The adaptation and validation of the Emotion Matching Task for preschool children in Spain

Abstract

The Emotion Matching Task (EMT; Izard, Haskins, Schultz, Trentacosta, & King, 2003) was developed to assess emotion knowledge in preschoolers and was demonstrated to show adequate convergent and predictive validity in an American sample (Morgan, Izard, & King, 2010). In light of the need for valid measures for assessing emotion knowledge in Spanish-speaking children, we adapted the EMT for preschoolers in Spain and conducted a psychometric validation with 110 Spanish children, aged 3 to 6. The EMT Spanish version showed good internal consistency and demonstrated expected concurrent validity with externalizing problems and adaptive abilities. Therefore the current study indicates that the adapted EMT is a valid tool to assess preschool children’s emotion knowledge in Spain.

Keywords

emotion knowledge Emotion Matching Task preschool Spanish validity

The facial expression is the basic route of emotional expression, has an important universal communicative function, and provides relevant situational information to an observer (Ekman, 1993; Izard, 1977). From infancy, children recognize facial expressions, imitate them, and develop their own cognitive framework regarding discrete expressions, as well as understanding the attached emotional meaning (Denham & Burton, 2003; McClure, 2000). Emotion knowledge is a set of skills that describes young children’s adaptive use of emotional information (Izard, 2001). It includes the ability to: (a) recognize emotional signals in facial expressions, voice, or behavior (receptive emotion knowledge); (b) label emotions correctly (expressive emotion knowledge); and (c) identify the causes or triggers of emotions and anticipate emotions in real or imaginary situations (emotion situation knowledge; Morgan et al., 2010). The individual components of emotion knowledge develop gradually in early childhood, with the receptive ability typically emerging first, followed by expressive emotion knowledge, and lastly emotion situation knowledge (Bassett, Denham, Mincic, & Graling, 2012; Morgan et al., 2010).

Emotion knowledge may play a key role in the social and emotional adjustment of young children as poorly developed emotion knowledge may result in an inappropriate response in a social situation. A recent meta-analytic study revealed a positive relation between emotion knowledge and social skills, as well as negative relations between emotion knowledge and both behavioral and emotional difficulties (Trentacosta & Fine, 2010). Moreover, emotion knowledge is posited to be the basis of emotion regulation development (Denham et al., 2011; Trentacosta & Izard, 2007).

Research investigating emotion knowledge development has demonstrated that it is related to both age and verbal ability (Bennett, Bendersky, & Lewis, 2005; Brown & Dunn, 1996; Denham, 1998; Widen & Russell, 2010). Children’s development and increasing exposure to emotion-eliciting experiences helps them to develop more advanced emotion knowledge. Similarly, the maturation of verbal ability fosters the acquisition and use of emotional language and paves the way for subsequent emotional understanding (Izard, Stark, Trentacosta, & Schultz, 2008; Zeidner, Matthews, Roberts, & McCann, 2003). Previous research findings are less clear regarding the relation between gender and emotion knowledge. In general, research shows that females score higher than males (Bennett et al., 2005; McClure, 2000; Widen & Russell, 2010), although some studies fail to find any gender differences (Dunn & Cutting, 1999; Miller et al., 2006).

Assessment of emotion knowledge

In order to study emotion knowledge in preschool-aged children, reliable and valid measures are needed (Zeidner et al., 2003). Various measures of emotion knowledge have been developed for use with American preschoolers, including the Emotion Matching Task (EMT; Izard et al., 2003). Unfortunately there is a dearth of preschool emotion knowledge assessments validated for use with Spanish preschoolers.

The EMT is a highly attractive tool as it assesses different facets of emotion knowledge in preschool children. It has demonstrated good reliability and validity in an American sample, and it is user-friendly. The tool assesses emotion knowledge in reference to happy, sad, anger, and fear/surprise, the four discrete emotions that emerge first in development (Lewis, 2008). The EMT is comprised of four parts: (1) expression matching; (2) emotion situation knowledge; (3) expressive emotion knowledge; and (4) receptive emotion knowledge. It provides a total emotion knowledge score, as well as a score for each emotion knowledge dimension, thereby aiding in the study of the development of the individual components of emotion knowledge. The EMT has been shown to relate to children’s emotion regulation and effortful control, as well as to behavioral problems (Morgan et al., 2010), which are all traditional correlates of emotion knowledge (Trentacosta & Fine, 2010). The EMT has also been shown to be sensitive to change in an emotion-based preventive intervention (Izard, King et al., 2008).

Current study

The aim of the current study was to investigate if the Emotion Matching Task is a reliable and valid assessment of emotion knowledge with preschoolers in Spain. We first examined the internal consistency of the EMT and then evaluated its concurrent validity with behavioral outcomes in the Spanish population. Suitable reliability indices for the Spanish version of the EMT were expected, similar to the original validation study (Morgan et al., 2010). It was hypothesized that there would be a positive association between emotion knowledge and adaptive abilities, as well as a negative relation between emotion knowledge and both externalizing and internalizing problems. We also explored the individual differences in preschoolers’ emotion knowledge profiles. Based on previous research, we expected a positive relation between emotion knowledge and both age and verbal ability. We further expected better performance for girls on the EMT.

Method

Participants

Participants were 110 children from Spain (56 boys and 54 girls). Data were collected from three preschools in Spain, two in Madrid and the other in Málaga, both urban areas. Children’s age ranged from 36 to 72 months (M = 52.25; SD = 10.21). In the current sample, 90.9% were white, 3.64% African-European, 2.73% Romani, and 2.73% biracial. All demographic areas involved in the study were middle class, though socioeconomic information was not collected for the entire sample, and thus no specific information on income level or socioeconomic status (SES) is available.

Measures

Emotion knowledge

The Emotion Matching Task (EMT; Izard et al., 2003) measures emotion knowledge, providing a total score and four dimension scores. It is divided into four parts, each with 12 items scoring 0 (incorrect) or 1 (correct). In part 1 (expression matching), the child is shown a photograph of an expression and asked to choose one of four photographs showing the same expression. In part 2 (emotion situation knowledge), the child is asked to indicate which of the four pictures matches a given situation, for example “show me who has just been pushed away from the table.” In part 3 (expressive emotion knowledge), the child is shown a single picture and is asked to look at the face and say how the child in the picture feels. In part 4 (receptive emotion knowledge), the examiner states an emotion and asks the child to point to the corresponding emotional expression on a sheet of four pictures. Alphas from the American validation of the EMT were .65, .54, .76, and .80 for parts 1 to 4, respectively, and .88 for EMT total score (Morgan et al., 2010).

The tool consists of a series of photographs of children (both boys and girls) whose emotional expressions are similar to those found in everyday life, thus making the test attractive for young children. Ethnic diversity of the children featured in the photographs is represented (African-American, Hispanic, White, and biracial).

Verbal ability

The Peabody Picture Vocabulary Test (Dunn, Dunn, & Arribas, 2006) assessed children’s receptive vocabulary. The examiner orally presented a word while displaying a sheet with four pictures from which the child was instructed to choose the one best representing the meaning of the given word. The difficulty increases progressively. The Spanish version of the Peabody Picture Vocabulary Test is reported to have good internal consistency (α = .91) and is highly correlated with other verbal ability and general intelligence measures (Dunn et al., 2006).

Assessment of behavior in children

The Spanish version of the Behavior Assessment System for Children (Reynolds & Kamphaus, 2004) is a 106-item questionnaire completed by teachers. The Behavior Assessment System for Children is a highly recognized and internationally used measure to evaluate children’s adaptive and maladaptive behavior. It results in scores on three main dimensions: externalizing problems, internalizing problems, and adaptive abilities. Alphas for the validated Spanish version are 0.95, 0.87, and 0.85, respectively (Reynolds & Kamphaus, 2004).

Procedure

The adaptation of the EMT items began with a process of double translation (Hambleton & Patsula, 1999) where, initially, two teams of translators familiar with literature on emotional content independently translated the EMT items into Spanish. The two translations were compared and discussed until an agreement was reached for each item. Subsequently, another two translators independently translated the items in the Spanish version back into English. Possible non-equivalences were examined with regard to meaning and the appropriate modifications were introduced. This version (see Appendix 1) was administered to a small group of children to identify possible difficulties with understanding or administration.

Educational psychologists trained in the EMT and Peabody Picture Vocabulary Test (PPVT) administered these measures to the children. The data were collected at schools in individual assessment sessions conducted in an area set aside for the purpose. The psychologists were introduced to each group of children before the tests were administered to ensure good rapport. The Peabody Picture Vocabulary Test was administered first, followed by the EMT. Finally, the teachers completed the Behavior Assessment System for Children.

Analytic plan

Internal consistency was examined using polychoric correlations by omega (Ω) coefficients, which are deemed a better estimator of reliability for items with a categorical nature, such as the EMT, as compared to alpha coefficients (Zumbo, Gadermann, & Zeisser, 2007). Omega values can be interpreted similarly to alpha coefficient values (Zinbarg, Revelle, Yovel, & Li, 2005). Descriptive analyses were conducted to describe the sample. Simultaneous regression analyses were then carried out to examine individual differences in emotion knowledge by age, gender, and verbal ability. Lastly, in order to study the EMT’s concurrent validity, given that there are no emotion knowledge tools for preschoolers in Spain to compare with, we conducted stepwise regression analyses to further examine the EMT as a predictor of criterion variables of adaptive and maladaptive behavior. Control variables (age, verbal ability, and gender) were entered on the first step followed by the individual parts of the EMT as predictors on the second step.

Results

Descriptive data

Table 1 provides the means, standard deviations, and ranges for age, verbal ability, EMT scores, and Behavior Assessment System for Children scores.

Table 1.

Means, standard deviations, and ranges for all study variables

Dimension	M	SD	R	N
Age (months)	52.25	10.21	36–72	110
PPVT	108.52	10.77	72–133	110
EMT total	32.40	8.88	6–46	110
EMT 1	8.05	2.23	2–12	110
EMT 2	6.81	2.38	2–11	110
EMT 3	8.35	3.09	0–12	110
EMT 4	9.34	2.62	1–12	110
Externalizing problems	14.65	12.19	0–49	109
Internalizing problems	7.92	5.88	0–30	109
Adaptive abilities	31.82	11.44	7–53	109

EMT internal consistency

We evaluated the reliability of the EMT total score and each of the parts through polychoric correlations. We calculated omega coefficients, obtaining values of 0.82 for the EMT total score, and 0.83, 0.82, 0.94, and 0.91 for parts 1 to 4, respectively.

Relations between emotion knowledge and age, verbal ability, and gender

In order to examine the pattern of relations between emotion knowledge and age, verbal ability, and gender, we conducted simultaneous regression analyses for the total score and each part of the EMT (see Table 2). Age and verbal ability were significant predictors of EMT total score, as well as all of the emotion knowledge dimensions, suggesting that emotion knowledge is more developed for older children and for children with stronger verbal abilities. The findings with gender were less consistent, with gender significantly predicting expression matching (EMT 1) and expressive emotion knowledge (EMT 3), as well as total EMT. Specifically, girls scored higher than boys did on these parts.

Table 2.

Simultaneous regression of age, verbal ability, and gender on the EMT

	β	Total R ² _ADJUSTED	N
Total EMT		.63**	110
Age	.62**
Verbal ability	.34**
Gender	−.16**
Expression matching (part 1)		.42**	110
Age	.51**
Verbal ability	.24**
Gender	−.17**
Emotion situation knowledge (part 2)		.50**	110
Age	.62**
Verbal ability	.25**
Gender	−.14
Expressive emotion knowledge (part 3)		.42**	110
Age	.50**
Verbal ability	.29**
Gender	−.16*
Receptive emotion knowledge (part 4)		.42**	110
Age	.47**
Verbal ability	.35**
Gender	−.08

*p < .05; **p < .01.

Concurrent validity

We conducted stepwise regression analyses to further examine the relation between the emotion knowledge parts and teacher-rated behavior in the domains of externalizing problems, internalizing problems, and adaptive abilities. Age, gender, and verbal ability were entered as control variables (see Table 3). Tolerance values and variance inflation factor were analyzed and they revealed no collinearity problems.

Table 3.

Stepwise multiple regression of EMT on children’s externalizing behavior and adaptive abilities

	ΔR²	β	TOTAL R² _ADJUSTED	N
Externalizing problems				109
Step 1	.09**		.08
Age		−.06
Verbal ability		−.12
Gender		.30**
Step 2	.04*		.12
Age		.03
Verbal ability		−.04
Gender		.26**
EMT part 1		.02
EMT part 2		.08
EMT part 3		.11
EMT part 4		−.21*
Adaptive abilities				109
Step 1	.30**		.28
Age		.37**
Verbal ability		.28**
Gender		−.17*
Step 2	.04*		.32
Age		.24*
Verbal ability		.21*
Gender		−.12
EMT part 1		.27*
EMT part 2		.12
EMT part 3		.08
EMT part 4		−.14

*p < .05; **p < .01.

EMT scores significantly predicted externalizing problems and adaptive abilities. Analyses indicated that gender and receptive emotion knowledge (EMT 4) were statistically significant predictors for externalizing problems, with boys scoring higher in externalizing problems and children with higher receptive emotion knowledge rated lower in externalizing problems. Results also revealed that expression matching (EMT 1) significantly and positively predicted teacher-rated adaptive abilities. The EMT scores did not predict teacher-rated internalizing problems.

Discussion

Literature supports that emotion knowledge may have an important role in social and emotional functioning in childhood, enabling a child to use the emotional information contained in social interactions (Bassett et al., 2012; Trentacosta & Fine, 2010). Reliable and valid measures of emotion knowledge are needed to study emotion knowledge development in young children (Zeidner et al., 2003), and they have to be adapted and validated for specific cultures. In the present work, we conducted a validation of the Emotion Matching Task (Izard et al., 2003) for Spanish preschool children. Although emotion knowledge has been extensively studied in America, less is known about emotion knowledge in other cultures (Cole & Tan, 2007; Friedlmeier, Corapci, & Cole, 2011).

The findings from the current study suggest that the EMT is an adequate tool to assess emotion knowledge in Spanish children. First, the EMT and the individual parts demonstrated good internal consistency. The relations found between emotion knowledge and both behavioral and emotional outcomes provide support about the EMT’s concurrent validity. As there are no other published tools to measure preschool emotion knowledge in Spain, we were unable to assess the EMT’s convergent validity; however, the EMT’s concurrent validity with established correlates of emotion knowledge provides initial evidence for an acceptable measure of preschool emotion knowledge (Morgan et al., 2010; Trentacosta & Fine, 2010).

We found that emotion knowledge was related to children’s adaptive abilities and externalizing problems as rated by a teacher. Expression matching positively predicted adaptive abilities. Furthermore, receptive emotion knowledge was negatively related to externalizing problems. The current study also explored individual differences in emotion knowledge, including age, verbal ability, and gender. As expected, age and verbal ability were highly related to all of the emotion knowledge dimensions. This result matches those found by Morgan et al. (2010) and previous research that demonstrates that older children and children with stronger verbal abilities have better developed emotion knowledge (Brown & Dunn, 1996; Denham, 1998; Denham et al., 2011; Miller et al., 2006; Trentacosta & Izard, 2007). The preschool period is one that involves significant development of children’s abilities to process emotional information, to assign a verbal label to an observed emotional expression, to associate similar emotional expressions, and to understand that certain situations may elicit specific emotions. Language abilities enhance children’s understanding of the emotional aspects of social interactions (Izard, Stark et al., 2008; Miller et al., 2006; Schultz, Izard, Ackerman, & Youngstrom, 2001; Trentacosta & Izard, 2007).

Gender differences in emotion knowledge have been less clear in previous literatures. The results from the current study suggest that girls had more developed emotion knowledge in the domains of expression matching and expressive emotion knowledge, as well as in total emotion knowledge. This result corresponds to previous findings about general emotion knowledge (Bennett et al., 2005; Brown & Dunn, 1996) and emotion recognition (McClure, 2000; Widen & Russell, 2010). This difference may be affected by different maturation of neurological structures, by differential emotional socialization in families, or by girls’ tendency to initiate and maintain more social interactions with adults (Brody & Hall, 2010; McClure, 2000).

The current study is the first to our knowledge to adapt an assessment tool to study preschool emotion knowledge in Spain. However, due to the lack of tools to measure emotion knowledge in Spanish contexts, convergent validity could not be assessed in the current study. However, it is important for future research to continue the development of emotion knowledge assessment tools for Spain and other cultures. It would also be worthwhile to continue to analyze the validity of the Spanish version of the EMT with other traditionally related constructs such as emotion regulation.

The Emotion Matching Task may be a useful tool for assessing the efficacy of a growing number of interventions aimed at developing emotional abilities in young Spanish-speaking children as a means of preventing behavioral and emotional problems (Denham & Burton, 2003; Domitrovich, Cortes, & Greenberg, 2007; Izard, 2002). Finally, findings from the current study may also help with promoting research on cultural differences in emotion knowledge development and outcomes (Cole & Tan, 2007; Friedlmeier et al., 2011).

Footnotes

Funding

This research project was supported in part by grant BF 109.346 from the Basque Government Department of Education, Universities and Research’s Training Program for Research Staff.

Appendix 1

References

Bassett

Denham

Mincic

Graling

(2012). The structure of preschoolers’ emotion knowledge: Model equivalence and validity using a structural equation modeling approach. Early Education and Development, 23, 259–279.

Bennett

D. S.

Bendersky

Lewis

(2005). Antecedents of emotion knowledge: Predictors of individual differences in young children. Cognition & Emotion, 19, 375–396.

Brody

L. R.

Hall

J. A.

(2010). Gender, emotion, and socialization. In Chrisler

J. C.

McCreary

D. R.

(Eds.), Handbook of gender research in psychology (pp. 429–454). New York, NY: Springer.

Brown

J. R.

Dunn

(1996). Continuities in emotion understanding from 3–6 years. Child Development, 67, 789–802.

Cole

P. M.

Tan

P. Z.

(2007). Emotion socialization from a cultural perspective. In Grusec

J. E.

Hastings

P. D.

(Eds.), Handbook of socialization: Theory and research (pp. 516–542). New York, NY: Guilford Press.

Denham

S. A.

(1998). Emotional development in young children. New York, NY: Guilford Press.

Denham

S. A.

Bassett

H. H.

Way

Mincic

Zinsser

Graling

(2011). Preschoolers’ emotion knowledge: Self-regulatory foundations, and predictions of early school success. Cognition & Emotion, 26, 667–679.

Denham

S. A.

Burton

(2003). Social and emotional prevention and intervention programming for preschoolers. New York, NY: Kluwer-Plenum.

Domitrovich

C. E.

Cortes

Greenberg

(2007). Improving young children’s social and emotional competence: A randomized trial of the preschool “PATHS” curriculum. The Journal of Primary Prevention, 28, 67–91.

10.

Dunn

Cutting

A. L.

(1999). Understanding others, and individual differences in friendship interactions in young children. Social Development, 8, 201–219.

11.

Dunn

L. M.

Dunn

L. M.

Arribas

(2006). PEABODY, Test de vocabulario en imágenes. Madrid, Spain: TEA.

12.

Ekman

(1993). Facial expression and emotion. American Psychologist, 48, 384–392.

13.

Friedlmeier

Corapci

Cole

P. M.

(2011). Emotion socialization in cross-cultural perspective. Social and Personality Psychology Compass, 5, 410–427.

14.

Hambleton

R. K.

Patsula

(1999). Increasing the validity of adapted tests: Myths to be avoided and guidelines for improving test adaptation practices. Journal of Applied Testing Technology, 1, 1–16.

15.

Izard

C. E.

(1977). Human emotions. New York, NY: Plenum.

16.

Izard

C. E.

(2001). Emotional intelligence or adaptive emotions? Emotion, 1, 249–257.

17.

Izard, C.E. (2002). Translating emotion theory and research into preventive interventions. Psychological Bulletin, 128, 796–824.

18.

Izard

C. E.

Haskins

F. W.

Schultz

Trentacosta

C. J.

King

K. A.

(2003). Emotion matching test. Unpublished test, Human Emotions Lab, Department of Psychology, University of Delaware, Delaware, USA.

19.

Izard, C. E., King, K. A., Trentacosta, C. J., Laurenceau, J. P., Morgan, J. K., Krauthamer-Ewing, E. S., & Finlon, K. J. (2008). Accelerating the development of emotion competence in Head Start children. Development & Psychopathology, 20, 369–397.

20.

Izard, C. E., Stark, K., Trentacosta, C., & Schultz, D. (2008). Beyond emotion regulation: Emotion utilization and adaptive functioning. Child Development Perspectives, 2, 156–163.

21.

Lewis

(2008). The emergence of human emotions. In Lewis

Haviland-Jones

J. M.

Feldman

(Eds.), Handbook of emotions (pp. 304–319). New York, NY: Guilford Press.

22.

McClure

E. B.

(2000). A meta-analytic review of sex differences in facial expression processing and their development in infants, children, and adolescents. Psychological Bulletin, 126, 424–453.

23.

Miller

A. L.

Fine

S. E.

Gouley

K. K.

Seifer

Dickstein

Shields

(2006). Showing and telling about emotions: Interrelations between facets of emotional competence and associations with classroom adjustment in Head Start preschoolers. Cognition & Emotion, 20, 1170–1192.

24.

Morgan

Izard

King

(2010). Construct validity of the emotion matching task: Preliminary evidence for convergent and criterion validity of a new emotion knowledge measure for young children. Social Development, 19, 52–70.

25.

Reynolds

C. R.

Kamphaus

R.W.

(2004). Sistema de evaluación de la conducta de niños y adolescentes. Madrid, Spain: TEA Ediciones.

26.

Schultz

Izard

C. E.

Ackerman

B. P.

Youngstrom

E. A.

(2001). Emotion knowledge in economically disadvantaged children: Self-regulatory antecedents and relations to social difficulties and withdrawal. Development and Psychopathology, 13, 53–67.

27.

Trentacosta

C. J.

Fine

S. E.

(2010). Emotion knowledge, social competence, and behavior problems in childhood and adolescence: A meta-analytic review. Social Development, 19, 1–29.

28.

Trentacosta

C. J.

Izard

C. E.

(2007). Kindergarten children’s emotion competence as a predictor of their academic competence in first grade. Emotion, 7, 77–88.

29.

Widen

S. C.

Russell

J. A.

(2010). Children’s scripts for social emotions: Causes and consequences are more central than are facial expressions. The British Journal of Developmental Psychology, 28, 565–581.

30.

Zeidner

Matthews

Roberts

R. D.

McCann

(2003). Development of emotional intelligence: Towards a multi-level investment model. Human Development, 46, 69–96.

31.

Zinbarg

R. E.

Revelle

Yovel

(2005). Cronbach’s alpha, Revelle’s beta, McDonald’s omega: Their relations with each other and two alternative conceptualizations of reliability. Psychometrika, 70, 123–133.

32.

Zumbo

B. D.

Gadermann

A. M.

Zeisser

(2007). Ordinal versions of coefficients alpha and theta for Likert rating scales. Journal of Modern Applied Statistical Methods, 6, 21–29.