10.5

Temperament

Although the overall development of emotions and emotion regulation capabilities is roughly similar for most children, there are also very large individual differences in children’s emotional functioning. Some infants and children are relatively mellow: they do not become upset easily, and they usually do not have difficulty calming down when they are upset. Other children are quite emotional: they get upset quickly and intensely, and their negative emotion persists for a long time. Moreover, children differ in their timidity, their expression of emotions, and in the ways they deal with their emotions.

Comparison of Reactions

Compare these two 3-year-old children, Maria and Bruce, as they react to Teri, an adult female stranger:

• When Teri walks over to Maria and starts to talk with her, Maria smiles and is eager to show Teri what she is doing. When Teri asks Maria if she would like to go down the hall to the play room (where experiments are conducted), Maria jumps up and takes Teri’s hand.

• In contrast, when Teri walks over to Bruce, Bruce turns away. He doesn’t talk to her and averts his eyes. When Teri asks him if he wants to play a game, Bruce moves away, looks timid, and softly says “no.” (N. Eisenberg, laboratory observations)

Such differences in how individual children react to similar situations led researchers to develop the concept of temperament. Temperament refers to individual differences in emotion, activity level, and attention that are exhibited across contexts and that are present from infancy; thus, they are thought to be genetically based (Bornstein et al., 2015). As we will see later in this section, the temperament of any given child is influenced both by the child’s genes and by the environment in which they live; therefore, the construct of temperament is highly relevant to our themes of individual differences and the role of nature and nurture in development.

Classification of Temperament

Due partly to variations in temperament, children often show very different reactions to the same situation. Alexander Thomas and Stella Chess are pioneers of temperament research (Thomas & Chess, 1977; Thomas, Chess, & Birch, 1968). They began by interviewing a sample of parents, repeatedly and in depth, about their infants’ specific behaviours. To reduce the possibility of bias in the parents’ reports, the researchers asked that instead of giving interpretive characterizations, such as “he’s often cranky” or “she’s interested in everything,” the parents provide detailed descriptions of their infants’ specific behaviours. On the basis of those interviews, nine characteristics of children were identified, including their mood, adaptability, activity level, attention span, and persistence. Further analyzing the interview results in terms of these characteristics, the researchers classified the infants into three groups:

• Easy babies adjusted readily to new situations, quickly established daily routines such as sleeping and eating, and generally were cheerful in mood and easy to calm.

• Difficult babies were slow to adjust to new experiences, tended to react negatively and intensely to novel stimuli and events, and were irregular in their daily routines and bodily functions.

• Slow-to-warm-up babies were somewhat difficult at first but became easier over time as they had repeated contact with new objects, people, and situations.

In the initial study, 40% of the infants were classified as easy, 10% as difficult, and 15% as slow to warm up. The rest did not fit into any single category but rather displayed characteristics of more than one category.

Of particular importance, some dimensions of temperament showed relative stability within children over time; temperament in infancy predicted how children were doing years later. For example, difficult infants tended to have problems with adjustment at home and at school, whereas few of the easy infants had such problems.

Measuring Temperament

Since the groundbreaking efforts of Thomas and Chess, researchers studying temperament no longer group children into categories such as easy, difficult, or slow to warm up, which reflects what is known as a between-person approach to understanding development. Rather, researchers now characterize every child along the same set of dimensions of temperament—a within-person approach to understanding development. According to this newer approach, every child has some level of each dimension of temperament. Not all researchers agree on exactly how many dimensions of temperament there are; however, one of the leading experts in this area, Mary Rothbart, has identified five key dimensions of temperament:

1. Fear

2. Distress/anger/frustration

3. Attention span

4. Activity level

5. Smiling and laughter (Gartstein & Rothbart, 2003; Rothbart et al., 2001).

She has created measures of temperament in both infancy (the Infant Behavior Questionnaire) and early childhood (the Child Behavior Questionnaire) that ask parents, teachers, or observers to rate each child along several dimensions of temperament (Gartstein & Rothbart, 2003; Rothbart et al., 2001).

Ratings and Stability of Temperament

The key advantage of parents’ reports of temperament is that parents have extensive knowledge of their children’s behaviour in many different situations. One important disadvantage of this method is that parents may not always be objective in their observations, leading to discrepancies with laboratory measures (Rothbart & Bates, 2006). Another disadvantage is that many parents do not have wide knowledge of other children’s behaviour to use as a basis for comparison when reporting on their own children; what is irritability to some parents, for example, may be near-placidness to others.

In contrast, the key advantage of laboratory observational data is that such data are less likely to be biased than is an adult’s personal view of the child. A key disadvantage is that children’s behaviour usually is observed in only a limited set of circumstances, and thus lab-based measures of temperament may not have external validity (see Chapter 1). Consequently, laboratory observational measures may reflect a child’s mood or behaviour at a given moment, in a particular context, rather than the totality of the child’s temperament.

Conclusion on Measuring Temperament

No measure of temperament is perfect, and it is prudent to assess temperament using a variety of different methods (Rothbart & Bates, 2006). Temperament is considered to be relatively consistent across time and across situations. Using parents’ ratings on the Infant Behavior Questionnaire, one study found that an infant’s level on each dimension of temperament is considerably stable across the first year of life but that there is some evidence of change over time as well (Bornstein et al., 2015). Across early childhood, children who are prone to more anger and distress than their peers at age 3 tend to be more angry and distressed than their peers at ages 6 and 8 (Rothbart, Derryberry, & Hershey, 2000); those prone to display happiness remain relatively happy across the same age range (Durbin et al., 2007; Sallquist et al., 2009). But this is not to say that temperament is entirely immutable; it can and does change over time.

Determinants of Temperament

Temperament is believed to have a strong basis in biology and genetics (Saudino & Wang, 2012). Evidence for a genetic component to temperament comes from dozens of studies. For example, identical twins are more similar to each other in these aspects of their emotion and regulation than are fraternal twins (Rasbash et al., 2011; Saudino & Wang, 2012). Studies have shown connections between an individual’s genes and aspects of temperament such as self-regulatory capacity (Depue & Fu, 2012; Goldsmith, Pollak, & Davidson, 2008; Saudino & Wang, 2012). For example, genes related to the functioning of dopamine and other neurotransmitters that affect voluntary attentional processes (executive attention) appear to be especially relevant for self-regulation (Posner, Rothbart, & Sheese, 2007).

Although genes are clearly important for temperament, the environment plays a role as well, even from before birth. In Chapter 2, we learned about teratogens, which are external agents that can cause birth defects in a fetus. Teratogens such as nutritional deficiencies, exposure to cocaine, or maternal stress and anxiety during pregnancy have each been found to predict infants’ and young children’s temperament-based abilities to regulate their attention and behaviour (Dennis et al., 2006; Huizink, 2008, 2012).

As children age, the behaviours of their parents become a strong influence on temperament. Children who grow up in home environments that are harsh or unstable tend to have problems with self-regulation and the expression of emotion (Lewis et al., 2007). Warm and responsive parenting can have the opposite effect; in a twin study, the twin who received more warm and responsive parenting had fewer emotional problems and exhibited more positive affect and more prosocial behaviours than the twin sibling (Deater-Deckard et al., 2001). Yet, as suggested by the active child theme, children’s temperamental characteristics can affect their environments and particularly their parents’ behaviours (Belsky, Bakermans-Kranenburg, & van IJzendoorn, 2007; Lee, Zhou et al., 2013).

Understanding Goodness of Fit

To determine whether genes or the environment have more influence over temperament, a group of researchers (Lemery-Chalfant et al., 2013) used a twin study, which is a research design we learned about in Chapter 3. In this study, parents of 807 pairs of twins (301 MZ [monozygotic, or identical], 263 same-sex DZ [dizygotic, or fraternal], and 243 opposite-sex DZ) reported on both of their children’s temperaments, and then statistical analyses were used to determine how similar the twins were. If MZ twins in the same home environment were more similar than DZ twins who shared the same home environment, then they would conclude that temperament is more determined by genes than the environment. Indeed, that is what they found: a large portion of the variance in three aspects of temperament (effortful control, negative affectivity, and extroversion) was explained by heritability (Lemery-Chalfant et al., 2013). They also found that key aspects of the home environments—namely how chaotic and unsafe they were—had a heritable component as well: parents’ temperaments affected children both directly through genetic transmission and indirectly through the home environment they created (Lemery-Chalfant et al., 2013). This latter contribution of genes to children’s temperament is referred to as a passive gene–environment correlation, a concept discussed in Chapter 8 with regard to the heritability of intelligence.

Undoubtedly, a combination of genetic and environmental factors jointly contributes to individual differences in children’s emotions and related behaviours (Rasbash et al., 2011; Saudino & Wang, 2012).

Conclusion

Children with different temperaments will have different reactions to the same situation, and thus some situations are better suited for certain temperaments. Imagine two 4-year-olds who arrive at a loud and chaotic birthday party. The child who has a high activity level will immediately join in, whereas the child high in fearfulness will observe from the perimeter before finding a way to play with the other children. The high-activity child is in their element; the high-fearfulness child is decidedly not. How well a child’s temperament matches the demands of a particular context is called goodness of fit. The family provides arguably the most important context for issues related to goodness of fit. Research indicates that children who are impulsive or low in emotion regulation seem to have more problems and are less sympathetic to others if exposed to hostile, intrusive, and/or negative parenting rather than to supportive parenting (Hastings & De, 2008; Kiff, Lengua, & Zalewski, 2011; Lengua et al., 2008). Similarly, children prone to emotions such as anger are more likely to have behavioural problems such as aggression if exposed to hostile parenting or low levels of positive parenting (Bates, Schermerhorn, & Petersen, 2012).

A Canadian study found that children whose parents rated them as high in shyness had fewer problems with peers and fewer emotional problems if their mothers were high in warmth (Coplan, Arbeau, & Armer, 2008).

It is important to note, however, that some children’s temperaments make them highly reactive to both positive and negative family environments. These children exhibit a characteristic known as differential susceptibility, which means that the same temperament characteristic that puts some children at high risk for negative outcomes when exposed to a harsh home environment also leads them to blossom when their home environments are positive (Belsky & Pluess, 2009). For example, children with impulsive temperaments have been found to exhibit high levels of behaviour problems in adolescence if they are raised in harsh family environments; but if raised in positive family environments, they exhibit low levels of behaviour problems (Rioux et al., 2016b). Children with such susceptible, or highly responsive, temperaments have been described as “orchids” who thrive when conditions are good but wither when they are bad. In contrast, children who are less sensitive to their environments and do well in all but the most high-risk environments are thought of as “dandelions” (Boyce & Ellis, 2005).

Although evidence for differential susceptibility has been found in several longitudinal studies including the Québec Longitudinal Study of Child Development (Rioux et al., 2016a), it is also the case that all children benefit from positive parenting environments, regardless of whether their genetic and temperamental characteristics make them likely to be reactive to their environments (Belsky et al., 2015).