Preliminary Analyses. At pretreatment the mean CES-D score for the samples was 24.71 (SD = 9.82). Mental health characteristics are summarized in Table 1. In both IC and WLC approximately two-thirds of the participants met the criteria for at least one axis I disorder. The most prevalent disorders were MDD (39%) and anxiety disorders (45%). Previous MDD was reported by 69% of the participants. There were no significant differences (all p
Preliminary Analyses. Because it was hypothesized that changes in perception of health status (as assessed with the SAS, HA subscale of the IAS and RQ) are correlated, the changes on these variables were transformed in one principal component accounting for as much of the variability in the data as possible. To this end a principal component analysis (PCA) was conducted on the 12 week follow-up residualised change scores on these measures (obtained by statistically correcting the follow-up scores for any baseline differences on these measures). Next, using the regression method a composite fac- tor score for change in perception of current health problems was calculated. The PCA on the residualised gain scores on the SAS, HA subscale and RQ at fol- low-up clearly yielded a one-factor solution (eigenvalue 1.75) accounting for 58.4% of the variance. Factor loadings were high (respectively 0.70, 0.82 and 0.77). Since there were no significant differences between the major non-western groups in change score in perception of current health problems (data not shown), we decided to operationalise ethnicity as western versus non-western. The ethnic difference in 113 changes in perception of health status was significant (t(295): -3.53, p<.001) and had a moderate effect size (d = 0.38). In Table 1 an overview is presented of characteristics of participants with a west- ern or non-western ethnicity. Except for gender, all demographic variables show a significant difference between residents of a western or non-western ethnicity. In addition, non-western residents reported more symptoms of fatigue, psychopatholo- gy, and post-traumatic stress and less health-related quality of life compared to west- ern participants at baseline.
Preliminary Analyses. The measure utilized in this study contained three vignettes that were diagnosed by participants. Each vignette was assigned a point value dependent on the diagnosis provided by the participant. The point values for each vignette ranged from 0 to 2 points. Two points were given for a diagnosis that matched the correct diagnosis for the vignette. One point was given to diagnoses that were in the same DSM-5 category as the correct diagnosis. Zero points were given to all other diagnoses. This scoring system yielded a cumulative score with a range of 0 to 6 points. The overall average cumulative score was 3.17 points while the standard deviation was 1.38 points. Table 4 Mean Cumulative Scores by Discipline Discipline Mean Standard Deviation Overall 3.17 1.38 Social Workers 3.13 1.31 Counselors 3.23 1.36 Hypothesis Testing To test the hypothesis that a significant difference existed between psychologists, counselors, and social workers in regards to diagnosing, an One-Way Analysis of Variance
Preliminary Analyses. Participants with and without missing values did not differ in terms of SES. Participants without missing values were more likely to be girls (54.2%) compared to those with missing values (46.6%): χ2 (1) =8.0, p=.005. In the sample used for further statistical analyses (n=1,363), cannabis users (n=400) did not differ from nonusers with respect to gender (χ2=.34, p=.561), SES (t=.97, p=.332), and familial vulnerability for internalizing behaviour (t=_-84, p=.404). Cannabis users and nonusers did, however, differ significantly on familial vulnerability for externalizing behaviour (t=-2.1, p=.037), externalizing behaviour (t=-6.4, p<.001), alcohol use (χ2=83.6, p<.001) and tobacco use (χ2=367.1, p<.001). Cannabis users scored higher on both familial vulnerability for externalizing behaviour (M=.2, SD=.4) and internalizing behaviour (M=.3, SD=.2) than nonusers (M=.1, SD=.4 and M=.2, SD=.2, respectively). In addition, cannabis users were more often monthly alcohol users (92.8% vs. 69.7%) and also more often weekly tobacco users than nonusers (60.8% vs. 11.0%). Familial vulnerability for externalizing behaviour, externalizing behaviour, tobacco use, and alcohol use were also related to social skills and therefore introduced as covariates in further statistical analyses (Table 1). Table 1: Correlation among variables Control variables Teacher- Reported Social Skills Cannabis use variables
Preliminary Analyses. A path analysis was conducted using Mplus 7.4 (Xxxxxx and Xxxxxx, 1998-2015). MLM estimator was chosen for its robustness to non-normality in data that does not contain missing values (Muthén and Muthén, 1998-2015). Prior to hypothesis testing, an analysis of goodness of fit was conducted. Based on the criteria of goodness of fit by Xx and Xxxxxxx (1999), RMSEA values lower than .08 and CFI and TLI values above .90 are indicators of good fits to the data. The hypothesized model did not fit the data well (RMSEA = .163, 90% CI = [0.130, 0.198], CFI = .95, TLI =.94). Given that previous research suggests that motives of entrepreneurship are associated with aspirations to grow a business (e.g., Verheul & xxx Xxx, 2011) and that motives of entrepreneurship are related to optimistic beliefs about the future (xxx xxx Xxxx et al., 2016), the hypothesized model was revised by adding direct paths from motives of entrepreneurship to future time perspective and general growth intentions. The analysis of goodness of fit revealed that the revised model fitted the data well (RMSEA = .060, 90% CI = [0.000, 0.0114], CFI = .99, TLI =.97;
Preliminary Analyses. Chi squared analyses comparing CHR and HC groups revealed significant differences in sex ratio (larger proportion of males in CHR versus control group), but no group differences in race or ethnicity. T-tests revealed significant differences in age between groups (CHR mean = 18.5 years, healthy controls mean = 19.7 years). See Table 2 for a summary of sample characteristics including demographics and mean symptom ratings for the CHR and control groups. Remaining analyses comparing CHR and control groups included age and sex as covariates. In order to determine if medication status at baseline was related to the sleep variable (in the CHR participants), multiple independent samples T-tests were conducted. Results from the T-tests showed that there was no significant association between sleep disturbance and medication use of any kind, including sleep medications (see Table 3. for breakdown by medication status). Medication status was therefore not included as a covariate in remaining analyses. The lack of association between sleep disturbance and use of sleep medication likely indicates that the use of sleep medication is effective at reducing sleep disturbance.
Preliminary Analyses. Correlations between each predictor variable and infant weight at each time point are shown in Table 2. To examine whether the assumption of normality was met, infant weight at each time point was examined. The Xxxxxxx-Xxxx test of normality indicated that raw infant weight scores at all time points were normally distributed (ps > 0.05). Thus for all subsequent analyses in HLM, raw weight at each time point was used. First, an unconditional growth model was run using weight as the outcome and time as the predictor at level 1 in order to examine whether weight changed significantly over the course of the study. The estimated mean slope for the weight variable was 0.026 (SE=0.0007), indicating that infant weight increased at an average rate of 0.023 kilograms per day from birth to six-months-of-age. This was significant at p < 0.001, and indicated significant increases in weight over time within the sample. In addition, there was significant variation among slopes of weight gain in this sample (χ2=163.55; p < 0.001). Before examining the indirect effect of overfeeding on the relationship between maternal stress and RWG, associations between maternal age, gestational age, preterm status, infant sex, socioeconomic status, and overfeeding were examined using bivariate analyses. Results indicated that only preterm status was significantly associated with overfeeding (p = 0.046). None of the other potential control variables was related to overfeeding (ps > 0.05). When preparing to examine any main effects of maternal stress and overfeeding on weight gain, all potential covariates were each entered separately as predictors of infant weight (intercept and slope) in HLM. Results indicated that preterm status and gestational age were both significantly associated with birthweight (p < 0.001). Entering both variables as covariates would have been redundant, as preterm status is a dichotomized variable created from gestational age. Thus, only gestational age was entered as a covariate at the intercept in all following HLM analyses. No other variables were associated with either birthweight (i.e., intercept) or the slope of weight gain. Hypothesis Testing Prior to examining any associations in HLM, a binary logistic regression analysis was conducted to examine whether maternal stress was associated with the overfeeding composite variable. A significant association would support a portion of the mediation hypothesis. Results indicated that maternal stress significantly ...
Preliminary Analyses. First, we investigated if there were differences between the Turkish immigrant and the Dutch group in maternal age and education. Turkish immigrant mothers (M = 26.86, SD = 2.99) were significantly younger than Dutch mothers (M = 32.71, SD = 4.19), t(138) = 9.52, p < .01, and Turkish immigrant mothers had a lower educational level on a scale of 1 to 5 (M = 2.83, SD = 0.72) than Dutch mothers (M = 3.40, SD = 1.08), t(138) = 3.68, p < .01. Parenting in the Dutch and the Turkish immigrant groups First, we compared Turkish immigrant and Dutch mothers on parenting behaviors (maternal sensitivity and control) without controlling for the effect of maternal age and education (see Table 3.1). We found significant differences between the mothers in overall maternal sensitivity and all its subscales, and in their use of authoritative control. Turkish immigrant mothers were less sensitive during the tasks: they were less supportive, gave less clear instructions, and were more intrusive than Dutch mothers. With regard to control strategies, Turkish immigrant mothers were less authoritative in their strategies during the clean-up task than Dutch mothers. No differences were found in authoritarian control. After controlling for maternal age and education, these differences between the groups remained (see Table 3.1). Table 3.1 Differences between the Dutch and Turkish immigrant groups on parenting behaviors Dutch (n = 70) Turkish (n = 70) Group differences (F-values) Mean (SD) Range Mean (SD) Range Uncorrected Corrected¹ Sensitivity 6.38 (1.74) 1.6 - 9.8 2.65 (3.05) -2.8 - 9.5 77.36** 32.95** Supportive presence 4.67 (0.93) 2.5 - 6.3 3.79 (1.38) 1.5 - 6.7 19.54** 6.01* Intrusiveness 2.84 (0.77) 1.8 - 5.0 4.06 (1.21) 1.5 - 6.5 50.74** 18.35** Clarity of instruction 4.55 (0.56) 3.5 - 5.8 2.98 (0.94) 1.0 - 5.3 143.24** 70.04** Control Authoritative 12.80 (5.04) 3.0 - 25.5 9.88 (4.87) 1.7 - 22.2 12.16* 4.41* Authoritarian 4.67 (4.24) 0.0 - 15.9 5.51 (4.11) 0.0 - 16.6 1.43 0.18 Note. ¹ Controlled for maternal age and education; *p < .05; **p < .001 Correlates of parenting behaviors in the Dutch and Turkish immigrant groups To examine the associations between maternal age, education, maternal sensitivity, and authoritarian and authoritative control Xxxxxxx correlations were computed (see Table 3.2). Higher maternal age was related to more sensitivity and supportive presence, and less intrusiveness in the Turkish immigrant group. In the Dutch group, age was not related to any o...
Preliminary Analyses. Internalizing and externalizing scores were correlated at pretreatment (r=.61, p<.001) and at post treatment (r=.77, p<.001). Table 2 presents the inter-correlations between stress measures. As can be seen, lifetime stress did not correlate with either of the other stress measures. Negative recent life events and daily hassles were positively correlated. Table 2 also depicts the significant correlation between daily hassles and basal cortisol levels. Cortisol sampled at pretreatment was not significantly associated with recent negative life events or lifetime stress. Potential Confounds Preliminary analyses revealed that the time of day that cortisol was collected was significantly associated with cortisol levels at awakening (r=.20, p<.05) and this variable was controlled for in all further analyses examining HPA axis dysregulation. Number of hours slept the night before collection, as well as the use of cigarettes, chewing tobacco, prescription medication, antihistamines, marijuana, steroids, sleep medications, allergy medications, cold/flu medication, and caffeine were not significantly related to cortisol levels at awakening. Youth age was not significantly correlated with externalizing behaviors, cortisol, or any of the stress measures. This nonsignificant correlation runs counter to some findings in the literature (Xxxxx et al., 2008). However, our sample was restricted to adolescence, as we sampled only from youth aged 12 to 18 years. In this study, SES was positively correlated with externalizing behaviors at pretreatment (r=.24, p<.01). SES was also significantly correlated with ethnicity such that Latino youth had significantly lower SES (r=-.35, p<.001) and white youth had significantly higher SES (r=.35, p<.001). SES was controlled for in all analyses. An ANCOVA controlling for SES revealed ethnic group differences for externalizing behaviors, such that youth who were White had higher externalizing behaviors at pretreatment. No significant differences between ethnicities were found for externalizing behaviors at post treatment. All further analyses exploring externalizing behavior at pretreatment statistically controlled for ethnicity (using a dummy coded variable). Zero-order correlations between stress, basal cortisol, and pretreatment externalizing behavior are presented in Table 5. Another consideration is that self-report measures are susceptible to response sets and styles and can lead to spurious results. Researchers in this study attemp...
Preliminary Analyses. See Table 1 for descriptive statistics of all measures. See Table 2 for intercorrelations between externalizing scores at pre- and post-treatment, delinquency scores at pre- and post- treatment, and internalizing scores at pre- and post-treatment. As seen in this table, all of the youth behavioral problem measures are significantly and positively correlated with the exception of self-reported delinquency at pre-treatment, which is only correlated with CBCL externalizing behaviors at pretreatment, and self reported delinquency post-treatment. A paired sample t-test found significant reductions in externalizing scores between pre- and post-treatment, t(167)=9.35, p<.001, and in delinquency scores between pre- and post-treatment, t(165)=4.66, p<.001. See Table 3 for intercorrelations between all stress measures. Once again, it can be seen that almost all of these measures are significantly and positively correlated, as would be expected. Potential Confounds In preliminary analyses, SES was not found to be significantly correlated with externalizing scores at pre- or post-treatment, internalizing scores at pre- and post- treatment, or gender. Due to the lack of associations between the primary dependent variables and SES, SES was not controlled for in the analyses.
