skin conductance reactivity as a moderator of associations ... · anne c. fletcher 1, cheryl...

https://doi.org/10.1177/0272431618812164

Journal of Early Adolescence2019, Vol. 39(8) 1154 –1176

© The Author(s) 2018Article reuse guidelines:

sagepub.com/journals-permissions DOI: 10.1177/0272431618812164

journals.sagepub.com/home/jea

Article

Skin Conductance Reactivity as a Moderator of Associations Between Youth Perceptions of Neighborhood Stress and Depressive Symptoms

Anne C. Fletcher1, Cheryl Buehler1, Amy L. McCurdy1, and Bridget B. Weymouth2

AbstractYoung adolescents (N = 68) completed questionnaires concerning perceptions of neighborhood stress (i.e., high negative influences, low cohesion, and connectedness). Youth self-reported their own depressive symptoms and participated in a public speaking task designed to be moderately stressful. Increases in skin conductance in response to this laboratory-based challenge task were measured (i.e., skin conductance level reactivity; SCLR). Higher levels of depressive symptoms were associated with greater perceptions of neighborhood stress, with the effects of neighborhood stress particularly strong among boys. For boys only, the association between perceived neighborhood stress and depressive symptoms was moderated by SCLR. For boys with average or above average levels of SCLR, greater perceptions of neighborhood stress were associated with more depressive symptoms.

1The University of North Carolina at Greensboro, NC, USA2The Pennsylvania State University, University Park, PA, USA

Corresponding Author:Anne C. Fletcher, Department of Human Development and Family Studies, The University of North Carolina at Greensboro, Greensboro, NC 27402-6170, USA. Email: [email protected]

812164 JEAXXX10.1177/0272431618812164Journal of Early AdolescenceFletcher et al.research-article2018

https://us.sagepub.com/en-us/journals-permissions

https://journals.sagepub.com/home/jea

mailto:[email protected]

http://crossmark.crossref.org/dialog/?doi=10.1177%2F0272431618812164&domain=pdf&date_stamp=2018-12-03

Fletcher et al. 1155

Keywordsskin conductance, depressive symptoms, reactivity, neighborhoods, adolescents

Evidence suggests that adolescence is a developmental period during which physiological stress reactivity is heightened and exposure to stressors may be particularly detrimental (Holder & Blaustein, 2014). As adolescents begin spending more unsupervised time outside of their homes and in the community, exposure to the neighborhood context becomes more develop-mentally salient (Delany-Brumsey, Mays, & Cochran, 2014). A large body of literature is devoted to the effects of neighborhood stressors on adolescent well-being. However, few studies have investigated the ways in which neighborhood stressors and individual-level characteristics may work together to impact adjustment. Greater consideration of person characteris-tics has the potential to provide a more comprehensive and holistic under-standing of ways in which stressors impact adolescents. Specifically, individual physiological stress reactivity may interact with perceived neigh-borhood stressors to predict depressive symptoms.

The purpose of the current study is to consider whether skin conductance reactivity moderates the association between neighborhood problems and youth depressive symptoms. A dual-risk model of vulnerability is used to generate hypotheses, to interpret findings, and to help elucidate why some adolescents develop depressive symptoms in the context of neighborhood problems and others do not.

Youth Depressive Symptoms

Early adolescence represents a particularly vulnerable time for the develop-ment of internalizing problems. During this developmental period, young people show increased physiological reactivity to stressors, which may cor-respond to increased symptoms of anxiety and depression (Holder & Blaustein, 2014). Analyzing data from a nationally representative sample of adolescents in Grades 6, 8, and 10, Saluja and colleagues (2004) reported that 18% of participants endorsed five or more items on an 8-item measure of depression based on the Diagnostic and Statistical Manual of Mental Disorders–Revised Third Edition (3rd ed., rev.; DSM-III-R; American Psychiatric Association, 1987). Percentages varied by gender, with 25% of girls but only 10% of boys in this nationally representative sample reporting five or more symptoms (Saluja et al., 2004). Research that has considered subclinical levels of depression has indicated even higher levels of

1156 Journal of Early Adolescence 39(8)

prevalence. Using a large, international sample of 14-year-old adolescents, Balázs et al. (2013) reported subthreshold depression rates of 38.24% in boys, with girls even more likely to experience clinical and subthreshold depression. Levels of depressive symptoms increase during adolescence (Bertha & Balázs, 2013), and these increases are particularly evident among girls (Bongers, Koot, van der Ende, & Verhulst, 2003).

Depressive symptomatology during adolescence is of concern, because youth who experience internalizing problems are also at higher risk for a variety of accompanying mental health problems, academic problems, and externalizing problems such as delinquency (Bhatia & Bhatia, 2007; Byrd, Loeber, & Pardini, 2012; Hallfors et al., 2004). Based on a systematic review of the literature on depressive symptomatology, Bertha and Balázs (2013) concluded that depressive symptoms constituted a significant health problem during adolescence, in part because they constitute a risk factor for major depressive episodes later in life. Such statistics suggest that research focused on understanding factors, both personal and contextual, that confer risk or resilience with respect to depressive symptoms meet a critical public health need.

Neighborhood Stress in Relation to Youth Depressive Symptoms

Links between characteristics of neighborhoods and depressive symptom-atology have been well documented among adults (Gariepy et al., 2015; Kim, 2010; Kohen, Leventhal, Dahinten, & McIntosh, 2008; Mair, Diez Roux, & Galea, 2008; Ross & Mirowsky, 2009), and a considerable amount of evidence indicates that such associations are present among adolescents as well. Evidence from longitudinal and cross-sectional research indicates that the impact of neighborhood effects becomes more pronounced as chil-dren reach adolescence (Browning, Burrington, Leventhal, & Brooks-Gunn, 2008; Delany-Brumsey et al., 2014). This change may correspond to neighborhood exposure; adolescents spend increasing amounts of time unsupervised in neighborhoods and outside the home, as compared to chil-dren. Thus, residing in neighborhoods that are perceived as dangerous or lacking in connection among residents constitutes a contextual stressor that has the potential to place youth at risk for distress, including heightened depressive symptomatology.

Ford and Rechel (2012) analyzed parent data (N = 32,499) from the 2007 National Survey of Children’s Health to assess contemporaneous associations between parent-perceived characteristics of neighborhoods and depressive symptoms and diagnoses among youth aged 12 to 17. More


depressive symptoms were linked with higher levels of neighborhood dis-order, social isolation, and lack of safety. Neighborhood disorder was also associated with a greater likelihood of a depression diagnosis among ado-lescents, despite the fact that such diagnoses are considerably less preva-lent during adolescence than are depressive symptoms (Ford & Rechel, 2012). Similarly, Latkin and Curry (2003) reported that the extent to which adolescents perceived their neighborhoods as having problems with crime, litter, and vacant housing at Time 1 predicted increases in depressive symptoms over a 9-month period.

A variety of mechanisms have been identified to explain associations between perceived neighborhood stress and depressive symptoms during early adolescence. Gilster (2014) tested a model based on neighborhood stress process theory, and findings indicated links between neighborhood stressors (e.g., neighborhood disorder, perceived neighborhood violence, neighborhood hazards, and limited neighborhood services) and youth depressive symptoms, with these associations explained in part by limited individual feelings of mastery. Others have implicated fear of crime as a mechanism linking residence in unsafe neighborhoods and depressive symptoms among adolescents (Pearson, Breetzke, & Ivory, 2015; Stevenson, 1998).

In addition to direct associations between neighborhood stressors and depressive symptoms, some researchers have documented that it is the combination of neighborhood stressors and vulnerability characteristics of adolescents that predict internalizing symptoms. For example, Rabinowitz, Drabick, and Reynolds (2016) reported that neighborhood stressors were associated with higher levels of internalizing symptoms only when adoles-cents were high in temperamental withdrawal. These findings elucidated one reason why neighborhood stressors represent a risk factor for some adolescents with internalizing problems, but not for others. Physiological characteristics of adolescents might also constitute a source of risk for internalizing symptoms if the youth are also exposed to neighborhood stressors. However, such a possibility has rarely been addressed within the literature connecting youth perceptions of neighborhood stress to their depressive symptoms.

Skin Conductance Reactivity

Dual-risk stress models posit that contextual stressors pose the greatest risk for youth who have physiological markers suggesting debilitating stress responses. Such models are consistent with the diathesis-stress model (Alloy, Hartlage, & Abramson, 1988), which suggests that the


likelihood of adjustment difficulties is due to both experiences of stressful events and preexisting vulnerability. However, a dual-risk model focuses on ways in which physiological functioning may explain individual differ-ences in vulnerability to environmental stressors. For example, El-Sheikh and Erath (2011) articulated a conceptual model considering ways in which family conflict and autonomic nervous system functioning work together to determine which children are the most likely to experience psychological problems. Similarly, Steinberg and Avenevoli (2000) pro-posed a modified version of the diathesis-stress perspective suggesting that individual differences in physiological reactivity to stress and expo-sure to contextual stressors work together to predict adjustment difficulties among children and adolescents.

A dual-risk perspective was used to generate hypotheses and interpret findings for the current study. Skin conductance reactivity is a physiological marker that is hypothesized, in this study, to interact with adolescent percep-tions of neighborhood stressors. We expect that the combination of greater reactivity and greater perceptions of neighborhood stress place adolescents at greater risk for higher levels of depressive symptoms. The sympathetic ner-vous system (SNS) is a branch of the autonomic nervous system (ANS) that is responsible for the “fight or flight” response in the presence of a perceived threat. Its role is to rapidly produce high levels of energy needed by the body during periods of challenge (Boucsein, 1992). When the SNS is activated, sweat gland activity increases, allowing the skin to better conduct electricity. The ability of the skin to conduct electricity is measured through electroder-mal activity (EDA), typically assessed by placing electrodes on two adjacent fingers of an individual and passing a small electric charge between elec-trodes, then measuring current flow. Skin conductance level (SCL) is linearly related to the rate at which sweat is produced.

Baseline SCL (SCLB) is measured during a time of relaxation—base-line—when an individual is not experiencing challenge. In contrast, SCL reactivity (SCLR) is measured during a period of challenge (typically a laboratory stressor), and is defined in terms of the change in SCL from baseline to challenge. SCLB and SCLR are both considered to represent stable differences among individuals (El-Sheikh, 2007). Higher levels of SCLR have been linked with higher levels of internalizing problems among children (El-Sheikh, 2005). It has been suggested that links between SCLR and depressive symptoms may be explained by brain mechanisms related to approach and avoidance (Gray, 1987), a premise that is supported by research linking higher levels of SCLR with behavioral inhibition and shy-ness (Kagan, Reznick, & Snidman, 1987), as well as internalizing symp-toms (El-Sheikh, 2005).


Interaction Between Neighborhood Stress and Skin Conductance Reactivity

The development of internalizing problems, and depressive symptoms, in particular, is the result of both biological and contextual factors (Thapar, Collishaw, Pine, & Thapar, 2012).

There is considerable evidence that (a) youth perceptions of neighborhood stress constitute a risk factor for depressive symptoms, and (b) physiological characteristics of youth represent a promising direction for understanding why children differ in terms of their vulnerability to perceived neighborhood stressors. Yet, to our knowledge, there has been no work considering the ways in which levels of SCLR during a simulated laboratory stressor may differentiate children who do versus do not experience depressive symptoms within the context of higher levels of perceived neighborhood stress. A num-ber of investigations have documented, however, that SCLR functions in this manner with respect to other types of contextual stressors.

Cummings, El-Sheikh, Kouros, and Keller (2007) assessed parents’ levels of depressive symptoms and children’s adjustment over a 2-year period. Of interest was whether children’s SCLR when exposed to adult arguments and engaged in problem-solving tasks would moderate associations between par-ents’ depressive symptoms and children’s experiences of internalizing behav-iors, externalizing behaviors, and social adjustment problems. Results supported a dual-risk perspective in that children with higher levels of SCLR were more vulnerable to parents’ depressive symptoms in terms of all three markers of child adjustment difficulties. El-Sheikh (2005) examined SCLR in response to a simulated argument as a potential moderator of associations between levels of parental reports of marital conflict and children’s external-izing, internalizing, and cognitive problems. For girls only, higher levels of SCLR increased vulnerability in terms of all three maladjustment outcomes. Once again, these findings were consistent with a dual-risk perspective.

The laboratory stressor in the above-referenced studies was child expo-sure to a simulated adult argument. It is unclear whether the same pattern of effects would emerge if youth were exposed to other types of stressors. In both the Cummings et al. (2007) and the El-Sheikh (2005) study, the simu-lated argument stressor was closely aligned with the contextual stressor under consideration in that both the laboratory stressor and the contextual stressor exposed youth to dysregulated adult behavior (conflict, depressive symp-toms). It is possible that it was not physiological stress reactivity in general that placed children in these studies at particular risk when also exposed to contextual stressors, but rather physiological reactivity to adult dysregulated behavior. Research involving laboratory stressors unrelated to the types of


contextual stressors being considered would clarify this issue. A laboratory-based public speaking task, or a task that focused on experiences of youth stress unrelated to adult behavior would permit exploration of whether physi-ological stress reactivity more broadly defined moderated associations between a range of contextual stressors and youth depressive symptoms. Public speaking is considered to be a stressful experience for individuals of all ages (Stein, Walker, & Forde, 1996).

Gender Differences

Several studies have indicated gender differences in the expression of inter-nalizing behaviors in response to neighborhood effects. Browning, Soller, Gardner, and Brooks-Gunn (2013) reported that social disorder within neigh-borhoods was associated with higher levels of internalizing symptoms among girls but not among boys. Similarly, Milam et al. (2012) found that greater perceptions of alcohol and drug use within neighborhoods was associated with higher levels of internalizing problems among girls, but not among boys. Conversely, neighborhood theories posit that adolescent boys will be impacted by neighborhood context to a greater extent than will girls, because boys are often subjected to less parental monitoring and have freedom to spend more time outside of the home (Fagan & Wright, 2012). However, problem behavior expression in response to neighborhood stressors may dif-fer for boys versus girls. Milam et al. (2012) suggested that boys may be more likely to express externalizing symptoms in response to neighborhood disorder, whereas girls respond to neighborhood disorder with internalizing symptoms, such as feelings of hopelessness and depression.

Research is inconsistent in terms of finding gender differences in the man-ner in which SCLR moderates associations between contextual stressors and child adjustment. Gender differences in associations have been found to depend on the type of SCL measured (SCLB, SCLR). El-Sheikh (2005) reported that associations between marital conflict and child maladjustment (internalizing, externalizing, cognitive problems) were significant only at higher levels of SCLR, and only for girls. Cummings et al. (2007) reported positive associations between parental depression and child maladjustment (internalizing, externalizing, social problems) only in the context of high SCLR—but these effects were present for both boys and girls. Erath and col-leagues (Erath, El-Sheikh, & Cummings, 2009) concluded that the role of SCLR as a moderator of associations between a contextual stressor (in their study, harsh parenting) and indicators of child adjustment (in their case, externalizing behavior) is more consistently present among boys than among girls. Thus, although there are conflicting perspectives regarding the ways in


which youth gender might shape associations among neighborhood stressors, youth physiological reactivity, and depressive symptoms, theory and research suggest that youth gender needs to be considered in analyses.

A study focused on understanding the interplay among youth perceptions of neighborhood stress, internalizing problems, and SCLR must consider the possibility that the nature of associations among these variables may differ for boys and girls.

The Current Study

Within the current study, we consider SCLR in response to a laboratory stressor as a potential moderator of associations between youth perceptions of neighborhood stress and their levels of depressive symptoms. Drawing upon a dual-risk perspective, we hypothesize that greater perceived neighborhood stress predicts more depressive symptoms only for youth who are high in SCLR. Given the extensive literature indicating gender differences in depres-sive symptomatology, as well as the impact of biological stress reactivity, we consider gender as a potential moderator of effects. However, given the incon-sistent findings regarding role of gender in moderating such effects, we make no specific hypotheses regarding the nature of possible gender differences.

Method

Participants

A total of 68 mother-youth dyads participated in this study. Youth were recruited during fifth grade from two elementary schools in the southeastern United States. Participation began the following school year when the chil-dren were in sixth grade. The youth sample was 56% female (n = 38) and 44% male (n = 30). Children identified themselves as European American (n = 37; 54%) and African American (n = 28; 41%), with one child each self-identifying as Hispanic, Native American, and Asian (not Hmong). In total, 43 participating mothers (63%) indicated that they were married, and 25 mothers (37%) indicated that they were single. The Economic Pressure Scale (which yields scores ranging from 1 to 5; Conger & Elder, 1994) was used as an indicator of economic well-being with scores in our sample ranging from 1.38 to 5 with a mean of 3.49 (SD = .85).

Procedures

Data collection occurred in a university research laboratory from 2011-2012. Upon their arrival, electrodes measuring SCL were attached to the


adolescents’ hands using small straps wrapped around two of their fingers. During a 10-minute acclimation period, youth wore the SCL finger bands while they completed questionnaires. Baseline SCL data were then collected during a 2-minute period during which the youth sat quietly at the computer wearing earphones and watching a relaxation video. Next, the youth partici-pated in the Trier Social Stress Test for Children (TSST-C; Kirschbaum, Pirke, & Hellhammer, 1993), which involved a public speaking task and a math activity. Youth were asked to construct a speech about a time in their lives that they felt left out, rejected, or excluded by a peer group—or when they had made another feel that way. Youth spent 3 minutes preparing the speech and 5 minutes performing the speech in front of an audience. The audience consisted of a video camera, the adolescents’ mother, and two expressionless research assistants. Mothers were instructed to maintain a neutral expression during the performance. Finally, youth participated in a mental arithmetic task in which they were asked to count backwards from 758 by sevens for 4 minutes in front of the same audience. At the conclusion of the math task, SCL cuffs were removed from adolescents’ hands while adolescents completed additional questionnaires. All study procedures were approved by the university’s Institutional Review Board. Mothers provided written consent for their own and youth participation, and youth provided written assent to participate.

Measures

Potential control variables. Youth sex (0 = female; 1 = male) and ethnicity (0 = European American; 1 = ethnic minority) were self-reported by youth. Socioeconomic status was assessed through mother completion of the 8-item Economic Pressure Scale (Conger, Ge, Elder, Lorenz, & Simons, 1994). The Economic Pressure Scale is designed to measure the perceived extent of fam-ilies’ financial pressure. Mothers were asked to rate their agreement with statements about whether their family could afford expenses such as “cloth-ing we should have,” and “leisure and recreational activities we want to par-ticipate in” on a scale of 1 (strongly disagree) to 5 (strongly agree). Items were averaged to create a summary score in which higher scores indicated lower perceived economic pressures (α = .89).

Adolescents reported on their physical development using the Pubertal Development Scale (PDS; Petersen, Crockett, Richards, & Boxer, 1988). Using the 5-item scale (some items were different based on sex), adolescents were asked to rate their development on a scale of 1 (has not started to develop) to 4 (development seems completed). The item scores were averaged to yield a summary score for which higher scores indicated higher levels of


physical pubertal development. Adolescents were asked, “Did you have any caffeine today?” (0 = no; 1 = yes). They were also asked to rate their overall physical health on a scale of 1 (very poor) to 5 (excellent).

Neighborhood stressors. To develop a measure of neighborhood stressors, we created a weighted composite of three measures of adolescent perceptions of neighborhood quality. The extent to which youth perceived negative neigh-borhood influences was evaluated using a 5-item subscale from the Neigh-borhood Quality Index (Cook, Herman, Phillips, & Settersten, 2002). A sample item is “You want to get away from this neighborhood as soon as possible.” Adolescents were asked to rate their agreement with statements on a scale from 1 (strongly disagree) to 5 (strongly agree). Items were averaged, and higher scores on this measure indicated more negative perceptions of neighborhood. One of the items was dropped after performing a preliminary factor analysis, leaving a total of 4 items for the scale as used in this study (α = .77). Adolescents’ perceptions of neighborhood social cohesion were assessed using a subscale from the Neighborhood Quality Index (Cook et al., 2002). This 4-item scale asked participants to rate their agreement with state-ments such as, “There are a lot of adults around here that I can look up to,” on a scale of 1 (strongly agree) to 5 (strongly disagree). Items were averaged to yield a summary score on which higher scores indicated less social cohe-sion (α = .79). Youth reported on their perceptions of neighborhood connect-edness using a 6-item subscale of the Neighborhood Cohesion Scale (Aber & Seidman, 1993; Adolescent Pathways Project, 1992). Adolescents were asked to rate their agreement with items such as “My neighborhood is a better place to live than other neighborhoods” on a scale of 1 (very true) to 4 (not at all true). The items were averaged to create a summary score for which higher scores indicated lower neighborhood connectedness (α = .79).

We transformed summary scores for all three measures of neighborhood stress to z scores. We then used structural equation modeling (SEM) in SPSS Amos 20.0 to test a confirmatory measurement model loading the three neighborhood stress z-scores scores onto a single latent factor. We then used factor loadings from this analysis to create a composite neighbor-hood stress variable that was used in subsequent analyses. The factor load-ings were .93 for negative influences, .75 for lower cohesion, and .87 for lower connectedness.

Skin conductance reactivity. SCL was measured using skin surface electrodes filled with a gel medium to capture the amount of sweat produced on the surface of the skin. SCL (units = microSiemens; μSiemens) was measured using UFI’s 3991x-GPP General Purpose Psychophysiological BioLog,


model 1081FG, with a sampling rate of 10 Hz. Before applying the elec-trodes, two sites on the youths’ fingers were wiped with alcohol pads in order to remove sweat or skin oils from the area. The laboratory environment was held at a standard room temperature to avoid impacting the amount of sweat produced. Next, the skin surface electrodes were attached to two of the ado-lescents’ finger pads at the distal phalange on the non-dominant hand, as recommended by Scerbo, Freedman, Raine, Dawson, and Venables (1992). A constant, low voltage current of 0.5 volts was passed between electrodes throughout the procedure as a bias (i.e., reference) point. An input connector ran from the electrodes to the BioLog. The experimenter had a key fob linked with the BioLog that was used to record event marks; specifically, the experi-menter pressed a button that communicated to the BioLog when an “event” started and ended. These events represented the different sections of the TSST-C (preparation, speech, math). SCL data were downloaded from the BioLog to a computer using BioLog GPP DPS software. Event markers were used to parse the data stream into task episodes. The mean SCL for each epi-sode was then calculated. Reactivity was calculated by subtracting average SCL during baseline from average SCL during each task episode (Erath et al., 2009; Erath, El-Sheikh, Hinnant, & Cummings, 2011). Thus, the SCLR score represents the change in skin conductance level from baseline to stress-inducing episode. Higher SCLR scores signify increases in SCL values from the baseline to the stress task. Evidence that the TSST-C was effective in producing a physiological response in the current study was demonstrated by the percentage of adolescents who experienced an increase in SCL from base-line to stress-inducing episodes, which was 91%, 94%, and 92% for the instructions, speech, and math episodes, respectively.

Participants also completed a subjective measure of negative affect before and after completion of the TSST-C, which indicated that 82% of participants self-reported more negative affect after completion of the TSST-C than they did before. Finally, our measure of SCLR was significantly different from 0 for the full sample, t(65) = 8.78, p < .01 for SCLR, for girls, t(36) = 7.03, p < .01, and for boys, t(28) = 4.70, p < .01.

Bivariate correlations among SCLR for the three task segments ranged from .50 to .81. To generate a single measure of SCLR across episodes, we used SEM in SPSS Amos to test a confirmatory measurement model loading the three SCLR scores (preparation, speech, math) onto a single latent factor. As such, one score was derived from comparing baseline SCL with SCL dur-ing the preparation period. The second score was derived from comparing baseline SCL with SCL during the speech challenge. The third score was derived from comparing baseline SCL with SCL during the math challenge. We then used factor loadings from this analysis to create a composite SCLR


variable that was used in subsequent analyses. The standardized factor load-ings were .75 for the preparation episode, 1.00 for the speaking episode, and .69 for the math episode.

Youth depressive symptoms. Youth depressive symptoms were assessed using the 10-item (short form) Children’s Depression Inventory (Kovacs, 1992) designed to measure cognitive, affective, and behavioral indicators of depres-sion in children and adolescents. For each item, youth were asked to choose the response that most described how they had felt during the past 2 weeks. For example, one item contained the responses (1) sad once in a while, (2) sad many times, (3) sad all the time. Items were averaged to create a total depressive symptoms score for which higher scores indicated more depres-sive symptoms (α = .82).

Analytic Strategy

To determine whether SCLR moderated associations between neighbor-hood stress and youth depressive symptoms, we adopted a multi-step approach that included all of the following. First, we used multiple regres-sion entering neighborhood stress, SCLR, and all potential control vari-ables as predictors of depressive symptoms. All control variables that were not significantly associated with depressive symptoms were eliminated from further analyses. Second, we performed regression analyses using SPSS Amos 20 to test a path model predicting depressive symptoms from baseline SCL, additional control variables identified in initial regression analyses, neighborhood stress, the composite SCLR variable, and the interaction between neighborhood stress and SCLR. Amos was used to conduct multiple-group analyses to determine whether this model differed from one in which paths were constrained to be the same for boys versus girls. If a chi-squared difference test indicated gender differences, critical ratios were used to identify structural paths that were significantly differ-ent for boys and girls. Finally, for cases in which significant interaction terms indicated a moderating effect of SCLR, these effects were probed using tests of simple slopes and regions of significance estimations (using the Johnson-Neyman technique). For main effects, we used a probability level of .05 as a cutoff for determining statistical significance. Whisman and McClelland (2005) have documented the increased level of power needed to detect interaction effects in non-experimental research and have recommend that research involving such interaction effects use higher probability levels so that potentially important patterns are not missed. Given our relatively small sample size, such an approach was particularly


appropriate. Accordingly, we used a probability value of .10 as a cutoff for determining the statistical significance of interaction effects and when probing the interactions.

Results

Preliminary Analyses

We performed a regression analysis to identify controls needed in the focal analyses. We predicted youth depressive symptoms from the composite neighborhood stress variable and the composite SCLR variable, controlling for ethnicity, economic pressure, pubertal development, caffeine consump-tion, self-rated physical health, and baseline RSA. None of these potential control variables was significantly associated with depressive symptoms. However, we decided to include baseline SCL as a control in subsequent analyses due to conventions within the field (Hinnant, Erath, Tu, & El-Sheikh, 2016). No other control variables were included in subsequent analyses.

Intercorrelations Among Model Variables and Gender Differences

Descriptive statistics and bivariate correlation coefficients for all model vari-ables are presented in Table 1 separately by gender. Among girls, higher SCL at baseline was associated with higher scores on the composite SCLR vari-able, r(38) = .48, p < .01. Among boys, there were no significant bivariate correlations. We also examined bivariate correlation coefficients for associa-tions between levels of depressive symptoms and SCLR scores for each com-ponent of the stress task and found that none were statistically significant.

T tests were conducted to test for gender differences in mean levels of all variables. Only one such difference emerged. Girls had higher levels of depressive symptoms than did boys, t(64.60) = 2.50, p = .01.

Associations Among Neighborhood Stress, Skin Conductance Level Reactivity, and Youth Depressive Symptoms

We used SPSS Amos 20 to conduct regression analyses testing whether youth perceptions of neighborhood stress (centered), SCLR (centered), and the interaction between the two, predicted youth depressive symptoms, control-ling for baseline SCL. Higher levels of perceived neighborhood stress pre-dicted higher levels of depressive symptoms, b = .04, β = .22, p = .04, 95% confidence interval (CI) = [–1.940, 2.020]. Contrary to the hypothesis, there


was not a significant interaction between perceived neighborhood stress and SCLR, b = .005, β = .09, p = .39, 95% CI = [–.004, .014].

We conducted a multi-group analysis comparing this model with a model in which the regression paths were constrained to equality across boys versus girls. The chi-square difference test comparing these models was significant, χ2(4) = 9.71, p = .04, indicating that there were gender differences in the strength of regression coefficients within the model. We then used critical ratios to identify structural paths that were significantly different for boys versus girls. There were no gender differences for the association between baseline SCL and depressive symptoms (significant for the full sample) or for the association between SCLR and depressive symptoms for the full sample (not significant for the full sample). However, for the association between neighborhood stress and depressive symptoms, and for the association between the interaction term and depressive symptoms, gender differences were such that paths were only significant for boys. For boys, higher neigh-borhood stress was associated with more depressive symptoms, b = .08, β = .60, p < .01, 95% CI = [.046, .125] (girls’ b = .01, β = .03, p =.55, 95% CI = [–.037, .071]). For boys, the interaction of neighborhood stress and SCLR was significant, b = .02, β = .68, p < .01, 95% CI = [–.1.962, 1.988] (girls’

Table 1. Descriptive Statistics and Intercorrelations Among Study Variables.

Variable 1 2 3 4 5 6 7

1. Skin Conductance—Preparation — .88** .64** .50* .90** −.15 −.202. Skin Conductance—Speech .76** — .83** .40* .95** −.13 −.043. Skin Conductance—Math .44* .75** — .43* .86** −.17 −.044. Baseline Skin Conductance .38* .37* −.07 — .49* −.14 −.29†

5. Skin Conductance Reactivity .79** .96** .84** .27 — −.16 −.136. Neighborhood Stress −.03 −.17 −.15 .19 −.11 — .147. Youth Depressive Symptoms .18 .20 .20 −.07 .18 .23 —Mean: Girls 1.75 2.28 2.52 5.20 6.97 5.99 1.34 Boys 1.30 2.21 2.47 5.46 6.12 5.69 1.16SD: Girls 1.32 1.42 1.45 3.31 4.33 1.95 .35 Boys 1.48 1.82 2.55 4.56 5.16 1.66 .24N: Girls 38 38 37 38 38 38 38 Boys 28 28 29 29 29 29 30

Note. Correlations for girls are above the diagonal; correlations for boys are below the diagonal.†p < .10. *p < .05. **p < .01.


b = −.00, β = −.04, p =.78, 95% CI = [–.017, .013]). We then probed the interaction between neighborhood stress and SCLR in the sample of boys.

This interaction was probed using tests of simple slopes as recommended by Aiken, West, and Reno (1991). Figure 1 shows the simple slopes graphed one standard deviation above and below the mean of SCLR. For boys with low SCLR, levels of perceived neighborhood stress were unassociated with depressive symptoms, b = −.02, β = −.18, t(25) = −.86, p = .39, 95% CI = [–.084, .032]. For boys with higher SCLR, higher levels of perceived neigh-borhood stress were associated with more depressive symptoms, b = .19, β = 1.38, t(25) = 3.89, p < .01, 95% CI = [.294, .098]. In other words, high SCLR placed boys at risk for the negative effects (in terms of depressive symptoms) of neighborhood stress.

To allow for a more detailed understanding of the way in which SCLR moderated associations between neighborhood stress and depressive symp-toms, regions of significance were estimated using the Johnson-Neyman technique. Results of this analysis are depicted graphically in Figure 2, which displays the 95% confidence bands for the association between levels of SCLR and the strength of associations between neighborhood stress and depressive symptoms among boys. All variables were converted to z scores in conducting this analysis for ease of interpretation. A significant association between neighborhood stress and depressive symptoms is present when the

Figure 1. Associations between perceived neighborhood stress and boys’ depressive symptoms.Note. SCLR = skin conductance level reactivity.*Slope significantly different from 0; p ⩽ .01.


value of 0 for this association falls outside of the confidence bands. This more careful examination of moderation effects indicated that greater percep-tions of neighborhood stress were associated with more depressive symptoms for boys scoring above –.10 on SCLR (mean centered). The significant region represented the top 51.72% of boys in the sample.

Discussion

In an attempt to understand variability in the extent to which perceptions of stressful neighborhoods are associated with depressive symptoms in early adolescence, we considered SCLR as a potential moderator of such associa-tions. Among all youth, higher levels of perceived neighborhood stress were associated with more depressive symptoms; however, such associations were particularly strong for boys. Among boys only, the association between

Figure 2. Confidence bands for observed sample values of skin conductance level reactivity.Note. SCLR values are centered. SCLR = skin conductance level reactivity.


perceived neighborhood stress and depressive symptoms was moderated by levels of SCLR during a laboratory-based public speaking task. For boys with average or above average levels of SCLR, perceptions of greater neigh-borhood stress were associated with more depressive symptoms.

In general, empirical research has indicated that adolescents who per-ceive their neighborhoods are more dangerous, or less cohesive, are at risk for higher levels of depressive symptoms (Ford & Rechel, 2012; Latkin & Curry, 2003). Our findings were consistent with this work, as we found that neighborhoods perceived as stressful by adolescents of both genders placed them at risk for depressive symptoms. However, boys appeared to be par-ticularly at risk in this respect. It has been suggested (Fagan & Wright, 2012) that boys may be more likely to experience problematic adjustment (including depression) when they reside in unsafe neighborhoods, given that boys spend more time unsupervised in the neighborhood context than do girls. Other researchers (Milam et al., 2012) have suggested that contex-tual stressors of all sorts (including neighborhoods) are more likely to be linked with externalizing behaviors among boys but internalizing symp-toms among girls. Given that levels of depression are lower among boys than among girls in adolescence (Saluja et al., 2004), the finding that per-ceived neighborhood stress was strongly linked with depressive symptoms among boys is noteworthy. Although girls are more susceptible to internal-izing problems in general, it may be that the presence of this particular contextual stressor—neighborhood problems—is especially relevant in eliciting symptoms of depression among boys.

Of primary interest in the current study was whether associations between perceptions of neighborhood stress and levels of depressive symptoms among youth varied based on levels of physiological reactivity to stress (SCLR). SCLR moderated associations between perceptions of neighbor-hood stress and boys’ depressive symptoms. Among boys, a specific set of contextual circumstances, combined with biological predispositions, increased vulnerability in this area. Interestingly, “vulnerability” for our sample was defined in terms of average or above average levels of SCLR. Specifically, higher levels of neighborhood stress were associated with more depressive symptoms problems for boys in the top 51.72% of the range of SCLR scores. Our findings are consistent with previous work (El-Sheikh, 2005) indicating that higher levels of SCLR constitute a risk factor for inter-nalizing problems. These findings extend the understanding of ways in which high SCLR places youth at risk when exposed to other types of con-textual stressors (perceptions of neighborhood stress)—and highlight the importance of considering how the moderating role of SCLR may vary for boys versus girls across these different contexts.


The results of this study suggest that the way in which the SNS responds to a laboratory stressor involving a public speaking task that is unrelated to exposure to dysregulated adult behavior plays a role in determining which individuals are, or are not, negatively impacted (in terms of reports of inter-nalizing symptoms) by exposure to dangerous and disconnected neighbor-hoods of residence. As intriguing as these findings may be, we note that the current study is not without limitations. These include a small sample size, a lack of ethnic diversity within the sample, and the artificial nature of our laboratory stressor.

Though a sample of 68 is limited in its power to detect effects, and mod-erator effects in particular, we were able to identify the moderating role that SCLR played with respect to associations between neighborhood stress and levels of internalizing problems among boys. These findings speak to the magnitude of this effect size and suggest that SCLR is important to consider when attempting to understand variability in boys’ responses to living in neighborhoods characterized by danger and lack of connectedness.

Our sample was also limited in that it was comprised predominantly of youth from two ethnic groups: African American and European American. In light of the increasing ethnic diversity in the United States, it is all the more important that researchers recruit samples that reflect this diversity. This was a challenge within the current study, given that the two most prev-alent ethnic groups within the geographic region in which this study was conducted were African American and European American. Future research in this area might target geographic areas that are more ethnically diverse—or that contain larger percentages of specific ethnic groups not represented in the current study.

Another limitation of the current study is its reliance on a laboratory task to elicit stress responses from youth. Such tasks are clearly artificial. However, the Trier Stress Task (Buske-Kirschbaum et al., 1997) is widely accepted as an effective way to induce moderate levels of stress among young adoles-cents. Engaging in a public speaking task in a strange environment might actually be more stress-inducing than engaging in a similar task in a more comfortable and familiar setting. Importantly, our adaptation of the Trier Stress Task required that youth prepare and present a speech about a poten-tially stressful experience in their past that was unrelated to experiences of neighborhood stress. The fact that this task raised levels of skin conductance among our participants, and that the magnitude of that increase was observed to moderate associations between experiences of neighborhood stress and internalizing problems, gives us confidence that, although not perfect, this laboratory stress task was effective in inducing moderate levels of stress among participants.


Findings from the current study highlight the individual and interactive roles of environment and physiology in relation to levels of depressive symp-tomatology in early adolescence. More importantly, they shed light on why the experience of living in dangerous neighborhoods that lack cohesion and connectedness is not uniformly associated with depressive symptoms during this developmental period. Neighborhood risk matters, as does baseline lev-els of SNS activation and levels of SNS reactivity during times of stress. Even more intriguing, these findings represent an important step in under-standing variability in the extent to which such associations are observed. Boys appear to be at particular risk when exposed to neighborhood risk; boys with high levels of SNS reactivity are even more susceptible to such effects. The potential ability to customize prevention programs based on context, gender, and physiological characteristics of young adolescents represents an intriguing glimpse into what prevention programs may look like in the future.

Acknowledgment

We appreciate the statistical advice from Jeffrey Labban.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by grants to Cheryl Buehler and Anne C. Fletcher from the University of North Carolina at Greensboro and to Christy M. Buchanan from Wake Forest University. Bridget Weymouth’s contribution to this manuscript was supported by the National Institute on Drug Abuse (NIDA) of the National Institutes of Health under award numbers P50 DA039838 and T32DA017629. The content is solely the responsibility of the authors and does not necessarily represent the official views of NIDA or the National Institutes of Health.

References

Aber, J., & Seidman, E. (1993, March). Are neighborhoods determinants, mediators, and/or moderators of urban adolescent development? Paper presented at the meetings of the Society for Research on Child Development, New Orleans, LA.

Adolescent Pathways Project. (1992). Neighborhood cohesion scale: Psychometric development summary. New York: New York University.

Aiken, L. S., West, S. G, & Reno, R. R. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.


Alloy, L. B., Hartlage, S., & Abramson, L. Y. (1988). Testing the cognitive diathesis-stress theories of depression: Issues of research design, conceptualization, and assessment. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 31-73). New York, NY: Guilford Press.

American Psychiatric Association. (1987). Diagnostic and statistical manual of men-tal disorders (3rd ed., revised). Washington, DC: Author.

Balázs, J., Miklósi, M., Keresztény, Á., Hoven, C. W., Carli, V., Wasserman, C., . . . Wasserman, D. (2013). Adolescent subthreshold-depression and anxiety: Psychopathology, functional impairment and increased suicide risk. The Journal of Child Psychology and Psychiatry, 54, 670-677. doi:10.1111/jcpp.12016

Bertha, E. A., & Balázs, J. (2013). Subthreshold depression in adolescence: A system-atic review. European Child & Adolescent Psychiatry, 22, 589-603. doi:10.1007/s00787-013-0411-0

Bhatia, S. K., & Bhatia, S. C. (2007). Childhood and adolescent depression. American Family Physician, 75, 73-80.

Bongers, I. L., Koot, H. M., van der Ende, J., & Verhulst, F. C. (2003). The normative development of child and adolescent problem behavior. Journal of Abnormal Psychology, 112, 179-192. doi:10.1037/0021-843X.112.2.179

Boucsein, W. (1992). Electrodermal activity. New York, NY: Plenum Press.Browning, C. R., Burrington, L. A., Leventhal, T., & Brooks-Gunn, J. (2008).

Neighborhood structural inequality, collective efficacy, and sexual risk behav-ior among urban youth. Journal of Health and Social Behavior, 49, 269-285. doi:10.1177/002214650804900303

Browning, C. R., Soller, B., Gardner, M., & Brooks-Gunn, J. (2013). “Feeling dis-order” as a comparative and contingent process: Gender, neighborhood con-ditions, and adolescent mental health. Journal of Health and Social Behavior, 54, 296-314.

Buske-Kirschbaum, A., Jobst, S., Wustmans, A., Kirschbaum, C., Rauh, W., & Hellhammer, D. (1997). Attenuated free cortisol response to psychoso-cial stress in children with atopic dermatitis. Psychosomatic Medicine, 59, 419-426.

Byrd, A. L., Loeber, R., & Pardini, D. A. (2012). Understanding desisting and per-sisting forms of delinquency: The unique contributions of disruptive behavior disorders and interpersonal callousness. The Journal of Child Psychology and Psychiatry, 53, 371-380. doi:10.1111/j.1469-7610.2011.02504.x

Conger, R. D., & Elder, G. H. (1994). Families in troubled times: Adapting to change in rural America. Hawthorne, NY: Aldine de Gruyter.

Conger, R. D., Ge, X., Elder, G. H., Jr., Lorenz, F. O., & Simons, R. L. (1994). Economic stress, coercive family process, and developmental problems of ado-lescents. Child Development, 65, 541-561. doi:10.2307/1131401

Cook, T. D., Herman, M. R., Phillips, M., & Settersten, R. A., Jr. (2002). Some ways in which neighborhoods, nuclear families, friendship groups, and schools jointly affect changes in early adolescent development. Child Development, 73, 1283-1309. doi:10.1111/1467-8624.00472


Cummings, E. M., El-Sheikh, C. D., Kouros, C. D., & Keller, P. S. (2007). Children’s skin conductance reactivity as a mechanism of risk in the context of parental depressive symptoms. The Journal of Child Psychology and Psychiatry, 48, 436-445. doi:10.1111/j.1469-7610.2006.01713.x

Delany-Brumsey, A., Mays, V. M., & Cochran, S. D. (2014). Does neighborhood social capital buffer the effects of maternal depression on adolescent behav-ior problems? American Journal of Community Psychology, 53, 275-285. doi:10.1007/s10464-014-9640-8

El-Sheikh, M. (2005). The role of emotional responses and physiological reactivity in the marital conflict-child functioning link. The Journal of Child Psychology and Psychiatry, 46, 1191-1199. doi:10.1111/j.1469-7610.2005.01418.x

El-Sheikh, M. (2007). Children’s skin conductance level and reactivity: Are these measures stable over time and across tasks? Developmental Psychobiology, 49, 180-186. doi:10.1002/dev.20171

El-Sheikh, M., & Erath, S. A. (2011). Family conflict, autonomic nervous system functioning, and child adaptation: State of the science and future directions. Development and Psychopathology, 23, 703-721.

Erath, S. A., El-Sheikh, M., & Cummings, E. M. (2009). Harsh parenting and child externalizing behavior: Skin conductance level reactivity as a moderator. Child Development, 80, 578-592. doi:10.1111/j.1467-8624.2009.01280.x

Erath, S. A., El-Sheikh, M., Hinnant, J. B., & Cummings, E. M. (2011). Skin con-ductance level reactivity moderates the association between harsh parenting and growth in child externalizing behavior. Developmental Psychology, 47, 693-706. doi:10.1037/a0021909

Fagan, A. A., & Wright, E. M. (2012). The effects of neighborhood context on youth violence and delinquency: Does gender matter? Youth Violence and Juvenile Justice, 10, 64-82. doi:10.1177/1541204011422086

Ford, J. L., & Rechel, M. (2012). Parental perceptions of the neighborhood context and adolescent depression. Public Health Nursing, 29, 390-402. doi:10.1111/j.1525-1446.2012.01015.x

Gariepy, G., Thombs, B. D., Kestens, Y., Kaufman, J. S., Blair, A., & Schmitz, N. (2015). The neighbourhood built environment and trajectories of depression symptom episodes in adults: A latent class growth analysis. PLoS ONE, 10, e0133603. doi:10.1371/journal.pone.0133603

Gilster, M. E. (2014). Neighborhood stressors, mastery, and depressive symptoms: Racial and ethnic differences in an ecological model of the stress process in Chicago. Journal of Urban Health, 91, 690-706. doi:10.1007/s11524-014-9877-4

Gray, J. A. (1987). The psychology of fear and stress. New York, NY: Cambridge University Press.

Hallfors, D. D., Waller, M. W., Ford, C. A., Halpern, C. T., Brodish, P. H., & Iritani, B. (2004). Adolescent depression and suicide risk: Association with sex and drug behavior. American Journal of Preventive Medicine, 27, 224-231. doi:10.1016/j.amepre.2004.06.001

Hinnant, J. B., Erath, S. A., Tu, K. M., & El-Sheikh, M. (2016). Permissive parenting, deviant peer affiliations, and delinquent behavior in adolescence: The moderat-


ing role of sympathetic nervous system reactivity. Journal of Abnormal Child Psychology, 44, 1071-1081. doi:10.1007/s10802-015-0114-8

Holder, M. K., & Blaustein, J. D. (2014). Puberty and adolescence as a time of vulnerability to stressors that alter neurobehavioral processes. Frontiers in Neuroendocrinology, 35, 89-110. doi:10.1016/j.yfrne.2013.10.004

Kagan, J., Reznick, J., & Snidman, N. (1987). The physiology and psychol-ogy of behavioral inhibition in children. Child Development, 58, 1459-1473. doi:10.2307/1130685

Kim, J. (2010). Neighborhood disadvantage and mental health: The role of neighbor-hood disorder and social relationships. Social Science Research, 39, 260-271.

Kirschbaum, C., Pirke, K., & Hellhammer, D. H. (1993). The “Trier Social Stress Test”—A tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology, 28, 76-81. doi:10.1159/000119004

Kohen, D. E., Leventhal, T., Dahinten, V. S., & McIntosh, C. N. (2008). Neighborhood disadvantage: Pathways of effects for young children. Child Development, 79, 156-169. doi:10.1111/j.1467-8624.2007.01117.x

Kovacs, M. (1992). Children’s Depression Inventory (CDI). New York, NY: Multi-Health Systems.

Latkin, C. A., & Curry, A. D. (2003). Stressful neighborhoods and depression: A prospective study of the impact of neighborhood disorder. Journal of Health and Social Behavior, 44, 34-44. doi:10.2307/1519814

Mair, C., Diez Roux, A. V., & Galea, S. (2008). Are neighbourhood charac-teristics associated with depressive symptoms? A review of evidence. Journal of Epidemiology & Community Health, 62, 940-946. doi:10.1136/jech.2007.066605

Milam, A. J., Furr-Holden, C., Whitaker, D., Smart, M., Leaf, P., & Cooley-Strickland, M. (2012). Neighborhood environment and internalizing problems in African American children. Community Mental Health Journal, 48, 39-44. doi:10.1007/s10597-011-9379-8

Pearson, A. L., Breetzke, G., & Ivory, V. (2015). The effect of neighborhood recorded crime on fear: Does neighborhood social context matter?. American Journal of Community Psychology, 56, 170-179.

Petersen, A. C., Crockett, L., Richards, M., & Boxer, A. (1988). A self-report measure of pubertal status: Reliability, validity, and initial norms. Journal of Youth and Adolescence, 17, 117-133. doi:10.1007/BF01537962

Rabinowitz, J. A., Drabick, D. A., & Reynolds, M. D. (2016). Youth withdrawal moderates the relationships between neighborhood factors and internalizing symptoms in adolescence. Journal of Youth and Adolescence, 45, 427-439. doi:10.1007/s10964-015-0324-y

Ross, C. E., & Mirowsky, J. (2009). Neighborhood disorder, subjective alien-ation, and distress. Journal of Health and Social Behavior, 50, 49-64. doi:10.1177/002214650905000104

Saluja, G., Iachan, R., Scheidt, P. C., Overpeck, M. D., Sun, W., & Giedd, J. N. (2004). Prevalence of and risk factors for depressive symptoms among young adolescents. Archives of Pediatrics & Adolescent Medicine, 158, 760-765. doi:10.1001/archpedi.158.8.760


Scerbo, A. S., Freedman, L. W., Raine, A., Dawson, M. E., & Venables, P. H. (1992). A major effect of recording site on measurement of electrodermal activity. Psychophysiology, 29, 241-246. doi:10.1111/j.1469-8986.1992.tb01693.x

Stein, M. B., Walker, J. R., & Forde, D. R. (1996). Public-speaking fears in a community sample: Prevalence, impact on functioning, and diagnostic clas-sification. Archives of General Psychiatry, 53, 169-174. doi:10.1001/arch-psyc.1996.01830020087010

Steinberg, L., & Avenevoli, S. (2000). The role of context in the development of psychopathology: A conceptual framework and some speculative propositions. Child Development, 71, 66-74. doi:10.1111/1467-8624.00119

Stevenson, H. C. (1998). Raising safe villages: Cultural-ecological factors that influ-ence the emotional adjustment of adolescents. Journal of Black Psychology, 24, 44-59.

Thapar, A., Collishaw, S., Pine, D. S., & Thapar, A. K. (2012). Depression in adoles-cence. The Lancet, 379, 1056-1067. doi:10.1016/S0140-6736(11)60871-4

Whisman, M. A., & McClelland, G. H. (2005). Designing, testing, and interpret-ing interactions and moderator effects in family research. Journal of Family Psychology, 19, 111-120. doi:10.1037/0893-3200.19.1.111

Author Biographies

Anne C. Fletcher is a professor in the department of Human Development and Family Studies at the University of North Carolina at Greensboro. She studies parenting and parent-child relationships across ages ranging from middle child-hood through adolescence, as well as physiological stress reactivity in early ado-lescence. Her research program involves qualitative, quantitative, and mixed meth-ods approaches.

Cheryl Buehler studies family processes, particularly in relation to children’s and adolescent’s adjustment and well-being. Her focus on family process includes marital interactions, parenting, and coparenting. She also looks at how cumulative risk in families and other socialization environments affect child and adolescent well-being, and factors that help provide needed protection.

Amy L. McCurdy is a master’s student enrolled in the department of Human Development and Family Studies at the University of North Carolina at Greensboro. Her professional interests broadly center around parenting and the conditions that affect parenting practices. She is also interested in parent-child relationships and family transitions within diverse family systems, especially in the context of adoption.

Bridget B. Weymouth is a postdoctoral fellow at The Pennsylvania State University. Her research focuses on adolescent development in the family context. Her specific research interests include family risk processes for adolescent social anxiety symp-toms, social anxiety as a risk mechanism for adolescent substance use, and parent-adolescent conflict.

skin conductance reactivity as a moderator of associations ... · anne c. fletcher 1, cheryl...

Documents