of
justifications scored 1 are when the participant begins to talk about a situation in his/her own life, never addressing the emotion question, or if the child avoids replying by simply answering with a response given previously. (For example: the participant previously answered, "sad" when asked how the character feels; when asked why the character feels that way, the participant says, "sad" again.)
Secondly, a justification can be given a score of 2. This occurs when the
participant simply repeats the narrative they have been given by the experimenter. In this case, the participant does not add any of his/her own explanation as to why the character is feeling this particular emotion, but simply quotes back what the experimenter has told him/her about the scene. (For example: in situation four the end o f the narrative is given by the experimenter: "They are going away for the
weekend." The participant then answers the emotion and emotion face questions. When asked why this emotion was chosen, the participant says, "Because they are going away for the weekend.") Finally, a score of 3 can be assigned for a
justification if the participant gives an appropriately embellished response. This is where the participant goes beyond repeating the stem that was given and further discusses the story or scene at hand, by relating the story to him/herself, by talking about him/herself in a way that is relevant to the story, and/or by any discussion of the wants, wishes, desires, or emotions of the character. Overall embellishment of justification is scored for each character in each cartoon scene separately and then
these scores are averaged. The average score is then rounded to the nearest whole number on the scale to arrive at an overall rating o f embellishment for each participant.
Coherence
Coherence of overall response is considered. "Overall response" in this case refers to the participant's responses on each scene to the initial question, "How d o e s__________ feel?", the secondary request "Choose a face that shows me what his/her face looks like now that he/she feels that way," and the justification
question "Why d o e s__________feel that way?" Firstly, a score o f 1 is assigned if the child has given an overall answer which is low in coherence. That is, the child has discussed an emotion, chosen a face for that emotion, and given a justification, and there is incongruence in these answers. (For example: the child has given "sad" for emotion, chosen "angry" for emotion face, and given an irrelevant justification such as, "our train was late today," "can't tell," or "I don't know".) Moderate coherence can be given score 2 or 3. Score 2 is where the participant has
discussed a certain emotion for which he/she has then proceeded to give an incongruent emotion face. The response that scores 2, however, does give a reasonable justification for either the face or emotion chosen. A "reasonable" justification can entail either simply repeating the stem or giving an embellished
response which coherently explains why the participant has discussed a particular emotion or why he/she has chosen the selected emotion face. What is at issue when a response is given a score o f 2 is the incongruence o f emotion discussed and face chosen, and yet the congruence o f justification given. Score 3 is where the participant discusses a given emotion, and chooses an appropriate emotion face, yet does not justify his/her response. Justification is the difficulty when the
response is given a score of 3, while emotion discussed and face chosen are congruent. Finally, a score of 4 can be given where a response is highly coherent, that is, where emotion discussed, face given for that emotion, and justification given are all congruent. A response can be given a 4 if the child simply repeats the stem for his/her justification, providing that repeating the stem is an appropriate response in relation to the emotion and face choices o f the participant. Coherence of overall response is scored for each character in each cartoon scene separately and then these scores are averaged. The average score is then rounded to the nearest whole number on the scale to arrive at an overall coherence score for each participant.
Matchins o f Parent-Child emotions
The task is assessed in light o f emotion-matching, or display o f empathy, in the family, by separating out the cartoons where parent and child are present and scoring these for matching o f emotion between child character, mother character.
and father character. Scoring is on a present/absent basis (1, present; 2, absent) where present means that the parent cbaracter(s) in the scene is given the same emotion, emotion face, and plausible justification as child character in the scene, and where absent means that the participant has not done this (i.e. the participant has given distinctive emotions, emotion faces, or justification to the family characters). "Matching o f emotion" is scored for cartoons 2 (the Bicycle scene), 5 (the Sibling scene), 7 (the Separation scene), and 11 (the Reunion scene). See above in Table 2.1 for a description of these narratives. In scenes where
participants are asked to discuss the emotions of the child character and both the mother and father characters, a score of 1 (present) is given when only one parent character is emotionally matched with child character as well as when both parent characters are matched with child character.
Multiple emotions
Several of the cartoon scenes used in the Emotion-Recognition Task are regarded by the author as containing sufficient ambivalence to act as potential elicitors for a discussion of 'mixed feelings.' The Building Blocks scene possesses this characteristic most prominently. This scene shows a teacher addressing a positive situation with one of her students who has been very successful at a task, while concurrently addressing a negative situation as another student is having difficulty with the task. Examples o f participants' descriptions o f mixed emotions with reference to several other scenes across the task may be used to illustrate the power o f these scenes to invoke such a discussion as well. For the Separation scene, one participant describes mother-character feeling both sad and happy while being separated from her child, explaining that the mother is sad about leaving her
child but happy that she can return home to finish her house work. Another example regards the Sibling scene, where one participant describes that while the mother character is happy to be feeding her baby, she is also sad that she doesn't have time to look at her older child's drawing. One participant responded with a discussion of mixed emotions when discussing the Biscuit scene. This participant said that the child-character was angry at his mother for saying he should not have the biscuits and happy to have his hands on them again.
Two multiple emotion faces are introduced to the participant in the list of basic emotion faces in phase one of the task. They are also made available in phase two of the task during the discussions of character emotions and expressions. These expressions are a surprised AND sad face, and a happy AND angry face (also called mischievous, cheeky, naughty, or sneaky). A child's discussion of multiple emotions may also utilise two o f the other "singular" emotions/emotion faces from the list, or from outside the list. A score of 1 is given when no
mention of multiple emotions is made by the participant. An overall score of 2 is given when participant discusses multiple emotion but neither gives a reasonable justification for this discussion nor uses the appropriate face(s) to express the
multiple emotion discussed. This rating can be assigned where a participant chooses two inappropriate emotion faces or where participant simply chooses one face which may or may not be appropriate for " h a lf of the multiple emotion discussed. When given a rating of 2 the participant has expressed an ability to discuss multiple emotion, but is showing an inability to complete that discussion, as he/she is unable to justify this choice or to assign appropriate facial expressions to it. When given an overall score of 3, a participant has used multiple emotions
in his/her response and displays a greater level of understanding, either by justifying that response OR by choosing the appropriate face(s) to express the
multiple emotion given. In this case the participant does not respond appropriately with both justification AND choice of emotion face(s), but shows a partial
understanding by achieving one or the other. A score o f 4 is given where a participant discusses multiple emotions and displays a complete understanding through justification and through the choice o f appropriate face(s) to express this multiple emotion. One multiple emotion score is given for each participant, reflecting his/her highest level of skill displayed across the task.
Table 2.2:
Examples from the Emotion-Recognition Task in the Sixth Year
Story and questions
(Example o f story coded as a 4 on coherence fo r each character, 3 on justification embellishment fo r each character, 4 on discussion o f multiple emotion, given a 'matched' score, and given a 2 fo r
range o f emotions discussed and faces utilised in this scene)
(experimenter showing cartoon scenes in sequence)
Now see it looks like it's half past eight and it's time for the little girl to go off to school. Here mummy is taking the little girl to school. There they are, and there is mummy saying goodbye and leaving her there.
How does that little girl feel? sad and happy
Choose a face for that little girl then if she is sad and happy, child chose sad and happy faces
Why does that little girl feel sad and why does she feel happy?
Because she is sad to leave her mummy but she is happy that she can go to play at school. How does the mummy feel?
sad and happy too
Choose a face for the mummy then if she is sad and happy too. child chose sad and happy faces
Why does the mummy feel sad and why does she feel happy?
Because she is sad that her little girl is sad but she is happy that she can go home to get some work done without the little girl under foot!
Story and questions
(Example o f story coded as a 4 on coherence fo r the two child characters and a 2 fo r the teacher character, a 3 on justification embellishment fo r all characters, 3 on discussion o f multiple emotion, and a 3 fo r range o f emotions discussed and on faces utilised in this scene)
(experimenter showing cartoon scenes in sequence)
Now see it looks like the teacher has come in with some building blocks for the boys to play with. Now, this little boy is doing really well building with his building blocks, and it looks like this other little boy is having some trouble with his building. Then the teacher comes back in to see how the building is going.
How does that little boy feel, the one who is having some trouble building? angry
Choose a face for him then if he is feeling angry, child chooses the angry face
Why does that little boy feel angry then?
because he just keeps trying to make a building, trying to make a tower, and it just keeps falling, falling...he's just no good at it. He wants to be like him.
How does the little boy feel, the one who is doing really well building? happy
Choose a face for him then if he is feeling happy, child chooses the happy face
Why does that little boy feel happy then?
because, he is the best, because he has the most, he is the best builder, and he [the other boy] wants to be like him.
How does the teacher feel?
Uhhh...she feels....uhhhh...she feels happy and sad Choose a face for her then if she is happy and sad, child chooses the happy and angry (naughty) face. Why does the teacher feel happy and sad then?
Well...she just doesn't know, um...she can't decide...she just, she's happy that he did a good one and sad that he did a bad one.
2,2.3.c. Inter-rater reliability fo r the coding system
Two raters were used for the present study. One was an undergraduate psychology student from America, studying in the U.K., and the author was the second. Final coding of the entire sample was completed by the author, subsequent to establishing inter-rater reliability on the above codes for 25 participants, 40% of the main sample. Again, the Emotion-Recognition Task categories, where double coding was considered, are as follows: spontaneous labelling responses, range o f emotions used, range o f faces utilised, overall embellishment o f justification, overall coherence, matching score fo r four family scenes, and multiple emotion score. Inter-rater reliability was calculated using Spearman Rank Correlation
Coefficient, a statistical test which measures the amount and significance of correlation between scores on two variables (Miller, 1991). In the case of phase one o f the task, where emotion labels are given by each child and the total number correct is documented by coders, rho value was found to be .9525 (p<=.000, one tailed) between the author's and the second rater's codes. Again, this coding requires rater to determine whether participant identified facial expressions correctly according to the above guidelines, and then to give the total number labelled correctly. With 95% agreement between the two raters, inter-rater reliability is strongly established in this case for coding of phase one of the task.
100% agreement was found between the coders' calculations of emotion range used over the task, as well as for their calculations of range of facial expressions utilised for the characters over the task by each participant. This is not surprising
considering that these categories are extremely objective — the coder is simply required to register the number of emotions discussed, and number o f expressions used to illustrate this discussion, for each participant. Doing this simply involves each coder counting mention of these variables in participant's discussion, from the interview transcripts. 94% agreement was found between coders for ratings of ability to discuss multiple emotion (rho=.9370, p<=.000, one-tailed). This category