Self-report methods are commonly used in developmental psychology. However, several important issues should be noted when using self-report data. First, age is an obvious concern because most children under age 8 have difficulty completing paper-and-pencil measures. Nevertheless, as we discuss briefly in this section, methods have been constructed that appear to successfully elicit young children's reports of some constructs. Second, the validity of self-report data sometimes is a concern because when self-report data are compared to other informants' reports, correlations among reporters often are variable (e.g., Achenbach, 1991). This calls into question what reports should be relied on most and how or if the data from multiple reporters should be used in combination. Third, many factors such as socioeconomic status, ethnicity, and social desirability have been found to affect responses to written measures (Knight & Hill, 1998). In response to potential cultural biases, Knight and Hill (1998) and others have begun working toward cross-cultural validation of some self-report measures (e.g., comparisons between internal consistency scores for Anglo Americans and Latino Americans on scales such as the Child's Depression Inventory; Kovacs, 1981).
One clear advantage of using self-report methods to study developmental psychology centers around the idea that the subjective experience of an individual has important implications for development. As Bronfenbrenner (1979) wrote, "the aspects of the environment that are most powerful in shaping the course of psychological growth are overwhelmingly those that have meaning to the person in a given situation" (p. 22). Certainly both objective and subjective reports provide important information; however, it may be the individual's personal subjective interpretation that is most influential in shaping development (Morris, Silk, et al, 2002).
Self-report measures are not always as objective as other methods and sometimes have low correlations with other methods. Additionally, self-report measures may not be able to assess concepts that are not always salient to the reporter (e.g., emotional reactivity). Nonetheless, self-reports often correlate with the reports of other people or observational measures. In those cases, structural equation modeling (SEM), discussed further in the longitudinal section, is a statistical technique that can be used to incorporate data from multiple informants (if they are related to some degree; e.g., Zhou et al., 2002).
Most child research before the late 1990s relied on reports and observations by trained observers or parents (Sessa, Avenevoli, Steinberg, & Morris, 2001), largely because of concerns about young children's ability to provide reliable and valid reports of their experiences. However, more recent reviewers and researchers have questioned the assumption that children cannot report important information (Miller & Aloise, 1989; Ridgeway, Waters, & Kuczaj, 1985). It is likely that developmental researchers have historically underestimated younger children's social cognitive competencies and their ability to report on their own experiences, primarily because of the methods used to assess children's beliefs and social understanding (Hart & Damon, 1986; Miller & Aloise, 1989). For example, most early researchers examining young children's person perception used open-ended interview techniques, which required extensive verbal production and expressive skills. Because young children's verbal comprehension skills are better than their verbal expressive skills (Kuczaj, 1986), observed age-related differences in children's use of dispositional terms in their descriptions of others likely reflected linguistic immaturity (Furman & Bierman, 1983). In addition, the demand characteristics of the standard interview research situation (i.e., being questioned by an unfamiliar adult) probably inhibited young children's ability to provide psychologically meaningful information in many studies. When children have been interviewed by more "benign" interviewers, such as puppets, children as young as 3Vi years old have been able to provide general descriptions of their own and others' internal states and emotions with adequate stability (e.g., Denham, 1986; Eder, Gerlach, & Perlmutter, 1987). Indeed, there has been a recent surge in research attempting to assess young children's perceptions of constructs like self-concept (e.g., Eder et al., 1987), parent-child relations (Morris, Silk, et al., 2002; Morris, Steinberg, et al., 2002; Sessa et al., 2001), sympathy and empathy (e.g., Miller, Eisenberg, Fabes, & Shell, 1996), and school engagement (Measelle, Ablow, Cowan, & Cowan, 1998). Obviously, infants and toddlers cannot provide self-report data, but this recent research with preschool and elementary-school-age children is promising.
To illustrate some of the issues involved in using self-report data in developmental research, we have chosen to highlight individuals' self-reports of mood/emotionality and coping. The advantages of using self-report for these types of constructs stem from the fact that many emotion-related processes and coping strategies are unobservable to others and, as a result, are difficult to measure. Self-report assessments tap individuals' own experience of emotion and coping, which is beneficial in understanding how self-construction and awareness affect developmental outcomes. However, self-report measures of this sort also have disadvantages; for example, some individuals may not be aware of the coping strategies that they use or may provide biased or self-serving reports of emotion or coping behavior (Eisenberg & Morris, 2002; Eisenberg, Morris, & Spinrad, 2005; Lennon, Eisenberg, & Carroll, 1983).
Researchers studying coping have identified a long list of strategies for managing stress (e.g., problem solving, cognitive restructuring, catastro-phizing, emotional ventilation, physical activities, acceptance, distraction, avoidance, wishful thinking, humor, social withdrawal, alcohol or drug use, and seeking social support; Compás, Connor, Saltz-man, Thomsen, & Wadsworth, 2001). Some of these coping strategies likely assess emotionality (e.g., emotional ventilation) or outcomes of coping. Thus, investigators need to carefully consider what specific measures of coping assess and how they relate to measures such as emotionality. Moreover, researchers studying children and adolescents' coping often use self-report measures that assess coping responses to hypothetical vignettes involving emotion and emotion management strategies (e.g., Band & Weisz, 1988; Saarni, 1997). These vignettes involve having children or adolescents read about a stressful situation or problem (or an experimenter reading to them the vignette) and then answer questions about how they would cope with the problem. This type of method usually is an attempt to present real-life stressful situations that occur in everyday life; however, when using hypothetical vignettes, investigators should consider that children may report the socially desirable response and not the response he or she would enact in real life. For example, Underwood, Coie, and Herbsman (1992) found that children nominated as aggressive by their peers did not differ from children classified as nonaggressive in their reactions to videotaped vignettes designed to elicit aggression-related emotions. Moreover, Lennon et al. (1983) found that children reported more empathy when interviewed by same-sex than other-sex adult interviewers.
Children's self-reported coping strategies have also been measured in a number of recent studies with survey measures such as the Children's Coping Strategies Checklist (CCSC; Ayers, Sandler, West, & Roosa, 1996). With this measure, children rate how often they used particular coping behaviors when they had a problem in the last month. This measure has produced four factors (Active Strategies, Avoidant Strategies, Distraction Strategies, and Support Seeking Strategies). There is also a parent-report version of this questionnaire. One advantage to this type of measure is that it calls for the participant to draw on actual events that occurred and to report their coping strategies. Nevertheless, individuals may be biased in how they remember events and their own coping behaviors in these situations.
A method for collecting self-report data on mood/emotion that is becoming more popular is the experience sampling methodology (ESM; also called Ecological Momentary Assessment; EMA). ESM is designed to assess the subjective experience of individuals in their typical environment (Stone & Litcher-Kelly, chap. 5, this volume). With this methodology, participants are signaled (e.g., by a beeper or cell phone) based on a sampling schedule designed to obtain a representative sample of a person's everyday experiences (Hormuth, 1986). When signaled, participants report on predetermined emotions, activities, or thoughts (whatever is the focus of the study). This methodology constitutes an "experience-near" approach to the study of affective phenomenon, rather than the "experience-distant" approach typically used in laboratory and questionnaire research (Silk, Steinberg, & Morris, 2003). For example, Silk et al. (2003) had adolescents report on anger, sadness, and anxiety and their emotion regulation strategies using ESM. Participants were "beeped" randomly throughout the course of a week during designated times (8:00 a.m. to 9:30 p.m. during the week and 8:00 a.m. to 11:00 p.m. on the weekend). During the target hours of each day, using a table of random numbers, one signal was programmed for every 90- to 150-minute block with the provision that no signals would occur within 60 minutes of one another. On average, participants were signaled 42 to 48 times. At each signal, adolescents completed a short checklist that assessed their anger, sadness, and anxiety. Specifically, adolescents reported how mad, sad, and nervous they were on a 5-point scale, at the current time of reporting and during their most negative experience in the past hour. Like other studies, the degree of variation in mood, or lability, was measured as the standard deviation of the individual's score across sampling points. Emotion regulation was indicated by a decrease in emotional intensity from the most negative event to the current reported mood.
One advantage to the immediate reporting of emotions/moods with ESM is that it circumvents potential memory distortions and allows for an assessment of emotional behavior in a natural context (Larson, 1987). Some disadvantages of this method involve participants not completing the measure at the time they are signaled, which is difficult to monitor, and the time involved for participants—individuals are typically signaled multiple times a day for a week. Hierarchical Linear Modeling (HLM) is a statistical technique that is often used to analyze these data because each time point of data collection can be used in this type of model (and not all participants have equal numbers of data points). See Drasgow and Chuah (chap. 7, this volume) for an illustration of this use of HLM.
Was this article helpful?
Have You Ever Wanted To Achieve A State Of Total Relaxation But Never Believed That Yoga Was For You? Has the stress of daily life made you tense, uptight and too wound up to be able to think clearly? If so, then you are not alone. 40 of Americans feel that their lives are too stressful and over 60 of Americans say that they find themselves in situations where they feel lost at least once a week.