ریسک پذیری فیزیکی در کودکان ابتدایی مدرسه: اندازه گیری و مسائل مربوط به تنظیم احساسات

Abstract Children who engage in physical risk taking experience more injuries. In the present study children 7–12 years completed laboratory tasks and standardized questionnaires to explore inter-relations between measures of risk taking and to determine if the Balloon Analog Risk Task (BART) is a valid index of physical risk taking. Emotion regulation skills also were assessed to determine if poorer skills are associated with greater risk taking, and what measures of risk taking show these relations. Results revealed that general measures of risk taking propensity predicted both BART performance and physical risk taking. Similarly, poorer emotion regulation skills were associated with greater risk taking scores for both the BART and physical risk taking task. Measures of physical risk taking related to one another, however, none of these scores related to performance on the BART. Implications for measuring physical risk taking and exploring emotional links to risk taking in future research are discussed.

. Introduction Injuries pose a significant health threat for children. In most developed countries they are the leading cause of death (World Health Organization, 2005). For elementary-school children, many injuries occur when they are away from home and making their own decisions about risk taking activities (National Center for Injury Prevention and Control [NCIPC], 2010 and Shanon et al., 1992). Past research has documented a number of temperament, cognitive, and emotional determinants that influence physical risk taking (i.e., engaging in a behavior that increases risk of physical injury when there are alternatives that pose less risk of injury, Morrongiello & Sedore, 2005) among children (Morrongiello et al., 2007, Morrongiello and Rennie, 1998 and Schwebel and Gaines, 2007). Such findings have paved the way for the development of effective programs to reduce these behaviors and children’s risk of injury (Morrongiello and Marks, 2008 and Morrongiello and Matheis, 2007). Of course, physical risk taking is only one of several different types of risk taking (e.g., social, academic). There is long standing debate about whether individuals show uniformity in risk taking across decision domains (Baucells and Rata, 2006, Hanoch et al., 2006 and Soane and Chmiel, 2005) and if there are underlying personality attributes that give rise to consistency in risk taking across domains (Hanoch et al., 2006 and Zuckerman and Kuhlman, 2000). Drawing on these issues, the current study examined inter-relations between different measures of risk taking, with a focus on determining if children show similar patterns of risk taking on the Balloon Analog Risk Task (BART) as they do on a physical risk taking task. The expectation was that these risk taking scores would be highly positively correlated if the BART was a good proxy measure for physical risk taking in elementary-school age children. Relations between emotion regulation and risk taking also were examined to determine if emotion dysregulation is associated with greater risk taking, and what measures of risk taking show these relations. 1.1. Measuring risk taking Developing reliable and valid assessment approaches for measuring risk taking is challenging. For elementary-school children, because the focus is often on physical risk taking that can lead to injuries, the measurement challenge is substantial. Innovative approaches to measuring children’s physical risk taking in ethically acceptable ways have included observations of children’s behavior in naturally hazardous situations (Ginsburg & Miller, 1982) and creating contrived situations in which they indicate intentions to risk take when they are led to believe they will have to demonstrate the behaviors endorsed (e.g., Morrongiello et al., 2007, Morrongiello and Matheis, 2007 and Morrongiello and Sedore, 2005). In contrast, for teenagers and adults, most risk taking measures involve self-reports about the health-related behaviors of interest (e.g., smoking, drinking). However, because self-reports about undesirable behaviors can be inaccurate (Ladouceur et al., 2000), there is increasing concern about their validity and interest in developing behavioral measures that index risk taking among teens and adults. Two such measures have gained in popularity, namely – the Iowa Gambling Task (Bechara, Damasio, Damasio, & Anderson, 1994) and the Balloon Analog Risk Task (BART, Lejuez et al., 2002). Although performance on the Iowa Gambling Task has been shown to relate to some aspects of health risk behaviors in adolescents (Bechara et al., 2001), evidence from school-age children indicates that it is not a valid measure of physical risk taking (Morrongiello, Lasenby-Lessard, & Corbett, 2009). For the BART, several studies report moderate associations with real world risk behaviors (smoking, gambling, alcohol use) for teens (Lejuez, Aklin, Zvolensky, & Pedulla, 2003) and adults (Lejuez, Simmons, Aklin, Daughters, & Dvir, 2004). Further support for the construct validity of the BART comes also from findings that it relates to temperament attributes among adolescents in ways one would expect. It is positively associated with measures of disinhibition, including sensation seeking and impulsivity (Aklin et al., 2005, Hunt et al., 2005, Lejuez et al., 2003 and Lejuez et al., 2002). Whether the BART can be used to index physical risk taking and whether it relates to temperament attributes associated with risk taking among elementary-school children, however, remains to be determined and was considered herein. 1.2. Emotion regulation and risk taking Historically, the field of decision making has neglected the influence of emotions on risk taking, preferring instead to focus on rational-based decision making. More recently, however, there is increasing evidence that emotions can undermine rational decision making and motivate risk taking or lead to impulsive decision making (Lerner and Tiedens, 2006 and Loewenstein et al., 2001). Emotion dysregulation has been conceptualized in many ways (e.g., impulsivity, lack of emotional awareness, emotional reactivity) and measured using a variety of questionnaires (e.g., Difficulties in Emotion Regulation Scale: Gratz & Roemer, 2004; Emotion Reactivity Scale: Nock, Wedig, Holmberg, & Hooley, 2008), but the general pattern of results is the same: individuals who are poor at regulating their emotions tend to engage in more risk taking (Cyders et al., 2007 and Yuen and Lee, 2003). In virtually all of this research, however, the focus has been on adolescents and adults (e.g., Cooper et al., 2003, Hessler and Katz, 2010 and Raffaelli and Crockett, 2003), leaving open the question of whether elementary-school children who are poor at emotion regulation show similar dysfunctional risky behavioral patterns. 1.3. Present study The present study had several aims. The first was to assess if the BART could be used as a proxy indicator of physical risk taking for elementary-school children. A computer-based behavioral measure of risk taking could be very useful because it would eliminate both under-reporting biases that can occur in physical risk taking questionnaire measures and data gathering limitations related to climate issues that occur when actual playground tasks are used. To address this aim, inter-relations of BART performance with other measures that tap physical risk taking were examined. Risk taking scores also were related to two temperament characteristics that have been linked to physical risk taking, namely – high intensity behavior (i.e., associated with increased frequency of physical risk taking and injury; Bijur et al., 1986, Nyman, 1987 and Vollrath et al., 2003), and inhibitory control (i.e., child’s ability to inhibit impulses when instructed to do so, which has been associated with reduced frequency of physical risk taking and injury; Morrongiello et al., 2006, Schwebel, 2004 and Schwebel and Plumert, 1999). A general index of risk taking propensity also was included to ascertain if this related in a similar way to performance on the BART and physical risk taking task. The second aim was to explore relations between emotion dysregulation and risk taking in order to determine if poor emotion regulation is associated with greater risk taking among school-age children, and if this pattern is evident for both the BART and physical risk taking task. Past research has shown that children attend to the emotions communicated by other children during risky play activities on playgrounds and they utilize this information in estimating injury risk (Morrongiello & Rennie, 1998). However, we are not aware of any research examining whether emotional dysregulation is associated with increased risk taking in children.

3. Results Inter-correlations between risk taking scores (see Table 1) revealed that the two BART scores were highly significantly positively related, but neither significantly related to physical risk taking on the playground task. An Analysis of Variance (ANOVA) with sex (2) as a between-participant factor replicated the common finding that boys endorsed more risk taking than girls for the playground task, [F(1, 68) = 7.90, p < .01, ηp2 = .36], but similar differences were not evident for the BART scores. Table 1. Averages (SD) and inter-correlations for the three risk taking measures (N = 70). Measure 1 2 3 MEAN (SD) 1. Playground – .02 .04 3.65a (2.81) 2. BART-pumps – – .95⁎⁎ 20.27 (10.76) 3. BART-pops – – 5.51 (3.57) ∗p < .05. a Male and female scores significantly differed (M = 4.92 and 2.39, SD = 3.92 and 1.70, respectively); no sex differences were obtained for any other scores. ⁎⁎ p < .01. Table options Table 2 presents inter-correlations depicting how these individual risk taking measures related to the questionnaire scores. Because of the high inter-correlation between the two BART measures, we limited the focus to the number of pumps score, as has been suggested in past research (cf. Lejuez et al., 2007). Scores on the BART and playground risk taking tasks both positively related to risk taking propensity (RPS). Similarly, scoring high in emotion dysregulation was associated with greater risk taking on both the BART and playground task. Scores reflecting high activity and intensity in executing behaviors (EATQ-HIP) related to physical risk taking, as has been found before (Bijur et al., 1986, Nyman, 1987 and Vollrath et al., 2003), but not to BART performance. Similarly, risk taking scores on the playground task negatively related to inhibitory control, as has been found before (Matheny, 1986, Morrongiello et al., 2006, Schwebel, 2004 and Schwebel and Plumert, 1999), but did not relate to BART performance. Finally, the IBC, which indexes physical risk taking, related to playground risk taking but not to performance on the BART. Table 2. Averages (SD) for questionnaires (N = 70) and inter-correlations of these with each risk taking measure. Questionnaire Risk taking measure MEAN (SD) PG B-pump IBC .77⁎⁎ .19 41.09 (13.87) RPS .59⁎⁎ .33⁎⁎ 18.56 (6.25) EATQ-HIP .36⁎⁎ .13 32.46 (8.19) EATQ-IC −.49⁎⁎ −.13 37.46 (8.77) EDS-C .42⁎⁎ .31⁎⁎ 47.80 (15.99) Note: IBC = Injury Behavior Checklist. RPS = Risk Propensity Scale. EATQ-HI = Early Adolescent Temperament Questionnaire-High Intensity. EATQ-IC = Early Adolescent Temperament Questionnaire-Inhibitory Control. EDS-C = Emotion Dysregulation Scale-overall average. PG = playground risk taking. B-pump = average # of pumps when balloons were not broken on the BART. ∗p < .05. ⁎⁎ p < .01. Table options To determine which factors predicted physical risk taking and performance on the BART (i.e., number of pumps), separate hierarchical regression analyses were conducted. For both regressions, in Step 1, age and sex were entered to control for any effect of these variables. All other variables that correlated with the criterion (see Table 2) were then entered in Step 2. The results replicated the correlation results. For physical risk taking, Step 1 contributed significantly to the model [F(2, 67 = 4.07, p < .05], accounting for 11% of the variance. Sex differences in risk taking were responsible for this effect, with females engaging in less risk taking than males, β = −.32, t = −2.78, p < .01. Step 2 also significantly contributed to the model [F(5, 62) = 19.30, p < .01], accounting for 54% additional variance. Significant predictors included injury-risk behaviors (β = .73, IBC: t = 9.20, p < .01), propensity for risk taking (β = .57, RPS: t = 5.99, p < .01), inhibitory control (β = −.43, EATQ-IC: t = −3.77, p < .01), intensity of behavior (β = .30, EATQ-HIP: t = 2.63, p < .01), and emotion dysregulation (β = .43, EDS-C: t = 3.77, p < .01. For BART performance, Step 1 variables (age, sex) did not predict risk taking. Step 2 significantly contributed to the model and accounted for 11% of the variance, F(2, 65) = 4.25, p < .05. Significant predictors included propensity for risk taking (β = .30, RPS: t = 2.63, p < .01) and emotion dysregulation (β = .30, EDS-C: t = 2.43, p < .05).