Probably one of the most common forms of assessment in developmental research is the use of questionnaires or interviews that assess some aspect of a participant's functioning from someone else's perspective, such as that of a parent, teacher, or peer (Neyer, chap. 4, this volume). Paper-and-pencil measures are often used because of their efficiency. Many participants can be questioned at one time, so obtaining these reports is much less time consuming and less expensive than observational research.
Using other reports has some advantages over using self-report or observational data. Compared to children and adolescents, adults are more skilled in answering questions, and parents and teachers have the opportunity to observe children over time and in a variety of social situations. Although adults' reports on children probably are more objective (on average) than children's reports about themselves, adults' reports are subject to certain biases. Parents' reports may be influenced by social desirability, and there is some evidence that teachers may rate academically skilled children more positively in general (Underwood, 1997). One way often used to assess the validity of parents' and teachers' reports is to examine agreement among informants. However, results regarding rater agreement are often inconsistent. For example, Guthrie et al. (1997) reported that the relation between teachers' and parents' assessments of children's reactivity were modest at best. In contrast, Eisenberg et al. (1996) found that teachers' and parents' reports of reactivity and regulation (particularly the latter) were positively correlated. Differences between informants are not surprising, however, and may not indicate low validity of a measure because individuals' expression of behaviors likely varies across contexts. Indeed, differences between parents' and teachers' reports reinforce the value of obtaining information from adults who have observed children in different settings.
It has been widely shown in the literature that as children grow older, peers increase in importance; therefore, peers' reports may be used to measure friendship status as it relates to constructs such as emotion regulation and social competence. Socio-metrics are the procedures used to measure peer relationships through a system of rating popularity. These procedures may be used with children as young as preschool through the use of pictures (e.g., a smile, frown, and neutral face; Asher, Singleton, Tinsley, & Hymel, 1979), or with older children using questionnaires to rank classmates and to determine peer status, acceptance, and rejection. In addition, preschool children and elementary school children appear to be relatively good reporters of peers' anger and related constructs (e.g., aggression; Eisenberg, Pidada, & Liew, 2001; Maszk, Eisenberg, & Guthrie, 1999).
Even though a variety of constructs have been measured with other informants, we choose to illustrate the use of others' reports in reference to temperament. Temperament can be defined as psychological qualities and behaviors that display considerable variation among infants and young children and have a relatively stable physiological basis that derives from the individual's genetic constitution (Kagan, 1994, p. 16). Most temperament theorists view temperament and biology as intertwined and see temperament as having stable, enduring properties that can be modified to a degree by contextual factors, such as parenting (Rothbart & Bates, 1998). Adult-report measures of child/infant temperament include scales that assess many dimensions of temperament such as regulation (e.g., attention focusing, inhibitory control), various types of emotion (e.g., positive emotion, anger/frustration, fear, sadness), activity level, impulsivity, surgency/approach, and soothability (e.g., the Infant Behavior Questionnaire [IBQ], Rothbart, 1981; the Toddler Behavior Assessment Questionnaire [TABQ], Goldsmith, 1996; the Child Behavior Questionnaire [CBQ], ages 3-7, Goldsmith & Rothbart, 1991, Rothbart, Ahadi, Hershey, & Fisher, 2001; the Early Adolescent Temperament Questionnaire [EATQ], ages 9-15, Capaldi & Rothbart, 1992; the Dimensions of Temperament Survey—Revised [DOTs-R], Windle & Lerner, 1986). Some examples of items that assess various temperamental constructs include, "pays close attention when someone tells him or her how to do something," from the attention scale of the EATQ, and "sometimes interrupts others when they are speaking," from the impulsivity scale of the CBQ. Items reflect typical behaviors across a variety of situations and settings, selected to tap an underlying temperamental predisposition. The CBQ and EATQ are currently available in parent-report formats; others have adapted and used the CBQ for teachers
(e.g., Eisenberg et al., 1997; Murphy, Eisenberg, Fabes, Shepard, & Guthrie, 1999). The DOTs-R is another parent-report measure that taps into several aspects of temperament (e.g., task orientation) that also has an adolescent self-report version.
Parents are likely excellent reporters of child temperament because parents have viewed their children in a variety of settings and over a length of time (Rothbart & Bates, 1998). Additionally, parent-report measures, like other types of questionnaires, are inexpensive to create, apply, and analyze. Moreover, these types of measures have shown moderate to strong convergent validity with observational methods. For example, Mathney, Wilson, and Thoeben (1987) found laboratory scores of temperament correlated with maternal report (rs ranged from .38 to .52 over a year). Additionally, Rothbart (1986) found moderate convergent validity between dimensions of the IBQ and home observations of temperament. Intercorrelations between overall reactivity measured by the IBQ and home observations ranged from .43 to .46. Yet despite the strength of using both parent reports and observation, not all facets of temperament are observable, and observational results may not always generalize to outside the laboratory or structured settings.
As one might imagine, some researchers have questioned whether parents' reports of their children's temperament assess characteristics of the parent more than of the child (Kagan, 1998). Although parental bias can distort reports of temperament, Lemery, Biersach, Chipongnian, Greenberg, and Goldsmith (2001) found that parental characteristics (i.e., personality, depression, family expressiveness) were equally related to parental report and lab measures of infant temperament, suggesting that parental reporting bias does not account for all the relations of temperament with other variables.
Was this article helpful?