J Abnorm Child Psychol (2017) 45:45–56

DOI 10.1007/s10802-016-0157-5

Emotion Knowledge and Attention Problems in Young Children:

a Cross-Lagged Panel Study on the Direction of Effects
Maria von Salisch 1 & Susanne A. Denham 2 & Tobias Koch 3

Published online: 28 April 2016

# Springer Science+Business Media New York 2016

Abstract Attention problems are likely to hinder children in of effects. How emotion knowledge impinges on attention
acquiring knowledge of their own and others’ emotions. problems is discussed.
Children with little knowledge of emotions tend to have dif-
ficulties with representing emotions, interpreting them, and Keywords Emotion knowledge . Attention problems . Verbal
sharing them, so that they are likely to spend more time in ability . Behavioral self-regulation . Kindergarten-age children
making sense of them and may thus appear to be inattentive.
In order to disentangle the direction of effects between emo-
tion knowledge and attention problems, 576 four- to- six-year- Young children’s attention problems tend to be persistent,
olds were interviewed at T1 and about 12 months later (T2) hinder them in their classroom engagement and their acqui-
about their emotion knowledge. Their kindergarten teachers sition of knowledge (Pagani et al. 2012), and often lead to
rated their attention problems, and their conduct problems at compromised outcomes in terms of academic achievement
T1 and T2. A cross-lagged panel model indicates that chil- (e.g., Duncan et al. 2007; Sjöwall et al. 2015a) and psychoso-
dren’s emotion knowledge at T1 contributed to the explana- cial adaptation (Kim and Deater-Deckard 2011). Difficulties in
tion of their attention problems at T2, after language ability paying or sustaining attention may diminish children’s ability
and attention problems at T1 were controlled. The other cross- to focus on relevant aspects of the situation, to benefit from
path from attention problems (T1) to emotion knowledge (T2) verbal input, and to retain knowledge in memory. These
was not significant. Adding gender, behavioral self-regula- attention-related difficulties are likely to extend to children’s
tion, working memory, conduct problems, or SES as alterna- acquisition of emotion knowledge (Denham et al. 2012;
tive explanations by third variables did not alter this direction Schultz et al. 2001; Trentacosta and Izard 2007). Emotion
knowledge includes the ability to recognize facial expressions
and situational antecedents of emotions as easier and the abil-
ity to understand mixed or concealed emotions as more diffi-
* Maria von Salisch
cult components in different theoretical models of children’s
[email protected]
emerging emotion knowledge (Bassett et al. 2012), but there
Susanne A. Denham may be other conceptualizations. In our view, emotion
[email protected] knowledge is related to emotion regulation in many ways (to
Tobias Koch
be discussed later), but does not include regulatory behaviors
[email protected] (Denham 1998).
However, the opposite direction of effects is also feasible:
1
Institute of Psychology, Leuphana University Lueneburg, Young children with little knowledge of their own and
Scharnhorststraße 1, D-21335 Lueneburg, Germany others’ emotions may have more difficulties in representing,
2
Department of Psychology MS 3F5, George Mason University, interpreting, and verbalizing emotions and may thus spend
4400 University Drive, Fairfax, VA 22030-4444, USA more time in making sense of them. Because emotions are
3
Center of Methods, Leuphana University Lueneburg, ever-present in kindergarten classrooms, teachers may con-
Scharnhorststraße 1, D-21335 Lueneburg, Germany sider these children inattentive and absorbed in their own
46 J Abnorm Child Psychol (2017) 45:45–56

thoughts (and accompanying emotions). Other caregivers emotional expressions and to connect emotions to typical sit-
are similarly likely to conclude that these children are having uational antecedents (e.g., Denham 1998). At the same time,
attention problems. This direction of effect is supported by a they begin to understand that a person’s wishes, memories,
study showing that first and second graders’ knowledge of and (sometimes mistaken) beliefs contribute to generating
emotions predicted their later ability to regulate their attention, his or her emotions (e.g., Harris 1989). Slightly later develop-
even after earlier attention ratings and other relevant covari- ments include an understanding of dissembled or concealed
ates had been taken into account (Trentacosta et al. 2006). The emotions and mixed emotions (e.g., Denham 1998; Harris
current study strives to disentangle the direction of effects 1989). Further components include knowledge of strategies
between attention problems and emotion knowledge in a het- of emotion regulation (e.g., Cole et al. 2008), as well as
erogeneous sample of four-to-six-year-olds, while controlling knowledge of the emotional consequences of moral trans-
for relevant third variables. gressions on the transgressor (e.g., Malti and Krettenauer
2013). A comprehensive measure of these nine components
of emotion knowledge, is provided by the Test of Emotion
Development of Attention and Attention Problems Comprehension (TEC; Pons and Harris 2002; Pons et al.
2002). Interindividual differences in young children’s emo-
Although even infants are capable of some voluntary control tion understanding are apparent from early on (Denham
of attention (Calkins and Marcovitch 2010), the conscious and 1998) and become increasingly stable (Pons and Harris
willful deployment of attention develops rapidly between 2 2005). Children’s advanced knowledge of emotions is pre-
and 5 years of age. According to the observation study by dictive of their academic and social success in kindergarten
Cole et al. (2011), young children’s ability to distract them- (e.g., Denham et al. 2012) and beyond (e.g., Trentacosta
selves in a frustrating waiting situation developed slowly in and Izard 2007). Thus, emotion knowledge contributes to
toddlerhood, and somewhat more rapidly after 36 months. children’s social, emotional, and academic adjustment.
With increasing age, children were better able to disengage
their attention away from a desirable gift and to start another
activity to occupy their minds. The ability to control attention The Attention Effect
at will, that is included in the construct of effortful control
(Eisenberg et al. 2013), seems to be important in children’s Because academic learning seems to be hindered by attention
regulation of anger and frustration (Cole et al. 2011), and in problems, acquiring knowledge about emotions may be equally
the reduction of externalizing problem behavior (Kim and impaired by inattention. This claim is supported by the following
Deater-Deckard 2011). clinical, cross-sectional, and longitudinal studies.

Attention Problems and Emotion Knowledge in at-Risk

Attention Problems Compromise Children’s and Clinical Samples A first line of evidence derives from
Learning studies of children who are at risk for or who show manifest
attention problems, usually in combination with hyperactivity.
Consistent interindividual differences in attention control are As predicted by Barkley’s (1997) model, boys who were at
apparent at age 5 and predict attention-related problems at age risk of attention deficit/hyperactivity disorders (ADHD) did
9 (Martin et al. 2012). In fact, attention problems may play a key not do as well as boys of a matched control group on gener-
role in the onset of learning difficulties over and above the ating emotion terms, connecting them to antecedent situations,
common comorbid conduct problems, because they limit chil- and understanding emotional states and emotion regulation,
dren’s ability to engage in learning activities (Pagani et al. 2012), even when their cognitive ability and their parents’ level of
to benefit from instruction, and to focus on educational tasks education had been controlled (Kats-Gold and Priel 2009).
(Arnold et al. 2012). According to the meta-analysis by Duncan Studies with children with clinical diagnoses of ADHD con-
et al. (2007), kindergarten children’s attention scores (which firm deficits in matching emotions to facial expressions and
included attention problems) predicted their later achievements antecedent situations (e.g., Da Fonseca et al. 2009). These
in mathematics and literacy, even after controlling for children’s clinical studies, however, did not rule out interindividual dif-
gender, SES, and respective pre-academic abilities. ferences in children’s executive function as an alternative ex-
planation, because deficits in working memory or in inhibitory
control that often accompany diagnoses of ADHD (e.g.,
Development of Emotion Knowledge Sjöwall et al. 2013) may also account for these findings. A
recent study that controlled for executive function and for
In the years leading up to school, children’s understanding of comorbid externalizing or internalizing problems corroborat-
emotions develops at a rapid pace. Children learn to label ed impairments in emotion recognition among children with
J Abnorm Child Psychol (2017) 45:45–56 47

ADHD (Sjöwall et al. 2013). These results highlight cross- knowledge in this domain. Because they have difficulties in
sectional associations between attention problems and deficits representing, verbalizing, and storing emotional events in
in emotion knowledge. They are limited, however, by the fact memory (Nelson and Fivush 2004), they may spend more
that no information is given on emotion knowledge prior to time in trying to determine (and to put into words) what they
the onset of ADHD or on the direction of effects. and their close associates are feeling, what situations may
have caused some emotions, and whether people are
Attention Problems and Emotion Knowledge in Typically concealing felt emotions. When confronted with emotional
Developing Children Cross-sectional studies with larger quandaries of this sort, young children with little emotion
samples of typically developing children concluded that knowledge may need to devote more attention (and perhaps
five- and six-year-olds’ emotion knowledge related to adult more working memory) to understanding the emotions of their
ratings of their concurrent attention to academic tasks social partners. Young children are highly motivated to do so,
(Trentacosta and Izard 2007, Trentacosta et al. 2006), and to because they depend on maintaining supportive relationships
tests of attention and attention problems (von Salisch et al. with close adults (Denham 1998). To their caretakers they
2013). More compelling evidence is provided by two longitu- may appear, however, to be more inattentive or absorbed than
dinal studies. One of them demonstrated that ratings of kin- their peers.
dergartners’ attention control predicted their recognition of Empirical support for the emotion knowledge effect comes
facial expressions and antecedent situations of emotions 1 year from a longitudinal study with a low income minority sample
later (Schultz et al. 2001) and the other study replicated it in which children’s emotion knowledge in preschool predicted
(Trentacosta and Izard 2007). Thus, multiple evidence sug- their attention ratings and test scores in kindergarten even after
gests that attention problems are related to later deficits in verbal ability and important social covariates had been taken
emotion understanding, but longitudinal studies that control into account (Rhoades et al. 2011). The longitudinal study by
for earlier emotion understanding are lacking. Trentacosta et al. (2006) corroborated that first and second
If attention problems hinder young children’s progress in graders with a better understanding of emotions received more
the domain of emotions, only longitudinal studies that take positive teacher ratings of their abilities to control their atten-
into account children’s earlier knowledge of emotions can tion some months later, even after their initial attention ratings
provide supporting evidence for this claim. We will call this (and the covariates age, gender, and verbal abilities) had been
first direction of effects, in which attention problems are re- controlled. Theoretical considerations and two longitudinal
sponsible for children’s slower progress in acquiring studies thus speak for the emotion knowledge effect, but no
knowledge about emotions, the attention effect. study so far has tested it and the attention effect in a single
model.

The Emotion Knowledge Effect

The Direction of Effects
Because neurobiological studies demonstrate close and
intertwined associations between the anterior attention and The first aim of the current study is to disentangle the direction
the emotion systems from infancy onward (Calkins and of effects between changes in young children’s inattention and
Marcovitch 2010), it is also feasible that children’s emotion changes in their emotion knowledge. This question can only
knowledge contributes to improvements in their ability to be answered in cross-lagged panel studies that take into ac-
focus, flexibly change, and sustain attention over time. An count initial emotion knowledge and initial attention problems
advanced understanding of emotions may help in amelio- when estimating the strength of the cross-lagged paths be-
rating children’s attention problems (above and beyond the tween earlier emotion knowledge and later attention problems
Bnormal^ age-related reduction in such problems). We will (i.e., the emotion knowledge effect) and earlier attention prob-
call this second direction of effects in which children’s lems and later emotion knowledge (i.e., the attention effect).
emotion knowledge impinges on their attention problems Key to this design is a comparison of the strength of the two
the emotion knowledge effect. cross-lagged paths. When both cross-paths are significant, bi-
Children with good emotion knowledge may be better at directional or reciprocal effects are assumed. We will venture
directing their attention at will because they may know when no prediction about the direction of effects, because some
to self-regulate. Children with little emotion knowledge, how- evidence speaks for the attention effect while other studies
ever, may experience an increased cognitive load when they support the emotion knowledge effect. Knowing the direction
encounter emotionally relevant situations (Hoehl et al. 2010). of effects will help in gaining a better understanding of how
In particular, the germane cognitive load that is devoted to the attention problems can be ameliorated and how emotion
processing, construction, and automation of emotional knowledge is acquired. Depending on whether the attention
schemas may be higher for them because they have less prior effect or the emotion effect (or both) are found, results from
48 J Abnorm Child Psychol (2017) 45:45–56

this study may assist in adding important components to early emotion knowledge (e.g., Cutting and Dunn 1999). Rater
interventions that target the promotion of emotion knowledge identity may influence results. In some Kindergartens, chil-
or attention. dren’s attention problems were rated by the same teacher at
It is well known that children’s verbal ability is related both T1 and T2, whereas in others they were rated by different
to their attention problems (e.g. Leonard et al. 2011) and to teachers. When the same teacher rates a child twice, his or
their emotion knowledge because language serves as means of her attitude may overshadow true changes in the child’s
representation that facilitates access to emotion knowledge attention-related behaviors. Finding no effect of rater identity
(Nelson and Fivush 2004). Because many studies support safeguards against source bias. Third variables were included
close associations between children’s emotion knowledge in order to avoid omitted variable bias in the rather sparse
and different components of their language abilities (e.g., cross-lagged models.
Pons et al. 2003), we included language ability in the basic
model. Because Trentacosta et al. (2006) demonstrated that
the relation between attention and emotion understanding
Aims
remained significant when controlling for verbal ability, we
do not expect it to affect the cross-lagged effects.
The first aim of the present study is to disentangle the direction
of effects between children’s attention problems and their
emotion knowledge in a cross-lagged panel study that controls
Excluding Alternative Explanations
for the well-known influence of language ability. The second
aim is to exclude alternative explanations and possible mod-
Third variables may explain or moderate the cross-lagged
erators of the cross-lagged associations between emotion
associations between emotion knowledge and inattention.
knowledge and attention problems that are provided by be-
Alternative explanations provided by behavioral self-regu-
havioral self-regulation, working memory, conduct problems,
lation, working memory, conduct problems, gender, and
gender, SES, and rater identity.
socio-economic status (SES) will be examined. Establishing
the robustness of the cross-lagged associations between
emotion knowledge and inattention is the second aim of this
study. Method
Behavioral self-regulation has been defined as Bdeliberately
applying multiple component processes of attentional or cog- Participants and Procedure
nitive flexibility, working memory, and inhibitory control to
overt, socially contextualized behaviors^ (McClelland et al. In the Emotional-Learning-is-Fantastic-study (Elefant-study,
2014, p. 2). This variety of executive function can be measured von Salisch et al. 2015a), a total of 265 children (133 girls
in a task that asks young children to inhibit their natural ten- and 132 boys) with complete data took part at T1 and again at
dency to follow the (misleading) instructions of the experi- T2 about 14 months later. Children’s mean age was M = 60.63
menter in their motor behavior (Cameron Ponitz et al. 2008). (5.65) months at T1 with a range of 44–74 months.
Children who are better in regulating their motor behavior at Participants came from 34 Kindergarten sites located in vil-
will tend to be more advanced in their emotion knowledge lages and towns in Northern Germany. These children were
(Denham et al. 2012; von Salisch et al. 2015b). Working part of a full sample of N = 576 children at T1. For the anal-
memory is an indicator of children’s executive functions and yses, we included information of the entire data set by using
thus related to attention problems. It is also associated with full information maximum likelihood (FIML) option imple-
emotion knowledge among preschool children (Blankson mented in Mplus 7.3.
et al. 2012). Conduct problems are related both to emotion Participation of the children and Kindergarten sites was
knowledge (Denham et al. 2002) and to attention problems voluntary. After a written parental permission was obtained,
(Sjöwall et al. 2013, Arnold et al. 2012). Gender differences children were tested individually by a trained interviewer in a
have been confirmed in that boys tend to have more attention quiet room at the Kindergarten site. Children were first
problems (e.g., Hölling et al. 2007) and are more often diag- interviewed about their emotion understanding (T1 and T2),
nosed with ADHD than girls (e.g., American Psychiatric then tested for their behavioral self-regulation (T1) and in a
Association 2000). Some evidence suggests that boys’ second visit (within 2 months) for their language ability. In
emotion knowledge is generally less advanced than girls’ addition, 96 Kindergarten teachers were asked about the chil-
(e.g., Denham et al. 2012). SES may affect both attention dren’s socio-demographic background (T1) and rated their
problems and emotion knowledge. Growing up in a house- behavior (T1 and T2). Children’s parents were asked about
hold with low SES tends to be associated with having their socio-demographic background at T2. 200 parents
more attention problems (e.g., Hölling et al. 2007), and less returned useable answers.
J Abnorm Child Psychol (2017) 45:45–56 49

Measures Strength-and-Difficulties-Questionnaire (SDQ; Goodman

1997). Five items could be answered with not true (0), partly
Emotion Knowledge Children’s emotion knowledge was true (1), or true (2). In order to differentiate between attention
assessed at T1 and T2 with the German version of the Test problems and hyperactivity, only the 3 items indicating atten-
of Emotion Comprehension (TEC; Pons et al. 2002, Janke tion problems on the SDQ scale (Beasily distracted, concen-
2008). The TEC is a mastery test which is based on Harris tration wanders^, Bthinks things out before acting (R)^, and
(1989) account of the development of emotion knowledge. Bgood attention span, sees work through to the endB) were
It consists of nine components. Component 1 measures used (whereas the two items indicating motoric overactivity
recognition of facial expressions for five basic emotions. were not used). Means were computed with higher values
Component 2 asks children to match emotional expres- indicating more problem behavior. Internal consistencies were
sions to five typical situational elicitors, such as happiness alpha = 0.76 (T1) and 0.80 (T2) for attention problems, and
when receiving a birthday gift. In component 3, children alpha = 0.71 for conduct problems. For attention problems,
are asked to attribute a character’s emotions depending on 11.4 % (T1) and 7.6 % (T2) scored at or above the cutoff of 1.4
whether the character’s food preferences (for Cola and in the clinical range, whereas another 8.7 % (T1) and 5.7 %
salad) are fulfilled or not. Component 4 calls upon emo- (T2) scored in the borderline range. For conduct problems,
tional perspective taking in a task that allows the child to 7.7 % scored in the clinical and another 11.1 % in the border-
have more information than the protagonist. This is a typical line range. The SDQ is validated with parents’ reports and
theory of mind task of the category Bprivileged information^. DSM-IV diagnostic interviews (e.g., Hawes and Dadds 2004).
Children’s theory of mind is again under scrutiny in com-
ponent 5, when they are asked which emotion the memory Language Ability Children’s receptive language ability was
(of the protagonist’s dead pet) may cause in him or her. In tested with a subtest of the German Language Development
component 6, children are asked about strategies of emotion Test for 3- to 5-year-old children (SETK 3–5; Grimm 2001). It
regulation in a sadness-eliciting situation. Component 7 asks children to enact increasingly complex sentences with
examines display rules, in particular the necessity for material given to them. Children’s correct enactments were
concealing a felt emotion in order to reach a goal. Children’s summed (0–15 points). Internal consistency was satisfactory
understanding of mixed emotions (e.g., both positive and neg- (Cronbach’s alpha = 0.79). The SETK 3–5 has shown good
ative feelings at the same time) is examined in component 8. predictive validity (Kiese-Himmel et al. 2012).
The ninth component elicits children’s understanding of moral
emotions after a transgression. Each component was scored as Behavioral Self-Regulation Behavioral self-regulation was
0 (not mastered) or 1 (mastered). measured with the Head-Toes-Knees-Shoulders-Task
The TEC was presented in a picture book with a male (HTKS; Cameron Ponitz et al. 2008). The HTKS consists of
or female child protagonist matched to the gender of the two training phases and two subsequent testing phases. In the
child. The interviewer read a story about the protagonist. beginning, children practiced doing the opposite of what the
Afterwards, the child made an emotional attribution by experimenter does (Bwhen I ask you to touch your head, you
selecting the most appropriate of four possible emotional touch your toes^ and Bwhen I ask you to touch your toes, you
outcomes. Children could obtain between 0 and 9 points touch your head^). This first rule was tested in a block of 10
depending on how many components they mastered. items. Afterwards children were asked to practice a second
Previous studies reported modest to good coefficients for rule: touching their shoulders instead of their knees and touch-
internal reliability and validity of the TEC (Pons et al. ing their knees instead of their shoulders. Adherence to both
2002). In the present study, internal consistencies were rules was tested in a second block of 10 trials. Children’s
low (with alpha = 0.63 at T1 and alpha = 0.62 at T2). correct reactions (2 points) and self-corrections (1 point) were
The TEC measures a very broad construct which tends to tallied (0–40 points). Internal consistency was excellent with
lower internal consistency. In fact, the TEC is a mastery test, alpha = 0.95. HTKS tends to predict children’s academic
which contains easier components (e.g. emotion recognition) achievements (McClelland et al. 2007).
that are passed earlier and more difficult components (e.g.
mixed emotions) that are passed later in development. The Working Memory Phonological working memory was
TEC is correlated with children’s language skills (Pons and measured with the subtest on number recall from the
Harris 2005) and their IQ (von Salisch et al. 2013) and relates Neuropsychological Test Battery for Number Processing and
to their social competence (Janke 2008). Calculation for Children-Kindergarten Version (ZAREKI-K)
(von Aster et al. 2009). In this forward digit span, the child
Ratings of Attention Problems and Conduct Problems repeats a sequence of numbers after the examiner. Internal
Kindergarten teachers were asked to rate each child’s hyper- consistency of the published version was alpha = 0.73. In
activity and attention problems and conduct problems on the the Elefant study, internal consistency was satisfactory with
50 J Abnorm Child Psychol (2017) 45:45–56

alpha = 0.75. Children’s raw scores were transformed into estimator implemented in Mplus and computed robust fit
T-values so that the mean was 50 and the standard devi- statistics (Asparouhov 2005).
ation was 10. Prognostic validity is excellent: Children’s
scores on the ZAREKI-K in Kindergarten identified
68.5 % of the children that were later diagnosed with
Results
dyscalculia (von Aster et al. 2009).
Descriptive statistics and intercorrelations of the measures are
Socio-Economic Status Kindergarten teachers were asked
presented in Table 1.
about the occupation of each child’s mother and father. In
The means presented in Table 1 show that children gained
addition, parents were asked about their highest educational
new knowledge of emotions from T1 to T2, whereas their
degree, their initial job training, and their current position.
attention problems decreased. Intercorrelations indicate that
Parents’ answers (and when missing, teachers’ answers) were
attention problems at T2 (but not at T1) were negatively relat-
coded according to the International Socio-Economic Index of
ed to emotion knowledge at T1 and T2. Attention problems
Occupational Status (ISEI) (Ganzeboom et al. 1992). The
were negatively associated with SES, language ability, behav-
highest ISEI-status in the family was used, independent of
ioral self-regulation, and working memory, and positively
whether it was the father’s or the mother’s ISEI. Highest
related to conduct problems. Emotion knowledge was cor-
ISEI (HISEI) had a mean of 55.80, a SD of 17.72, and varied
related positively with language ability and behavioral self-
between 17 and 88 (with a possible range between 16 and 90).
regulation. As in many other studies, boys outscored girls
The sample was socio-economically diverse with a large
in problems with attention and conduct. Next, we present
proportion of families in the middle of the bell-shaped
the findings concerning the direction of effects (question 1)
distribution.
and the third-variable influences on the cross-lagged associa-
tions (question 2).
Rater Identity One hundred fifty-one children were rated by
the same teacher at T1 and T2, while 114 were rated by dif-
ferent teachers. Rater identity was scored to be true (1), or not Aim 1: Direction of Effects between Emotion Knowledge
true (0). and Attention Problems

Data Analysis Figure 1 contains a graphical representation of the cross-

lagged path model on the relations between emotion
To examine the direction of effects, we specified a cross- knowledge and attention problems when language ability
lagged panel model, which allows for the simultaneous pre- was controlled. On the left hand side, this model shows the
diction (a) of later emotion knowledge from earlier attention cross-sectional correlations between emotion knowledge and
problems and (b) of later attention problems from earlier emo- language ability, and language ability and attention problems
tion knowledge. Please note that this model comprises (a) the at T1. Stability estimates (autoregressive effects), which are
longitudinal (cross-lagged) regression between earlier and lat- indicated on the horizontal lines, were substantial for attention
er indicators, (b) the stability of the indicators over time problems, and less so, but still highly significant for emotion
(autoregressive effects), (c) the contemporaneous correlation knowledge.
at T1 and T2, and (d) language ability as a control variable for Among the two cross-lagged paths, only the parameter
known effects on emotion knowledge at T2. Cross-lagged estimate between children’s emotion knowledge at T1
panel models allow researchers to examine to the amount of and their attention problems at T2 showed a significant
total variance of the outcome variables that can be explained negative relationship. Four- to six-year-olds who were more
by the cross-lagged associations over and above the influence knowledgeable about emotions at T1 had fewer attention
of the longitudinal stabilities (as well as the influences of problems at T2 (with control of the attention problems they
language ability). had manifested at T1 and the correlation between emotion
In addition, we included third variables (covariates) in the knowledge and attention problems at T1) than did their less
cross-lagged model in order to test whether the effects knowledgeable age-mates. Children with higher initial levels
remained invariant when controlling for gender, rater identity, of emotion knowledge improved in teacher-ratings of atten-
behavioral self-regulation, socio-economic status, working tion over time. The cross-lagged path between attention prob-
memory, and conduct problems. Main and interaction lems at T1 and emotion knowledge at T2 was near zero.
effects of the third variables were tested in separate The model fitted perfectly (Hu and Bentler 1999), even
cross-lagged models (six models in total) to avoid loss of when fixing the cross-lagged path from attention problems
statistical power. Because some variables were non-normally T1 to emotion knowledge T2 and the cross-lagged path
distributed, we used the robust maximum likelihood (MLR) from language ability to emotion knowledge T2 to zero,
J Abnorm Child Psychol (2017) 45:45–56 51

Table 1 Means (M), standard deviations (SD), ranges, and intercorrelations of the variables

1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11)

1) Emotion knowledge T1
2) Emotion knowledge T2 0.41
3) TR Attention problems T1 −0.10 −0.06
4)TR Attention problems T2 −0.28 −0.23 0.55
5) Language ability T1 0.33 0.28 −0.24 −0.24
6)TR Conduct problems T1 −0.06 0.00 0.50 0.40 −0.11
7) Working memory T2 0.34 0.23 −0.24 −0.32 0.32 −0.10
8) Gender 0.04 −0.01 −0.27 −0.35 0.00 −0.21 0.01
9) Rater identity 0.11 0.06 −0.07 −0.10 0.06 −0.04 0.10 0.05
10) SES −0.01 −0.02 −0.15 −0.14 0.14 −0.06 0.05 0.01 −0.01
11) Behavioral self-regulation T1 0.45 0.29 −0.29 −0.31 0.37 −0.11 0.38 0.11 0.14 0.02
M 4.15 5.74 0.25 0.19 10.60 0.34 54.43 0.00 1.40 58.55 25.88
SD 5.35 1.49 0.17 0.17 8.63 0.11 8.37 1.0 0.49 19.70 12.05
Range 0–8 2–9 0–0.67 0–0.67 1–15 0–1.5 30–74 −1-1 1–2 17–88 0–40

N = 576 with FIML

Values above r = 0.12, p < 0.05, values above r = 0.21, p < 0.000

χ2 = 0.778(2), p = 0.678, CFI = 1.000, RMSEA = 0.000, socio-economic status, and working memory. We first tested
SRMR = 0.008. the interaction effects of each single covariate using separate
models (Model 1–6). Based on these results, we specified a
Aim 2: Third Variable Influences on the Direction full model that included all covariates (Model 7). Multi-group
of Effects models were used for testing the interaction effects with re-
gard to the categorical covariates (i.e., sex and rater identity).
In order to safeguard against omitted variable bias, we tested Single-group models were used for the continuous covariates.
in Table 2 whether the cross-lagged path coefficients were Overall, we found no significant interaction effects with re-
moderated by one or more of the following covariates: sex, gard to the above covariates, except for a significant interac-
rater identity, conduct problems, behavioral self-regulation, tion between conduct problems at T1 and emotion knowledge

Fig. 1 Cross-lagged path model

on emotion knowledge and
attention problems with language
ability as covariate
52 J Abnorm Child Psychol (2017) 45:45–56

Table 2 Cross-lagged path

coefficients when including Model including third variables Emotion knowledge Attention problems
different third variables - > attention problems with third - > emotion knowledge with third
variable control variable control
Standardized Standardized
Path Coefficient Path Coefficient
a
Model 1: SES −0.220 (0.042)*** −0.008 (0.049)
b
Model 2: Gender −0.239 (0.050)*** −0.020 (0.056)
a
Model 3: Behavioral self-regulation −0.185 (0.046)*** 0.033 (0.050)
a
Model 4: Working memory −0.183 (0.043)*** 0.036 (0.049)
a
Model 5: Conduct problems −0.213 (0.042)*** −0.007 (0.056)
b
Model 6: Rater identity −0.235 (0.049)*** −0.020 (0.056)
a
Model 7: Including all third variables −0.161 (0.045)*** 0.016 (0.056)
a
Model included interaction terms between the third variable and the key variables measured at T1
b
Multigroup analysis was conducted for the categorical covariates, assuming that all coefficients were identical in
the groups. The fit of the models were excellent. Standard errors are provided in parentheses
***
: p ≤ 0.000

at T1 on the development of attention problems at T3 to the knowledge of emotions contributed to explaining the reduced
effect that conduct problems attenuated the effect of emotion teacher-perceptions of attention problems beyond the general
knowledge at T1 on attention problems at T2. This interaction reduction that is typical in this age group and add that this was
effect was, however, only marginally significant (p = 0.071) the case even after the other cross-path (and language ability
when it was included in the full model. As Table 2 shows, the as a control variable) were included in the model. Emotion
attention effect was always negligible and did not change knowledge may be one mechanism that intervention re-
when any of the third variables were included. Children who searchers may wish to consider when they aim to improve
understood more about emotions tended to improve in their children’s academic and social functioning through improving
teacher-rated attention over time, independent of their gender, their attention (Bunford et al. 2015a). Perhaps, emotion
their parents’ SES, their behavioral self-regulation, their work- knowledge could be a primary treatment target for attention
ing memory, and possible conduct problems. Whether the problems. Future research will show whether emotion
same or a different teacher at T1 and T2 rated children’s be- knowledge improves not only children’s attention problems
havior, also made no contribution. These factors cannot be but also their academic and social impairment through attention
used to explain our finding that children who were initially problems. Furthering children’s emotion knowledge is also ex-
more advanced in their emotion knowledge tended to show pected to have direct positive effects on their social functioning
fewer attention problems over time. because correct emotion processing is vital for social informa-
tion processing and competent behavior (Denham 1998).

Discussion How Does Emotion Knowledge Influence Attention

Problems?
The Direction of Effects
When children are able to verbally label their sometimes dif-
The Elefant study confirmed the emotion knowledge effect on fuse, ambiguous, or upsetting emotions, they are less preoc-
attention in a cross-lagged panel model. Theoretical accounts cupied and distracted by them, and can pay better attention to
(e.g., Blair and Dennis 2010), longitudinal studies (e.g. events in the Bouter world^. This includes the emotions of
Denham et al. 2012, Duncan et al. 2007), and intervention their social partners and academic tasks, among many other
studies (e.g., Jones et al. 2011) all agree that control over things. Awareness of their own emotions can help children in
attention is an important mediator for social and academic focusing their attention (e.g., through self-talk; Riggs et al.
success in childhood and beyond. Earlier studies pointed to 2006).
emotion knowledge as contributing to children’s growing abil- Being less aware of their own emotions and difficulties in
ity to willfully control their attention in academic settings identifying felt emotions is characteristic of adolescents with
(Rhoades et al. 2011; Trentacosta et al. 2006), but the ADHD (Bunford et al. 2014). Children who are unable to put
Elefant study is the first study that tested both directions of their feelings into words are likely to produce misunderstand-
effect in a single model. Results underscore that advanced ings with peers and adults because their social partners do not
J Abnorm Child Psychol (2017) 45:45–56 53

understand why the child is acting in this way. These misun- direct their attention at will and to inhibit interfering stimuli
derstandings may further occupy their minds. Young children (in a Stroop test). The PATHS intervention resulted in less
who can verbalize their emotions are also at an advantage externalizing problem behavior (which includes attention
because it will be easier for them to find caretakers and friends problems and hyperactivity) than in the control group (Riggs
to help them in making sense of their feelings and in amelio- et al. 2006). A meta-analysis over 213 studies confirmed that
rating them when necessary, so that the youngsters can take interventions on SEL contributed to improving school-chil-
their mind off them. Children’s knowledge of their own emo- dren’s self-management, including paying attention when
tions thus facilitates different means of emotion regulation asked to do so (Durlak et al. 2011). Thus, lower attention
(Denham 1998). problems are frequent outcomes of interventions that include
Not knowing which emotions their social partners typically lessons on emotion knowledge. However impressive these
feel and show, may add cognitive load to children’s working interventions are in terms of sample size, they are limited by
memory (Hoehl et al. 2010). It makes it harder for them to the fact that it is not clear whether emotion knowledge or
disregard emotionally arousing scenes, to accurately appraise another factor in their array of treatments is responsible for
(and reappraise) their partners’ intentions, and to modulate the reduction of children’s attention problems.
their behavior in socially acceptable ways, to mention just a
few ways of emotion regulation (Gross 1998) that may be The Attention Effect
influenced by emotion knowledge. Lacking knowledge of
other peoples’ emotions thus impairs youngsters’ emotion That attention problems did not predict gains in emotion
regulation and is likely to cause social problems (Bunford knowledge may have to do with measurement. Measuring
et al. 2014) that may further absorb their attention. attention problems by teacher ratings has the advantage of
Longitudinal studies underscore that more hostile attributions ecological validity, but does not map well onto test measures
of intent (Choe et al. 2013), more intensive expressions of of attention that are used in some of the studies of executive
anger (Trentacosta et al. 2006), more aggressive behavior with function and learning mentioned in the introduction (e.g.
peers (Denham et al. 2002), less productive relationships with Sjöwall et al. 2015b). In our Elefant study, concurrent corre-
teachers and peers (Mostow et al. 2002), and compromised lations between teacher ratings and test measures of attention
academic achievements (Denham et al. 2012; Rhoades et al. at T2 were significant at p < 0.001, but did not exceed 0.27
2011; Trentacosta and Izard 2007) are explained by deficits in (von Salisch et al. 2013). Attention tests measure the efficacy
young children’s emotion knowledge. of processes that are available to children for behavioral con-
Furthermore, emotion knowledge includes knowing trol in order to achieve an optimal performance in the presence
about strategies of emotion regulation (Cole et al. 2008). of an experimenter, whereas attention ratings tend to capture
Better regulated emotions make fewer demands on chil- their typical use of attention in everyday situations in implicit
dren’s attentional capacities. A recent study showed that comparison with their age mates (Toplak et al. 2013).
one variant of emotion dysregulation in preschool predicted
inattention at age 18 after controlling for earlier ADHD The Emotion Knowledge Effect Is Robust
symptoms (Sjöwall et al. 2015b). Cross-sectional associa- against Alternative Explanations
tions between impairments of emotion regulation and
symptoms of inattention are confirmed in two studies A strong point of the Elefant study lies in the examination of
(Bunford et al. 2014, Sjöwall et al. 2015b) and held even after third variables that may explain the longitudinal associations
controlling for executive function. between emotion understanding and attention problems.
Although boys were rated to be more inattentive than girls,
Intervention Studies as in many other studies (e.g. Hölling et al. 2007), there were
no differences between boys and girls in the direction of the
Further evidence on the attenuating influence of children’s cross-lagged effects. Living in a family with high or low SES
emotion knowledge on attention problems comes from inter- did not influence the longitudinal evidence of emotion
ventions on social-emotional learning (SEL) that usually con- knowledge contributing to the reduction of attention problems
tain lessons on some components of emotion knowledge. over time. In regard to child characteristics, neither a test of
Improvements of children’s attention problems have been their behavioral self-regulation nor of their working memory
found, for example, in a 2-year-intervention in New York changed the significant effect of emotion knowledge on later
City primary schools that yielded lower teacher reports of attention problems and the non-significant effect of inattention
ADHD-symptoms and lower child reports of hostile attribu- on later emotion knowledge. The non-significant results for
tional biases (Jones et al. 2011). Analyses of primary school behavioral self-regulation and working memory underscore
children showed that the Promoting Alternative THinking that interindividual differences in these components of chil-
Strategies (PATHS) curriculum improved their abilities to dren’s executive functions can be excluded as an explanation
54 J Abnorm Child Psychol (2017) 45:45–56

for the emotion knowledge effect. Finding no effect of behav- attention problems, is a predictor for later academic success
ioral self-regulation on the emotion knowledge effect further (Rhoades et al. 2011; Trentacosta and Izard 2007). In the years
suggests that deficits in emotion processing among children leading up to school, both emotion knowledge and regulation
with attention problems may be impaired above and beyond of attention grow at a rapid pace. Future studies with more
behavioral inhibition (Barkley 1997). Adding conduct prob- measurement points should look into the mechanism that
lems did not alter the direction of effects, either. This result translate children’s knowledge of their own and others’ emo-
safeguards against the interpretation that children’s acquisition tions into better control of their attention (and their emotions).
of (emotion) knowledge is compromised by the conduct prob- Future studies need to examine which influences strengthen or
lems that often accompany attention problems (Arnold et al. dampen the positive effect of emotion understanding on ame-
2012, Bunford et al. 2014). It can thus be ruled out that the liorating the widespread attention problems among young
reduced ratings of attention problems among children who children.
were initially more knowledgeable about emotions were due
to a range of factors extraneous to the always rather lean cross- Acknowledgments We would like to thank the kindergarten teachers
and children for their patience in answering our many questions and Dr.
lagged panel models.
Julie Klinkhammer, Dr. Martha Hänel, Uta Kraft, and our student inter-
viewers for asking them. We also thank Dr. Jens Vogelgesang for statis-
Strengths and Limitations of the Current Study tical advice.

The main strength of the Elefant study lies in the large and Compliance with Ethical Standards
heterogeneous sample drawn from the hitherto understudied
Funding This study was funded by the Ministries of Science and of
population of children in German Kindergarten. Another Education of the land of Lower Saxony that was handed out to
strength lies in the cross-lagged panel design that makes it the Research Association for Early Childhood Education and
possible to examine the direction of effects over time and Development Lower Saxony (Forschungsverbund Frühkindliche
the (lacking) influence of third variables that strongly suggest Bildung und Entwicklung Niedersachsen).
causal relations. It is important to recognize the limitations of Conflict of Interest Maria von Salisch declares that she has no conflict
this research. First, only four- to six-year-olds were examined. of interest, Susanne Denham declares that she has no conflict of interest,
Future research should extend the age range in order to find and Tobias Koch declares that he has no conflict of interest.
out whether emotion understanding contributes to the reduc-
Ethical Approval All procedures performed involving human partici-
tion not only of attention problems, but also of other executive pants were in accordance with the ethical standards of the institutional
functions in older and especially in younger age-groups (as research committee and with the 1964 Helsinki declaration and its later
suggested by Blankson et al. 2012). Second, more differenti- amendments or comparable ethical standards.
ated results about the different components of emotion
Informed Consent Informed consent was obtained from all individual
knowledge can be obtained with improved measures of emo- children, their parents, and their kindergarten teachers included in the
tion understanding. Third, evidence for the direction of effects study.
can be strengthened with more than two points of measure-
ment in the cross-lagged models. Fourth, other third variables
that are associated both with emotion knowledge and with
attention problems can be used to further probe the robustness
of the emotion knowledge effect. Likely candidates are inter- References
personal factors, like emotionally invalidating parenting
(Bunford et al. 2015b), and household chaos (Raver et al. American Psychiatric Association (2000). Diagnostic and Statistical
Manual of Mental Disorders, 4th edition, Text Revision (DSM-IV-
2014). Fifth, experimental work is needed to test hypotheses
TR). Washington, DC: American Psychiatric Association.
on the influence of children’s emotion knowledge on their Arnold, D. H., Kupersmidt, J. B., Voegler-Lee, M. E., & Marshall, N. A.
online processing of emotions. Sixth, results need to be repli- (2012). The association between preschool children’s social function-
cated, preferably in other cultures. Seventh, multilevel analy- ing and their emergent academic skills. Early Childhood Research
sis insures against effects of shared classroom membership Quarterly, 27, 376–386. doi:10.1016/jecresq.2011.12.009.
Asparouhov, T. (2005). Sampling weights in latent variable modeling.
that tends to inflate individual-level associations.
Structural Equation Modeling, 12, 411–434.
Barkley, R. A. (1997). Behavioral inhibition, sustained attention, and exec-
utive functions: constructing a unifying theory of ADHD.
Conclusions Psychological Bulletin, 121, 65–94. doi:10.1037/0033-2909.121.1.65.
Bassett, H. H., Denham, S. A., Mincic, M., & Graling, K. (2012). The
structure of preschoolers’ emotion knowledge: Model equivalence
The present study is an initial step in examining the direction and validity using a structural equation modeling approach. Early
of effects between emotion understanding and attention prob- Education and Development, 23, 259–279. doi:10.1080/10409289.
lems. Emotion knowledge, both alone and in conjunction with 2012.630825.
J Abnorm Child Psychol (2017) 45:45–56 55

Blair, C., & Dennis, T. (2010). An optimal balance: the integration of Eisenberg, N., Edwards, A., Spinrad, T. C., Sallquist, J., Eggum, N. D., &
emotion and cognition in context. In S. Calkins & M. A. Bell Reiser, M. (2013). Are effortful and reactive control unique con-
(Eds.), Child development at the intersection of emotion and structs in young children? Developmental Psychology, 49, 2082–
cognition (pp. 17–35). Washington, DC: APA. 2094. doi:10.1037/a0031745.
Blankson, A. N., O’Brien, M. O., Leerkes, E. M., Marcovitch, S., Calkins, Ganzeboom, H. B. G., DeGraaf, P. M., & Treiman, D. J. (1992). A
S., & Weaver, J. M. (2012). Developmental dynamics of emotion and standard international socio-economic index of occupational status.
cognition processes in preschoolers. Child Development, 84, 346– Social Science Research, 21, 1–56.
360. doi:10.1111/j.1467-8624.2012.01841.x. Goodman, R. (1997). The strengths and difficulties questionnaire: a re-
Bunford, N., Evans, S. W., & Langberg, J. M. (2014). Emotion dysregu- search note. Journal of Child Psychology and Psychiatry, 38, 581–
lation is associated with social impairment among young adoles- 586. doi:10.1111/j.1469-610.1997.tb01545.x.
cents with ADHD. Journal of Attention Disorders, 1–17. doi:10. Grimm, H. (2001). SETK 3–5: Sprachentwicklungstest für drei- bis
1177/1087054714527793. fünfjährige Kinder [Test of language development for 3-to-5-year
Bunford, N., Brandt, N. E., Golden, C., Dykstra, J. B., Suhr, J. A., & old children] (2nd ed.). Göttingen: Hogrefe.
Owens, J. S. (2015a). Attention-deficit/hyperactivity disorder symp- Gross, J. J. (1998). The emerging field of emotion regulation: an integra-
toms mediate the association between deficits in executive function- tive review. Review of General Psychology, 2, 271–299.
ing and social impairment in children. Journal of Abnormal Child Harris, P. L. (1989). Children and emotion: The development of psycho-
Psychology, 43, 133–147. doi:10.1007/s10802-014-9902-9. logical understanding. Oxford: Blackwell.
Bunford, N., Evans, S. W., & Wymbs, F. (2015b). ADHD and emotion Hawes, D. J., & Dadds, M. R. (2004). Australian data and psychometric
dysregulation among children and adolescents. Clinical Child and properties of the strengths and difficulties questionnaire. Australian
Family Psychology Review, 18, 185–217. doi:10.1007/s10567-015- and New Zealand Journal of Psychiatry, 38, 644–651.
0187-5. Hoehl, S., Brauer, J., Brasse, G., Striano, T., & Frederici, A. (2010).
Calkins, S. D., & Marcovitch, S. (2010). Emotion regulation and execu- Children’s processing of emotions expressed by peers and adults:
tive functioning in early development: integrated mechanisms of an fMRI study. Social Neuroscience, 5, 543–559.
control supporting adaptive functioning. In S. D. Calkins & M. A. Hölling, H., Erhardt, M., Ravens-Sieberer, U., & Schlack, R. (2007).
Bell (Eds.), Child development at the intersection of emotion and Verhaltensauffälligkeiten bei Kindern und Jugendlichen. Erste
cognition (pp. 59–78). Washington: APA. Ergebnisse aus dem Kinder- und Jugendgesundheitssurvey
Cameron Ponitz, C. C., McClelland, M. M., Jewkes, A. M., McDonald (KiGGS) [Problem behavior among children and adolescents. First
Connor, C., Farris, C. L., & Morrison, F. J. (2008). Touch your toes! results from the child and youth survey]. Bundesgesundheitsblatt -
Developing a direct measure of behavioral regulation in early child- Gesundheitsforschung - Gesundheitsschutz, 50, 784–793.
hood. Early Childhood Research Quarterly, 23, 141–158. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covari-
Choe, E. N., Lane, J. D., Grabell, A. S., & Olson, S. L. (2013). ance structure analysis: conventional criteria versus new alterna-
Developmental precursors of young school-age children's hostile tives. Structural Equation Modeling, 6, 1–55.
attribution bias. Developmental Psychology, 49, 2245–2256. doi: Janke, B. (2008). Emotionswissen und soziale Kompetenz [emotion
10.1037/a0032293. knowledge and social competence]. Zeitschrift für empirische
Cole, P. M., Dennis, T. A., Smith-Simon, K. E., & Cohen, L. H. (2008). Pädagogik, 22, 127–144.
Preschoolers’ emotion understanding: relations with emotion social- Jones, S., Brown, J. L., & Aber, L. (2011). Two-year impact of a universal
ization and child self-regulation. Social Development, 18, 324–352. school-based social-emotional and literacy intervention: an experi-
Cole, P. M., Tan, P. Z., Hall, S. E., Zhang, Y., Crnic, K. A., Blair, C. B., & ment in translational developmental research. Child Development,
Li, R. (2011). Developmental changes in anger expression and at- 82, 533–554.
tention focus: learning to wait. Developmental Psychology, 47, Kats-Gold, I., & Priel, B. (2009). Emotion, understanding, and social
1078–1089. skills among boys at risk of attention deficit hyperactivity disorder.
Cutting, A., & Dunn, J. (1999). Theory of mind, emotion understanding, Psychology in the Schools, 46, 658–678. doi:10.1002/pits.20406.
language, and family background: individual differences and inter- Kiese-Himmel, C., Auberlen, S., & von Steinbüchel, N. (2012). Ausgewählte
relations. Child Development, 70, 853–865. Sprachentwicklungsstandsfacetten von Kindergartenkindern mit
Da Fonseca, D., Seguier, V., Santos, A., Poinso, F., & Deruelle, C. (2009). Migrationshintergrund in Deutschland. [Selected facets of
Emotion understanding in children with ADHD. Child Psychiatry language development status in German kindergarten children
and Human Development, 40, 111–121. doi:10.1007/s10578-008- with immigration background]. Zeitschrift für Medizinische
0114-9. Psychologie, 21, 8088.
Denham, S. A. (1998). Emotional development in young children. New Kim, J., & Deater-Deckard, K. (2011). Dynamic changes in anger, exter-
York: Guilford. nalizing, and internalizing problems: attention and regulation.
Denham, S. A., Caverly, S., Schmidt, M., Blair, K., DeMulder, E., Caal, Journal of Child Psychology and Psychiatry, 52, 156–166. doi:10.
S., et al. (2002). Preschool understanding of emotions: contributions 1111/j.1469-7610.2010.02301.x.
to classroom anger and aggression. Journal of Child Psychology and Leonard, M., Milich, R., & Lorch, E. P. (2011). The role of pragmatic
Psychiatry, 43, 901–916. language use in mediating the role between hyperactivity and social
Denham, S. A., Bassett, H. H., Way, E. E., Mincic, M., Zinsser, K., & skills problems. Journal of Speech, Language, and Hearing
Graling, K. B. (2012). Preschoolers’ emotion knowledge: self- Research, 54, 567–579.
regulatory foundations, and predictions of early school success. Malti, T., & Krettenauer, T. (2013). The relation of moral emotion
Cognition and Emotion, 26, 667–679. attributions to prosocial and antisocial behavior: a meta-analysis.
Duncan, G. J., Dowsett, C. J., Claessens, A. M. K., Huston, A. C., Child Development, 84, 397–412. doi:10.1111/j.1467-8624.2012.
Klebanov, P., Pagani, L. S., et al. (2007). School readiness and later 01851.x.
achievement. Developmental Psychology, 43, 1428–1446. doi:10. Martin, A., Razza, R. A., & Brooks-Gunn, J. (2012). Sustained attention
1037/0012-1649.43.6.1428. at age 5 predicts attention-related problems at age 9. International
Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Journal of Behavioral Development, 36, 413–419. doi:10.1177/
Schellinger, K. B. (2011). The impact of enhancing students’ social 0165025412450527.
and emotional learning: a meta-analysis of school-based universal McClelland, M. M., Cameron Ponitz, C. E., Connor, C., Farris, C.,
interventions. Child Development, 82, 405–432. Jewkes, A., & Morrison, F. J. (2007). Links between behavioral
56 J Abnorm Child Psychol (2017) 45:45–56

self-regulation and preschoolers’ literacy, vocabulary, and math Schultz, D., Izard, C. E., Ackerman, B., & Youngstrom, E. (2001).
skills. Developmental Psychology, 43, 947–959. Emotion knowledge in economically disadvantaged children.
McClelland, M. M., Cameron, C. E., Duncan, R., Bowles, R. P., Acock, Development and Psychopathology, 13, 53–67.
A. C., Miao, A., & Pratt, M. E. (2014). Predictors of early Sjöwall, D., Roth, L., Lindqvist, S., & Thorell, L. B. (2013). Multiple
growth in academic achievement: the Head-Toes-Knees- deficits in ADHD: executive dysfunction, delay aversion, reaction
Shoulders-task. Frontiers in Psychology, 5, 1–14. doi:10.3389/ time variability, and emotional deficits. Journal of Child Psychology
fpsyg.2014.00599.Article599. and Psychiatry, 54, 619–627. doi:10.1111/jcpp.12006.
Mostow, A., Izard, C. E., Fine, S., & Trentacosta, C. (2002). Modeling Sjöwall, D., Bohlin, G., Rydell, A. M., & Thorell, L. B. (2015a).
emotional, cognitive, and behavioral predictors of peer acceptance. Neuropsychological deficits in preschool as predictors of ADHD
Child Development, 73, 1775–1787. doi:10.1111/1467-8624.00505. symptoms and academic achievements in late adolescence. Child
Nelson, C. A., & Fivush, R. (2004). The emergence of autobiographical Neuropsychology. Advance online publication. doi:10.1080/
memory: a social cultural developmental theory. Psychological 09297049.2015.1063595.
Review, 111, 486–511. Sjöwall, D., Backman, A., & Thorell, L. B. (2015b). Neuropsychological
Pagani, L. S., Fitzpatrick, C., & Parent, S. (2012). Relating kindergarten heterogeneity in preschool ADHD: investigating the interplay be-
attention to subsequent developmental pathways of classroom en- tween cognitive, affective, and motivation-based forms of regula-
gagement in elementary school. Journal of Abnormal Child tion. Journal of Abnormal Child Psychology, 43, 669–680. doi:10.
Psychology, 40, 715–725. doi:10.1007/s10802-011-9605-4. 1007/s10802-014-9942-1.
Toplak, M. E., West, R. F., & Stanovich, K. E. (2013). Practitioner re-
Pons, F., & Harris, P. L. (2002). Test of Emotion Comprehension – version
view: do performance-based measures and ratings of executive func-
2. Oxford: Oxford University Press.
tion assess the same construct? Journal of Child Psychology and
Pons, F., & Harris, P. (2005). Longitudinal change and longitudinal sta-
Psychiatry, 54, 131–143. doi:10.1111/jcpp.12001.
bility of individual differences in children’s emotion understanding.
Trentacosta, C., & Izard, C. (2007). Kindergarten children’s emotion
Cognition and Emotion, 19, 1158–1174.
competence as a predictor of their academic competence in first
Pons, F., Harris, P. L., & Doudin, P.-A. (2002). Teaching emotion under- grade. Emotion, 7, 77–88. doi:10.1037/1528-3542.7.1.77.
standing. European Journal of Psychology of Education, 17, 293– Trentacosta, C., Izard, C. E., Mostow, A. J., & Fine, S. E. (2006).
304. doi:10.1007/BF03173538. Children’s emotional competence and attentional competence in
Pons, F., Lawson, J. L., Harris, P. L., & de Rosnay, M. (2003). Individual early elementary school. School Psychology Quarterly, 21, 148–
differences in children’s emotion understanding: effects of age and 170. doi:10.1521/scpq.2006.21.2.148.
language. Scandinavian Journal of Psychology, 44, 347–353. doi: von Aster, M.G., Bzufka, M.W. & Horn, R.R. (2009). ZAREKI-K - neu-
10.1111/1467-9450.00354. ropsychological test battery for number processing and calculation
Raver, C. C., Blair, C., Garrett-Peters, P., & Family Life Project Key for children-kindergarten version. Hallbergmoos: Pearson.
Investigators (2014). Poverty, household chaos, and interparental von Salisch, M., Haenel, M., & Freund, P. A. (2013). Emotion under-
aggression predict children's ability to recognize and modulate neg- standing and cognitive abilities in young children. Learning and
ative emotions. Development and Psychopathology, 37, 114. doi:10. Individual Differences, 26, 15–19. doi:10.1016/j.lindif.2013.04.001.
1017/S0954579414000935. von Salisch, M., Klinkhammer, J. & Haenel, M. (2015a). Welche
Rhoades, B. L., Warren, H. K., Domitrovich, C. E., & Greenberg, M. T. kognitiven Fähigkeiten von jungen Kindern sagen ihre Fortschritte
(2011). Examining the link between preschool social–emotional beim Wissen über Emotionen voraus? [Which of young children's
competence and first grade academic achievement: the role of atten- cognitive abilities predict their acquisition of knowledge about emo-
tion skills. Early Childhood Research Quarterly, 26, 182–191. doi: tions?] In Cloos, P., Koch, K. & Mähler, C. (Hrsg.). Entwicklung
10.1016/j.ecresq.2010.07.003. und Förderung in der frühen Kindheit [Development and training in
Riggs, N. R., Greenberg, M. T., Kusche, C. A., & Pentz, M. A. (2006). early childhood] (pp. 115–132). Weinheim: BeltzJuventa.
The mediational role of neurocognition in the behavioral outcomes von Salisch, M., Haenel, M., & Denham, S. (2015b). Self-regulation,
of a social-emotional prevention program in elementary school stu- language skills, and emotion knowledge in young children. Early
dents. Prevention Science, 7, 91–102. doi:10.1007/s11121-005- Education and Development, 26, 792–806. doi:10.1080/10409289.
0022-1. 2015.994465.