The correct expression. Ratings were out of 3, with 3 being unsuitable, 2 being suitable, and 1 being jir.2013.0113 very suitable. Any participants with a suitability rating of 3 for any emotion were excluded. For the rest, a total suitability score across all emotions was calculated by adding up the 6 individual emotion scores for each participant. The images from the best scoring 15 participants in each of the four age/sex sets were then taken to be used to make prototypes. Prototype creation. Prototype faces of the 4 age/sex groups pulling each of the 6 expressions were Linaprazan web created using Psychomorph [12]. Each of the 360 images was delineated by placing 172 points in the same position on the features of each face. This allowed colour and texture information to be averaged across all 15 faces [12]. The intensity of the expressions was then increased. In order to do this, all 90 images from each age/sex group were averaged to produce 4 emotion ambiguous prototypical faces [17]. The linear vector difference of each delineation point on the emotion ambiguous prototype compared to emotion expression prototype was then increased to Cyclosporine web effectively `caricature’ the expression, making it more intense [18]. Colour and texture information from the emotional prototype was then rendered into this new shape. The level of caricature for each expression was chosen so that the face did not distort or look unnatural. An ambiguous rather than a neutral expression was used for caricaturing because the intention was to increase the perceptual difference between the 6 expressions. It makes sense therefore to choose an average of all expressions as the starting point, rather than a neutral expression which is often perceived as similar to an angry expression [19, 20]. Study stimuli. Original prototypes and caricatured prototypes for each of the 6 expressions on the 4 prototypical child faces described above were included in the study. Both the original prototypes and caricatured prototypes were included to check that increasing the intensity of the expression had been successful in making the expressions more recognisable. In addition to the child face prototypes, existing adult face prototypes were included. These stimuli had been developed and used in other unpublished studies. They were created from photos of individuals between the ages of 19?3 who were recruited from the undergraduate population at the University of Bristol through notices on campus. Individuals took part in a jir.2013.0113 20 minute photograph session and were reimbursed ?. Information on the exact age of these individuals is not available but all were within the narrow age range noted above. Table 2 summarises information about the faces used to construct the adult prototypes and those used to construct the child prototypes.Table 2. Faces used for prototypes. Younger Child Male n Age range (years) Age mean (years) Age SD (years) doi:10.1371/journal.pone.0125256.t002 15 5? 7 1.8 Female 15 5? 6 1.4 Male 15 9?2 10 1.2 Older Child Female 15 9?2 10 1.0 Male 12 19?3 Adult Female 12 19?PLOS ONE | DOI:10.1371/journal.pone.0125256 May 15,4 /No Own-Age Advantage in Children’s Recognition of EmotionFig 1. Examples of the 6 face prototypes each showing one of the 6 emotional expressions at each intensity. From left to right: age 5? female happy, age 5? male disgusted, age 9?2 female angry, age 9?2 male afraid, adult female sad, adult male surprised. The top row contains the the more intense versions and the bottom row contains.The correct expression. Ratings were out of 3, with 3 being unsuitable, 2 being suitable, and 1 being jir.2013.0113 very suitable. Any participants with a suitability rating of 3 for any emotion were excluded. For the rest, a total suitability score across all emotions was calculated by adding up the 6 individual emotion scores for each participant. The images from the best scoring 15 participants in each of the four age/sex sets were then taken to be used to make prototypes. Prototype creation. Prototype faces of the 4 age/sex groups pulling each of the 6 expressions were created using Psychomorph [12]. Each of the 360 images was delineated by placing 172 points in the same position on the features of each face. This allowed colour and texture information to be averaged across all 15 faces [12]. The intensity of the expressions was then increased. In order to do this, all 90 images from each age/sex group were averaged to produce 4 emotion ambiguous prototypical faces [17]. The linear vector difference of each delineation point on the emotion ambiguous prototype compared to emotion expression prototype was then increased to effectively `caricature’ the expression, making it more intense [18]. Colour and texture information from the emotional prototype was then rendered into this new shape. The level of caricature for each expression was chosen so that the face did not distort or look unnatural. An ambiguous rather than a neutral expression was used for caricaturing because the intention was to increase the perceptual difference between the 6 expressions. It makes sense therefore to choose an average of all expressions as the starting point, rather than a neutral expression which is often perceived as similar to an angry expression [19, 20]. Study stimuli. Original prototypes and caricatured prototypes for each of the 6 expressions on the 4 prototypical child faces described above were included in the study. Both the original prototypes and caricatured prototypes were included to check that increasing the intensity of the expression had been successful in making the expressions more recognisable. In addition to the child face prototypes, existing adult face prototypes were included. These stimuli had been developed and used in other unpublished studies. They were created from photos of individuals between the ages of 19?3 who were recruited from the undergraduate population at the University of Bristol through notices on campus. Individuals took part in a jir.2013.0113 20 minute photograph session and were reimbursed ?. Information on the exact age of these individuals is not available but all were within the narrow age range noted above. Table 2 summarises information about the faces used to construct the adult prototypes and those used to construct the child prototypes.Table 2. Faces used for prototypes. Younger Child Male n Age range (years) Age mean (years) Age SD (years) doi:10.1371/journal.pone.0125256.t002 15 5? 7 1.8 Female 15 5? 6 1.4 Male 15 9?2 10 1.2 Older Child Female 15 9?2 10 1.0 Male 12 19?3 Adult Female 12 19?PLOS ONE | DOI:10.1371/journal.pone.0125256 May 15,4 /No Own-Age Advantage in Children’s Recognition of EmotionFig 1. Examples of the 6 face prototypes each showing one of the 6 emotional expressions at each intensity. From left to right: age 5? female happy, age 5? male disgusted, age 9?2 female angry, age 9?2 male afraid, adult female sad, adult male surprised. The top row contains the the more intense versions and the bottom row contains.
