YOUNG PRESCHOOLERS' UNDERSTANDING OF EMOTIONS

The effects of age, response modality, and specific emotion on knowledge of happiness, sadness, anger, and fear, in 45 preschoolers aged 26-54 months (M =40.7 months) were examined. A contextually valid measure involving puppets was devised to capture the subjects' attention, and to embed the assessment within ongoing social interaction. Results with this measure indicated that: (a) the older subjects named and recognized emotional expressions better than the younger subjects; (b) the ability to recognize emotional expressions was greater than the ability to name them, especially for the fearful expression; (c) the ability to recognize happy expressions was greater than that for negative emotions (i.e., sadness, anger, and fear), whereas naming of the happy expression exceeded that for the anger or fear expressions; (d) the abilities to both recognize and name sad expressions exceeded corresponding abilities for anger or fear expressions; (e) the happy and sad situations were easiest to interpret; and (f) the subjects made specific errors for particular emotions. Theoretical and empirical implications of young preschoolers' substantial understanding of emotions are discussed.

Recent empirical work has suggested that both expressive and situational sources of information are important in emotion judgments, even at a very early age (Bretherton, Fritz, Zahn-Waxler, & Ridgeway, 1986; Bullock & Russell, 1986; Camras, 1986; Gnepp, 1983, 1989). The two sources of information go hand-in-hand; for example, the young child probably needs to be proficient in identifying emotional expressions before s/he can reliably associate them with certain situations (i.e., to know event scripts for specific emotions, e.g., "when I receive a gift, I feel happy"; see Bullock & Russell, 1986). Further, either source of information may be used by preschoolers when confronted with others' emotions, depending on the circumstances in which the emotion judgment is made (Camras, 1986). Because of the importance of both aspects of emotion knowledge, this study extends much previous research by assessing and comparing knowledge of both facial and situational sources of information in investigating two- to four-year-olds' understanding of emotions.

The first major goal of this research was to assess young preschoolers' abilities to identify emotional expressions by naming them (labeling) and pointing at them (recognition). Age change within the two response modalities also were specified. Ridgeway, Waters, and Kuczaj (1985) have confirmed that 60% to 90% of 24-month-old s recognize the four basic emotion terms (no t expressions; happy, sad, mad, and scared), whereas 83% to 97% use them in their own speech by 36 months. Similar age trends would be expected in identifying actual expressions (cf. Michalson & Lewis, 1985; Stifter & Fox, 1986). This study extends previous research by examining the abilities of younger children, in greater numbers than even Michalson & Lewis, to both verbally and nonverbally identify emotional expressions.

Further, given the lag cited between recognition and use of emotional terms (Ridgeway et al., 1985), it is also probable that children in the early preschool period would be more able to identify emotional expressions by pointing than by naming. Such a discrepancy could be due to developing language ability, as well as the relative ease of activating recognition versus recall of emotional memory. This study extends previous research by specifying emotional expressions for which this lag exists.

The second major goal was to examine age change in interpretation of emotional situations. An age-related increase in comprehension of emotional event scripts has previously been found in an older preschool sample (Borke, 1971), and with a sample which included some two-year-olds (Michalson & Lewis, 1985). The inclusion of labeling and recognition measures in this study, however, made it possible to explore the relation among identification of emotional expressions and situations; only one, or both, may change with age; age change in knowledge of emotional expressions may underlie age change in knowledge of emotional situations (cf. Bullock & Russell, 1986; Smiley & Huttenlocher, 1989).

The third major goal was to examine the effect of specific emotions (i.e., happy, sad, angry, and afraid) on naming and pointing at emotional expressions, and on interpretation of emotional situations. Theoretically, the personal and social information signalled by each emotion is quite different (Izard, 1971). Discrete emotions theory would also predict that children would come to understand emotions in the order that these emotions become adaptive in the young child's life. Thus, it was predicted that understanding of emotions would vary across emotions. Some emotional expressions and situations would be easier than others for young preschoolers to identify by either pointing or naming.

Happy expressions and situations are predicted to be the easiest to identify accurately (Borke, 1971, 1973; Michalson & Lewis, 1985). It is also likely that sad expressions and situations will be accurately identified, but that distinguishing among negative emotions, especially correctly identifying anger and fear, will be difficult for young preschoolers (Borke, 1971, 1973; Bullock & Russell, 1984; Michalson & Lewis, 1985; Stein & Jewett, 1986). Although others have explored emotion effects, this study allowed for generalizations across knowledge of expressions and situations. Moreover, this study's subjects, on the average, were younger than those in the previously cited studies.

The fourth major goal of this study was to go beyond specifying emotion-specific effects on accuracy, to identify emotion-specific errors in naming and pointing at emotional expressions, and in interpretation of emotional situations. Differentiation of happy versus nonhappy expressions and situations may be acquired very early (with few errors). Differentiating between anger and fear may, however, be difficult for young preschoolers (Borke, 1971, 1973; Bullock & Russell, 1984; Michalson & Lewis, 1985; Stein & Jewett, 1986). Anger also may be confused with sadness (Felleman, Barden, Carlson, Rosenberg, & Masters, 1983; Stein & Jewett, 1986).

Such errors also may change with age, with such age change differing for emotional expressions and situations. For example, in one previous study, preschoolers often pointed at the sad and angry faces, as well as the fearful face, when asked to find the face that was "afraid"; this tendency decreased over time (Bullock & Russell, 1986). Errors for identifying certain emotionally negative situations may, however, persist until later preschool years.

Further, a major unique contribution of our research was its attempt to ameliorate competence/performance errors of previous research. One way to eliminate the competence/performance problem is to contextualize the measurement of understanding of emotions. In this study, contextualized assessment of young preschoolers' interpretation of emotional situations was achieved through embedding the measurement of the dependent variables within ongoing play interaction with a familiar examiner, using attention-capturing, nonstatic emotional stimuli (i.e., puppets and the puppeteer's face and voice), and minimizing task demands, particularly for verbalization (Lewis, 1989). Further, memory for the categories of emotional situations were not assessed; instead, the child's ability to interpret ongoing emotional situations, similar to those experienced with peers, parents, and self, were measured.

Our study, thus, adds to the data reported by Borke (1971,1973) and Michalson and Lewis (1985), which it most closely parallels. Further, in comparison with Michalson and Lewis, more synonyms of the emotional terms for naming the expressions were accepted as correct. Moreover, photographic representations of emotional expressions were not used, in favor of simpler, less distracting, but still emotionally correct, schematic drawings of emotional expressions. Regarding the interpretation of emotional situations, pictorial representations of vignettes were not used because their use may overwhelm two-year-olds and many three-year-olds. Instead, a previously-validated play method was used (Denham, 1986). Thus, many methodological changes were made to eliminate competence/performance confounds of earlier research, and to extend that research.

To recapitulate, the major goals of this study were to specify: (a) the effects of age, specific emotion, and response modality on early identification of emotional expressions; (b) the effects of age and specific emotion on interpretation of emotional situations; and (c) errors which occur for each emotion, and their change with age.

Method

Subjects

Subjects were 45 two-, three-, and four-year-olds from two classrooms of a day care center in a large town in predominantly rural surroundings and two classrooms of a laboratory preschool in a suburban metropolitan area (overall mean age = 40.7 months; 20 girls, mean age 38.8 months, range 26-50 months; and 25 boys, mean age 42.3 months, range 30-54 months). Children were heterogeneous as to SES, judged by income and housing.

Procedure

There were two types of measures for each subject: (a) expression identification via verbal naming and nonverbal pointing; and (b) interpretation of emotional situations. Children left their classrooms one at a time with one of two adult female experimenters, whom they knew well and with whom they were comfortable. Subjects were selected in random order for testing. The measures of emotion knowledge were embedded within a structured play session described at length in Denham and Couchoud (1989).

Measures

Expression identification. Children examined four faces made of felt, on which the expressions of happy, sad, angry, and afraid were drawn (see Borke, 1971, and Izard, 1971, for the prototypical facial expressions which were drawn on the faces). These emotions were chosen from a discrete emotions theory perspective, and to parallel other studies' usage. The specific drawings have been validated as representing the facial expressions of happiness, sadness, anger, and fear.

Subjects were asked to name these four facial expressions (i.e., to answer the question, "what is this face feeling?"). Next, they. were required to point at each expression in answer to the question "where is the (underbar) face?." Thus, the first task included both verbal identification (naming) and recognition (pointing).

To ensure random order of questioning for each subject, the four faces were shuffled and laid on the table before each set of questions (i.e., naming and pointing). During the pointing task, all four faces were available on each trial; thus, subjects were not able to answer correctly due to the process of elimination. Subjects received one point for correctly naming or pointing at the expressions, for a possible total score of four on both the naming and the recognition measures. All spontaneous verbalizations, whether during naming or pointing tasks, were also included for error analyses. Cronbach's alphas were .73 for both the naming and recognition measures.

Interpretation of emotion situations. Family puppets made of cloth, with neutral facial expressions, were used to enact eight vignettes, similar to those of Borke (1971). Puppets were used to lessen cognitive demands of more representational media and to engage the children (Ridley, Vaughn, & Wittman, 1982). Protagonist puppets were the same gender as the subject.

In each vignette, the puppet felt the way most people would feel. For some emotions, such as happy, item creation is a simple task. Other emotional situations pose a problem in that it is possible for people to feel either of two ways. Because the potential developmental decline of interemotion confusion is, in fact, being studied here, its inadvertent inclusion in the measure itself would be a confound. Thus, the script of each vignette was chosen specifically to involve one major emotion (Lewis, 1989).

Each vignette, presented in a different random order for each subject, was accompanied by standardized vocal and visual affective cues emitted by the puppet/experimenter. Such systematically controlled cues from the experimenter were included in order to approximate the components of children's naturally-occurring emotion scripts. Thus, the puppet/experimenter showed standard peak facial expressions (see Izard, 1971), and expressed standard vocalizations, for each specific emotion. Props, such as miniature blocks, were also used. The experimenter correctly renamed the faces used in the previous task before proceeding.

Vignettes and standard facial and vocal expressions emitted by the examiner, as well as "body language" of the puppet, are described in Table 1. After seeing each vignette the children were asked "How does [the puppet] feel?" They then affixed the proper face, from the four identified in the previous task, on the puppet. Verbal responses were not required; such a requirement would substantially increase the difficulty of the measure. Scoring proceeded as in the nonverbal emotion identification task; thus, because two situations for each emotion were included, the maximum score per emotion was two, with a maximum total of eight. Cronbach's alpha equalled .82.

Results

Naming and Pointing at Emotional Expressions

Mean scores for the emotional expression naming measure and the emotional expression recognition measure for younger children, older children, and the total sample, are shown in Table 2. Mixed model Age (2) x Emotion (4) ANOVAs were performed for each measure. The mean age of the Younger group was 35.6 months, and the mean age of the Older group was 46.4 months. For simplicity, these groups will be called the three-year-olds and the four-year-olds.

Gender was not used as a factor in these and subsequent analyses. Preliminary Age (2) x Sex (2) ANOVAs for each emotion, for pointing, naming, and situation identification tasks, showed that its effects were routinely nonsignificant. The only exceptions were that girls showed greater ability than boys to name fear, p < .05, and a marginal tendency to point at angry expressions more readily, p < .10.

For both naming and pointing at emotional expressions, effects of age were significant, with four-year-olds' scores higher [Fs (1,43) = 9.71 and 5.54, ps < .01 and .05, respectively]. For both modalities, the effect of emotion was significant [Fs (3,129) = 34.37 and 6.21, ps < .0001 and .001]. Emotion x Age interactions were not significant for either task.

Planned comparisons for within-subjects designs indicated that identification of happy expressions was clearly the easiest naming or pointing task for the children. These were more readily named than either angry or fearful expressions, Fs (1,43) = 8.61 and 79.75, respectively, ps < .001, and more readily pointed at than all three negative emotions, Fs = 5.92, 14.20, and 19.87, ps < .05, .001, and .001, respectively. Sad and angry expressions were more readily named than fear expressions Fs = 60.21, and 35.20, respectively, ps < .001. There were no significant differences among negative expressions on the pointing measure.

Another mixed-model ANOVA was also performed with Age, Response Modality, and Emotion factors. The main effect of response modality was significant, F (1,43) = 13.17, p < .001. The mean naming score was thus lower (0.60) than that for pointing (0.69). The Emotion x Modality also was significant, F (3,129) = 13.21, p < .001. Simple main effects analysis showed that the locus of the interaction effect of modality lies in the difference between naming and point scores for the fear expression, F (1,43) = 25.04, p < .001. The mean score for naming fearful expression was .17, whereas the mean for pointing at these expressions was .56. The Age x Modality interaction was nonsignificant.

Interpretation of Emotional Situations

Table 2 also shows mean scores for each age group and the total sample on the emotional situation measure. An Age (2) x Emotion (4) repeated-measures ANCOVA design, with the sums of naming and pointing scores for each of the four emotions as covariates, was used to analyze these data. The main effect of age was nonsignificant, F (1,42) = 1.95, with the significant covariate partialled out, F (1,42) = 50.37, p < .0001. Thus, ability to interpret emotional situations did not increase with age, independent of the increasing ability to identify emotional expressions.

However, the emotion effect was significant, F (3,128) = 8.30, p < .01, even after partialling nonsignificant effects of the covariate, F (1,128) = 0.83. Planned comparisons for within-subjects designs indicated that understanding of happy emotions was again a strength in this age range. Happy situations were more easily identified than either angry or fearful situations, Fs (1,43) = 23.16 and 42.93, respectively, ps < .001. Happy situations were marginally more readily understood than were sad situations, F = 4.04, p < .06. Understanding of sadness was a strength among negative emotions; sad situations were identified more readily than either angry or fearful situations, Fs = 6.49 and 20.96, ps < .05 and .001, respectively. Angry situations were more readily identified than fearful ones, F = 4.49, p < .05. The Age x Emotion interaction was again nonsignificant, F (1,128) = 0.83.

Error Analyses

Total and age-group frequencies of each error made for each emotion within each task are listed in Tables 3,4 and 5. To investigate distributions of errors made by children during each of the tasks across the four emotions, chi-square analyses were performed where possible (i.e., no expected frequencies less than one, less than 20% of expected frequencies less than five). Where nonparametric analyses were not appropriate, only descriptive discussions follow.

Chi-square analyses were appropriate for the distribution of naming errors for fearful expressions; the distribution was nonrandom, X² (5) = 13.43, p < .05. Errors were most often "I don't know." The frequency tables also show that the happy expression was unlikely to be called a negative emotion, and anger and fear were often called other negative emotions. No chi-square analyses for pointing errors were appropriate, because of low expected frequencies. The frequency tables show that children pointed at sad and angry expressions when asked to identify the fearful expression, and the sad and fearful expressions when asked to identify the angry expression.

Age trends for errors in naming and pointing at expressions suggested that "I don't know" errors decreased with age, except for naming the afraid face. Confusing the angry or afraid faces with the sad face also decreased with age. Pointing at the angry face instead of the correct happy face decreased with age.

Regarding interpretation of emotional situations, appropriate chi-square analyses showed that error distributions were nonrandom for sad, angry, and fearful situations, X²s (3) = 11.00, 13.05, and 12.46, ps < .02, .01, and .01, respectively. Sad situations were likely to be called happy (perhaps an attempt at reparation), but unlikely to be called fear. Angry situations were often called sad, but unlikely to be called fear. Errors for the fearful situations were most often sad, but also unlikely to be angry. Descriptively, errors for happy situations were especially likely to be "I don't know," but unlikely to be angry or afraid. Age trends on this task also suggested that "don't know" responses decreased with age. Confusing happy and sad situations was an error made almost totally by three-year-olds. Older children were less apt than younger ones to call angry situations happy or sad; similarly, calling fearful situations happy or angry decreased with age.

Discussion

The predictions made regarding the basic aims of this study were generally upheld. These contextually valid, multi-method results emphasize the emerging but sizable understanding of emotion in young preschoolers. Such results have both theoretical and empirical implications.

Theoretical Implications

Our findings fit well with several theories of emotional development: Izard's (1971,1977) discrete emotions theory formed the basis for the hypotheses regarding identification of separate emotions' expressions and situations. This theory suggests that there exist separate neural mechanisms for each emotion, with emotional expressions emerging when adaptive in the infant's life. Izard's overall theory appears to be supported; the extent of abilities to understand both expressions and situations generally varied in the same order as the emergence of corresponding discrete emotional expressions (i.e., from greatest to least ability: happy, sad, angry, afraid; see Izard, 1971).

Lewis and Michalson's (1983) theory emphasizes that socialization factors are vital in the development of emotional expression and understanding. First, the infant comes to perceive the gestures and changes in face, voice, and posture in the emotional displays of others (Field & Walden, 1982). This perceptual ability may form the structural basis for lexical differentiation of expression terms later (Michalson & Lewis, 1985). Adult models emit and imitate happy expressions most often, and fear less often (Malatesta & Haviland, 1982), possibly explaining the ease with which children at the age levels investigated here are so proficient at identifying and interpreting happiness, and have such relative difficulty with fear. Those emotions which adults talk about more frequently may also become part of the child's growing emotion lexicon more readily.

Next, the child begins to assign meaning to these facial stimuli; for example, happy faces are seen when someone gets something, and sad faces when someone is hurt or doesn't get his/her way. Thus, the collection of emotional scripts available to the young child gradually enlarges (Bullock & Russell, 1986), until s/ he is as sophisticated as subjects tested here, with emotional script development depending on differentiation of facial expressions.

Empirical Implications

Age predictions. Ability to both name and point at emotional expressions increased with age, as expected. Results suggest, however, that the ability to identify emotional situations is dependent on expression identification. It seems that as children become able to recognize and name emotional facial expressions, they become capable of identifying the emotions expressed in situations (cf. Bullock & Russell, 1986; Smiley & Huttenlocher, 1989). The general linear increase of situation interpretation across age levels, which others report (Borke, 1971, 1973; Michalson & Lewis, 1985), and which is found in this study when expression identification is not covaried out [F (1,129) = 9.75, p < .01], may be artifactual, dependent on children's age-related increase in comprehension of emotional expressions.

Response modality predictions. Ability to name expressions was generally higher in this study than previously found, perhaps because of the contextually valid nature of the measure. In contextualizing measures to measure young children's abilities accurately, investigators need to remember this: Relative inability in verbalizing about certain emotions does not mean that very young children do not understand them (Lewis, 1989). Even so response modality (verbal vs. nonverbal) exerted an effect on identification of emotional expressions (see also Michalson & Lewis, 1985). As Michalson and Lewis assert, one of the general rules of language learning is that comprehension precedes production.

Furthermore, it is important to note that the locus of the Modality x Emotion interaction is the naming/pointing difference for the fear expression. Naming of and pointing at expressions are roughly equivalent across modalities for the other emotions, at fairly high levels. Further, though there were overall age effects for naming and pointing at emotional expressions, there was little improvement for naming of the fear expression across age levels. Lentz (1985) has shown that preschoolers also describe their own fears in ways which, to adults, are unsophisticated. Young children seem relatively able to point at the expression of fear, but to have trouble talking about it; identifying the reason for this dichotomy is a fruitful avenue for future research.

Specific emotion predictions and error analyses. Regarding comparisons among emotions, results for interpretation of situations replicated earlier findings with other, usually older, samples (cf. Borke, 1971; Carlson, Felleman, & Masters, 1983; Gomby & Ellsworth, 1985; Michalson & Lewis, 1985; Riess & Cunningham, 1986). Our results also closely follow others' results for identification of emotional expressions (e.g., Bullock & Russell, 1984; Walden & Field, 1982), that is, interpretation of happiness was closely followed by or equal to interpretation of sadness, which exceeded interpretation of anger, with fear being most difficult to identify either as an expression or situation.

Error frequencies show that children rarely confused happiness with negative emotions in identifying happy expressions and situations. In fact, these young children's identification of emotional expressions and situations was most like adults' in the case of happiness (Bullock & Russell, 1984, 1986; Izard, 1971; Michalson & Lewis, 19850. Moreover, our subjects' concepts of emotional expressions and situations were organized along the pleasant/unpleasant dimension. They understood that happy expressions and situations were pleasant and that sad or "bad" expressions and situations were negative.

In contrast, negative emotions, compared to happiness, were indeed difficult to differentiate at these age levels. The children demonstrated "fuzzy borders" for some of their negative emotion concepts (Bullock & Russell, 1984, 1986). For example, they often said "I don't know" in naming negative emotional expressions, and often picked the sad face when either of the other negative emotions was the correct choice, especially in the expression recognition and situation identification tasks.

Situations involving certain negative emotions, particularly sad and angry, may continue to confuse young children because both involve similar motivational bases in interpersonal contexts (Farber & Moely, 1979; Stein & Jewett, 1986). In a given situation, either emotion may be appropriate, depending on the emphasis given by the particular person --either of the blocked state, resulting in anger, or the loss of the wanted object, leading to sadness. Thus, although young children are confusing anger for sadness in interpreting situations, some of this confusion may be inherent in the nature of our expressions of these emotions in our social world. Another possibility is that the children chose sadness instead of anger in situations because of social desirability (Felleman et al., 1983). It also is noteworthy that neither angry nor sad situations are confused with fear-provoking situations; in the case of fear, an unwanted state of affairs is anticipated, and this crucial difference may support the rejection o f fear in these cases. Regardless, age trend s for errors on all three tasks, except for fearful expressions and situations, and perhaps anger situations, suggest that four-year-olds have become very competent indeed at understanding emotions.

General Conclusions

Two- and three-year-olds in this study, when compared with those in Michalson and Lewis' study (1985), were more adept at naming happy, sad, angry, and fearful faces, pointing at faces, and interpretation of angry and fearful situations. Thus, for the abilities posited to be more difficult (i.e., naming emotional expressions and interpreting more complex negative emotional situations), the current methodology may truly obviate earlier competence/ performance difficulties. Lessened memory demands and approximation of real-life emotional interaction appears to aid two- and three-year-olds in demonstrating their substantial emotional knowledge.

Further, effects of age and specific emotion were rendered more interpretable for both types of emotion knowledge because the same subjects were used in measuring each. Moreover, this design made possible demonstration of the relative dependence of interpretation of emotional situations on identification of emotional expressions. Errors which young preschoolers make in their judgments about emotional expressions and situations also have been more clearly delineated than before. Knowledge of such errors may facilitate adult understanding of the specific misperceptions very young children make in identifying and interpreting emotions. Such adult understanding could lead to more precise emotion socialization in daycares and preschools.

Footnotes

An earlier version of this paper was presented at the meeting of the American Psychological Association, August, 1986, in Washington, DC. The authors gratefully acknowledge the assistance and participation of the directors, teachers, and, especially, the children, of The Carroll County Committee for the Daycare of Children, Unit III, Westminster, MD (Mrs. Ann Fair, Director), and The Project for the Study of Young Children (Dr. Christine Burger, Director).

Reprint requests can be sent to:

Susanne A. Denham
Department of Psychology
George Mason University
4400 University Drive
Fairfax, VA 22030

TABLE 1 Vignettes and Indicators of Emotion Shown by Puppet and Puppeteer in Situation Interpretation Measure

The following chart reads as follows:

Row 1: Vignette Content

Row 2: Hand Puppet's "Body Language"

Row 3: Puppeteer's Facial Cues

Row 4: Puppeteer's Vocal Cues

1 Ice Cream Cone: Happy

      "Bounces," spreads arms

      Broad smile

      Relaxed, "pearly" tones

2 Pushed Down by Peer: Sad

      Wipes eyes, Head downcast

      Eyes & mouth down-turned

      Whiny, crying tones

3 Block Tower Destroyed by Peer: Angry

      Clenched

      Eyebrows down, lips pursed

      Gruff, growling, abrupt

4 Nightmare: Fear

      Hands up Rigid

      Eyes wide, mouth gaping

      High-pitched, unwavering

5 Going to to the Zoo: Happy

      "Bounces," spreads arms

      Broad smile

      Relaxed, "pearly" tones

6 Big Wheel Stolen: Sad

      Wipes eyes, Head downcast

      Eyes & mouth down-turned

      Whiny, crying tones

7 Dark: Fear

      Hands up Rigid

      Eyes wide, mouth gaping

      High-pitched, unwavering

8 Forced to Eat: Angry

      Clenched

      Eyebrows down, lips pursed

      Gruff, growling, clipped, abrupt

TABLE 2 Mean Scores for the Naming, the Recognition, and the Situation Measures by Age Group and for Total Sample

                                  Age Groups
Emotion                    Younger        Older[a]         Total

Naming Expressions

Happy                         0.67           1.00           0.82
                            (0.48)         (0.00)             --

Sad                           0.62           0.90           0.76
                            (0.49)         (0.30)             --

Angry                         0.50           0.76           0.62
                            (0.51)         (0.44)             --

Afraid                        0.08           0.28           0.18
                            (0.28)         (0.46)             --

Age Group Mean                0.47           0.74             --

Pointing at Expressions

Happy                         0.75           1.00           0.87
                            (0.44)         (0.00)             --

Sad                           0.62           0.76           0.69
                            (0.49)         (0.44)             --

Angry                         0.50           0.81           0.64
                            (0.51)         (0.40)             --

Afraid                        0.46           0.67           0.56
                            (0.51)         (0.48)             --

Age Group Mean                0.58           0.81             --

Situation Identification

Happy                 1.46/1.40[b]      1.81/1.49           1.62
                            (0.78)         (0.51)             --

Sad                      1.12/1.14      1.62/1.45           1.36
                            (0.95)         (0.74)             --

Angry                    0.62/0.75      1.33/1.21           0.96
                            (0.88)         (0.91)             --

Afraid                   0.42/0.74      1.10/1.24           0.73
                            (0.78)         (0.89)             --

Age Group                     0.91           1.46             --
Mean

Note: Standard deviations are in parentheses.

a Mean age of younger group = 35.6 months, of older group = 46.4 months.

b Means adjusted for covariate displayed after slash.

TABLE 3 Error Frequencies for Naming Expressions

Legend for Chart:

A - Errors: Emotion
B - Errors: Happy
C - Errors: Sad
D - Errors: Angry
E - Errors: Afraid
F - Errors: Bad
G - Errors: Don't Know
H - Errors: Other

    A             B      C      D       E      F      G        H

Total            --      2      0       0      0      5        0

Happy

Three's          --      2      0       0      0        5      0
Four's           --      0      0       0      0        0      0

Total             0     --      1       0      3        5      2

Sad

Three's           0     --      1       0      1        5      2
Four's            0     --      0       0      2        0      0

Total             0      3     --       2      3        6      3

Angry

Three's           0      3     --       0      0        5      2
Four's            0      0     --       1      2        1      1

Total             3      5      7      --      4       14      4

Afraid

Three's           3      5      4      --      1        7      2
Four's            0      0      3      --      3        7      2

TABLE 4 Error Frequencies for Pointing at Expressions[a]

                         Errors
Emotion    Happy    Sad    Angry     Afraid    Bad    Don't
                                                      Know

Total         --      0       3          1       0       3
Happy         --     --      --         --      --      --
Three's       --      0       3          1       0       3
Four's        --      0       0          0       0       0
Total          1     --       3          5       2       3
Sad           --     --      --         --      --      --
Three's        1     --       2          2       1       3
Four's         0     --       1          3       1       0
Total          1      4      --          6       2       3
Angry         --     --      --         --      --      --
Three's        1      4      --          3       1       3
Four's         0      0      --          3       1       0
Total          1      9       5         --       2       4
Afraid        --     --      --         --      --      --
Three's        1      6       3         --       1       3
Four's         0      3       2         --       1       1

a There were no "other" non-verbal naming responses possible; only four faces were available from which to choose.

TABLE 5 Error Frequencies for Identifying Situations[a]

                      Errors
Emotion    Happy    Sad    Angry    Afraid    Don't Know

Total        --       5       1        1            8
Happy        --      --      --       --           --
Three's      --       5       1        1            8
Four's       --       0       0        0            0
Total        12      --       4        1            8
Sad          --      --      --       --           --
Three's      10      --       2        0            8
Four's        2      --       2        1            0
Total         8      19      --        3           12
Angry        --      --      --       --           --
Three's       6      12      --        1           12
Four's        2       7      --        2            1
Total        12      22       5       --           10
Afraid       --      --      --       --           --
Three's      11      11       4       --           10
Four's        1      11       1       --            0

a Errors for two stories per emotion.

References

Borke, H. (1971). Interpersonal perception of young children: Egocentrism or empathy? Developmental Psychology, 5, 263-269.

Borke, H. (1973). The development of empathy in Chinese and American children between three and six years of age: A cross-cultural study. Developmental Psychology, 9, 102-108.

Bretherton, I., Fritz, J., Zahn-Waxler, C., & Ridgeway, D. (1986). Learning to talk about emotions: A functionalist perspective. Child Development, 56, 529-548.

Bullock, M. & Russell, J. A. (1984). Preschool children's interpretations of facial expressions of emotion. International Journal of Behavioral Development, 8, 1538.

Bullock, M., & Russell, J. A. (1986). Conceptual emotions in developmental psychology. In C. E. Izard & P. Read (Eds.), Measurement of emotions in children (Vol. 2, pp. 203-237). New York: Cambridge University Press.

Camras, L. A. (1986). Judgments of emotion from facial expression and situational context. In C. E. Izard & P. Read (Eds.), Measuring emotions in infants and children (Vol. 2, p. 75-89). Cambridge: Cambridge University Press.

Carlson, C. R., Felleman, E. S., & Masters, J. C. (1983). Influence of children's emotional states on the recognition of emotion in peers and social motives to change another's emotional state. Motivation and Emotion, 7, 71-79.

Denham, S. A. (1986), Social cognition, prosocial behavior, and emotion in preschoolers: Contextual validation. Child Development, 57, 197-201.

Denham, S. A., & Couchoud, E. A. (1989). Situational contributors to preschoolers' responses to adult sadness, anger and pain. Unpublished manuscript.

Farber, E. A., & Moely, B. F. (1979,April). Inferring others' affective states: Their use of interpersonal vocal and facial cues by children of three age levels. Paper presented at the biennial meeting of the Society for Research and Child Development, San Francisco.

Felleman, E. S., Barden, R. C., Carlson, C. R., Rosenberg, L. R., & Masters, J. C. (1983). Children's and adults' recognition of spontaneous and posed emotional expressions in young children. Developmental Psychology, 19, 405-413.

Field, T. M., & Walden, T. A. (1982). Production and perception of facial expression in infancy and early childhood. In H. W. Reese and L. P. Lipsitt (Eds.), Advances in child development and behavior (Vol. 16, pp. 171-211). New York: Academic Press.

Gnepp, J. (1983). Children's social sensitivity: Inferring emotions from conflicting cues. Developmental Psychology, 19, 805-814.

Gnepp. J. (1989). Children's use of personal information to understand other people's feelings. In C. Saarni & P. L. Harris (Eds.), Children's understanding of emotions (pp. 151-180). Cambridge: Cambridge University Press.

Gomby, D. S., & Ellsworth, D. S. (1985,August). Preschool childrens' understanding of emotions. Paper presented at the annual meeting of the American Psychological Association, Los Angeles, CA.

Izard, C. E. (1971). The face of emotion. New York: Appleton.

Izard, C- E. (1977). Human emotions. New York: Plenum Press.

Lentz, K. A. (1985). The expressed fears of young children. Child Psychiatry and Human Development, 16, 3-13.

Lewis, M. (1989). Cultural differences in children's knowledge of emotional scripts. In C. Saarni & P. L. Harris (Eds.), Children's understanding of emotions (pp. 350372). Cambridge: Cambridge University Press.

Lewis, M., & Michalson, L. (1983). Children's emotions and moods: Developmental theory and measurement. New York: Plenum Press.

Malatesta, C. Z., & Haviland, J. M. (1982). Learning display rules: The socialization of emotion expression in infancy. Child Development, 53, 991-1003.

Michalson, L., & Lewis, M. (1985). What do children know about emotions and when do they know it? In M. Lewis & C. Saarni (Eds.), The socialization of emotion (pp. 117-140). New York: Plenum Press.

Ridgeway, D., Waters, E., & Kuczaj, $. A. (1985). The acquisition of emotion descriptive language: Receptive and productive vocabulary norms for ages 18 months to 6 years. Developmental Psychology, 21, 901-908.

Ridley, C A., Vaughn, S. R., & Wittman, S. K. (1982). Developing empathic skills: A model for preschool children. Child Study Journal, 12, 89-97.

Riess, J. A., & Cunningham, J. A. (1986,April). Differentiation and labeling of affective states by two- and three-year-olds. Poster presented at the annual meeting of the American Psychological Association, Washington, DC.

Smiley, P., & Huttenlocher, J. (1989). Young children's acquisition of emotional concepts. In C. Saarni & P. L. Harris (Eds.), Children's understanding of emotions (pp. 27-49). Cambridge: Cambridge University Press.

Stein, N. L., & Jewett, J. L. (1986). A conceptual analysis of the meaning of negative emotions: Implications for a theory of development. In C. Izard & P. Read (Eds.), Measurement of emotion in infants and young children (pp. 238-268). Cambridge: Cambridge University Press.

Stifter, C A., & Fox, N. (1986). Preschool children's ability to identify and label emotions. Journal of Nonverbal Behavior, 10, 255-266.

Walden, T. A., & Field, T. M. (1982). Discrimination of facial expressions by preschool children. Child Development, 53, 1312-1320.

~~~~~~~~

By Susanne A. Denham and Elizabeth A. Couchoud, George Mason University