The effect of socioeconomic status on the association between maltreatment experienced by parents during their childhood and current daily parenting hassles

13-9-2019

The effect of socioeconomic status on the

association between maltreatment

experienced by parents during their

childhood and current daily parenting

hassles

Name: Kim van Nispen

Student number: 2224267

University: University of Leiden

Faculty: Faculty of Social Sciences

Educational Course: Child and Education studies: Forensic

Family Sciences

Supervisor: Anneloes Slappendel

Second reader: Lenneke Alink

Index

Abstract ________________________________________________________________________ 3 Introduction _____________________________________________________________________ 4 Theoretical frame work ___________________________________________________________ 4

Childhood traumas ____________________________________________________________ 4 Daily parenting hassles (DPH) __________________________________________________ 5 Childhood trauma and parenting _________________________________________________ 5 Socioeconomic Status (SES) _____________________________________________________ 6

Aim and research question ________________________________________________________ 6

Method _________________________________________________________________________ 8 Design ________________________________________________________________________ 9

Participants ____________________________________________________________________ 9

Instruments ___________________________________________________________________ 10

Childhood Trauma Questionnaire Short Form (CTQ-SF) _____________________________ 10 Parenting Daily Hassles (PDH) _________________________________________________ 11 SES _______________________________________________________________________ 11

Analysis ______________________________________________________________________ 11

Results ________________________________________________________________________ 13 Univariate data exploration _______________________________________________________ 13

Missing data ________________________________________________________________ 13 Outliers ____________________________________________________________________ 13 Descriptive statistics __________________________________________________________ 13

Bivariate data exploration ________________________________________________________ 16

Hypothesis 1: Regression ________________________________________________________ 18

Hypothesis 2: One-way ANOVA __________________________________________________ 18

Hypothesis 3: Multiple Regression _________________________________________________ 19

Discussion ______________________________________________________________________ 21 Implications ___________________________________________________________________ 22

Limitations ___________________________________________________________________ 22

Abstract

The aim of this study was to assess the moderating effect of socioeconomic status on child maltreatment and parents’ perceptions on daily parenting hassles. Mothers with childhood maltreatment experiences encounter more parental stress due to the high impact of traumatic

experiences. Besides experiencing more stress, experiences of childhood maltreatment are associated to a lower overall socioeconomic status (Pereira, Li, and Power, 2017). One-hundred infants around the age of one-and-half years and one of their caregivers were included in this research. 88 mothers and 12 fathers participated together with 48 boys and 44 girls. All participating families were living in and around the Dutch city of Leiden at the time of recruitment. The research data was collected within the project ‘Changing Chaos’ at Leiden University, using the Childhood Trauma Questionnaires Short Form (CTQ-SF) and Parenting Daily Hassles (PDH). Additionally, to acquire information on

socioeconomic status, the participating parent reported their highest completed grade in school. Besides a multiple regression, a regular regression analysis and a Kruskal-Wallis test was conducted. No association between childhood maltreatment and daily parenting hassles was found, neither when socioeconomic status was added as a moderator. The moderation hypothesis could not be verified. Also, no difference between the socioeconomic groups for daily parenting hassles was found. Findings from this current study could provide knowledge about the influence of socioeconomic status, which can be used in preventions and interventions for children from higher risk groups.

Introduction Theoretical frame work

Childhood traumas. In 2010 118.836 children were found to be a victim of child maltreatment (Alink et al., 2010). The first three years of a child’s life appear to have the highest incidence of maltreatment, since 41% of these victims were zero to three years old (Lieberman, Chu, Van Horn, & Harris, 2011; Alink et al., 2010). A broad distinction can be made between four types of child maltreatment: physical abuse, sexual abuse, psychological or emotional abuse and neglect. Gilbert et al. (2009) suggested that witnessing intimate partner violence should also be defined as a form of maltreatment. Physical abuse is defined as the intentional use of brute force towards a child that results in physical injury, or at least has the potential to (Ay & Erbay, 2018; Gilbert et al., 2009). Sexual abuse is an attempt to or completed sexual act, sexual contact or sexual interaction towards a child. It is often underreported, due to the victim feeling ashamed, guilty, but most importantly dependent of the perpetrator (McElvaney, Greene, & Hogan, 2014). When the perpetrator is a relative or an acquaintance, the child might be afraid to upset the parents or to cause trouble for the family. Psychological or emotional abuse is defined as verbal behavior, such as threats and humiliation towards the child with the intentions of hurting ones emotional health and/or development. However, psychological or emotional abuse may also be unintentional (Ay & Erbay, 2018; Gilbert et al., 2009). Neglect is described as a situation in which a child does not get the physical or emotional care and support that they need, such as hygiene, nutrition and shelter (Ay & Erbay, 2018; Gilbert et al., 2009). Intimate partner violence includes threatening behavior, abuse or violence between two adults who are or have been intimate partners or family members (Gilbert et al., 2009). Alink et al. (2010) reported the highest prevalence for emotional neglect (7214 official reports), followed by witnessing intimate partner violence (5497 reports) and emotional abuse (2453 reports). Experiencing one or more of these maltreatment types can have long-term consequences in different dimensions of life, such as lower educational achievements and mental health problems (Gilbert et al., 2009). Besides lower educational achievements, maltreated children are more likely to receive special education and often work the semi-skilled or low-status jobs. Furthermore, after examining a maltreated group and control group (𝑀𝑀_{𝑎𝑎𝑔𝑔𝑔𝑔}= 28.7 years), only 41% of the maltreated children remained in employment the past five years, while this was 58% for the control group (Gilbert et al., 2009).

As noted before, experiencing one or more types of child maltreatment can have a

considerable influence on mental health status. Maltreated children are at a greater risk of developing behavioral problems, such as internalizing (depression and anxiety) and externalizing problems (aggression) (Gilbert et al., 2009). Besides the internalizing and externalizing consequences,

experiencing child maltreatment has a major effect on the hypothalamic-pituitary-adrenal axis function (HPA-axis), which causes heightened reactivity to stress (Ay & Erbay, 2018; DeSantis et al., 2011). Experiencing child maltreatment also increases the risk of posttraumatic stress disorder (PTSD), causing the child to have anxious thoughts, sleeping difficulties and reliving the trauma (Gilbert et al.,

2009). Suicide probability was found to be associated with childhood trauma, independent of whether the child had developed depression or anxiety (Ay & Erbay, 2018; DeSantis et al., 2011).

Furthermore, maltreated children are more likely to get involved in delinquent behaviors (Yun, Ball, & Lim, 2011; Steketee, Aussems, & Marshall, 2019). Steketee et al. (2019) suggested that all

maltreatment types are associated to violent delinquency. However, Yun et al. (2011) only found an association for sexual abuse and neglect. It should be noted that Yun et al. (2011) used a nationally representative sample, in contradiction to the high risk sample used by Steketee et al. (2019). Additionally, experiencing trauma, such as child maltreatment, can also affect parental experiences (Finegood, Raver, DeJoseph, & Blair, 2017).

Daily parenting hassles (DPH). Experiencing child maltreatment can cause attentional biases to threat. Because of that, their perceptions of stress in parenting can also increase (Finegood et al., 2017). A common form of a parental stress source is daily parenting hassles. Parenting hassles are defined as irritating, stressful and disturbing events in daily parenting (Crnic & Greenberg, 1990). An example of these hassles could entail children being resistant or having difficulties during meals (Finegood et al., 2017). Experiencing these hassles on a daily basis can have consequences for the parent but also for the child, since humans develop themselves in interaction with their environment and their parents play an extensive role in this (Belsky, as cited in Crnic & Greenberg, 1990; Gülseven et al., 2016; Ungar, Ghazinour, & Richter, 2013). According to Belsky (1984), sources of stress have a big influence on parenting which can influence the child’s development. Due to daily hassles parents may show less warmth and sensitivity, but also use more physical punishment towards the child as a result of being overwhelmed by the stress that the hassles cause (Pereira et al., 2012; Gülseven et al., 2016). Hence, children of parents who experience frequent hassles are also at a greater risk of

exposure to maltreatment (Gilbert et al., 2009; Fenerci & DePrince, 2018). Furthermore, children who grow up in families with frequent parenting hassles often experience less positive parent-child

interactions. Additionally, the hassles can lead to more externalizing behavioral problems in children (Crnic & Greenberg, 1990; Finegood et al., 2017). So, stress sources experienced by the parent on a daily basis can cause problematic behavior and development for the child.

Childhood trauma and parenting. Furthermore, mothers who have experienced childhood maltreatment encounter more parental stress due to the high impact of traumatic experiences (Pereira et al., 2012). Steketee et al. (2019) highlights the social learning theory developed by Bandura. This theory illustrates how behaviors in adult life are connected to behaviors acquired in childhood. Humans acquire these behaviors through observational learning and/or imitation, particularly from their parents (Bandura, as cited in McRea, Daire, Abel, & Lambie, 2017). For example, children’s experiences with aggression or abuse create a distorted view on the appropriateness of this kind of behavior. The consequences of this distorted view are visible later in life through experienced problems in parenting interactions, causing abusive, hostile and anxious interactions (Schwerdtfeger, Larzelere, Werner, Peters, & Oliver, 2013; Schwartz, Dohrenwend, & Levav, 1994). This is also

referred to as the cycle of violence hypothesis, or intergenerational transmission, since the violence or abuse spreads itself through generations (Steketee et al., 2019). Correspondingly, a meta-analysis regarding the intergenerational transmission of child maltreatment was conducted by Assink et al. (2018). The meta-analysis included 84 studies reporting 285 effect sizes for intergenerational transmission. They found an overall effect size 𝑑𝑑 = .289. This implies that parents who have experienced child maltreatment are approximately at an three times bigger risk of experiencing child maltreatment in their own family than parents who have not experienced child maltreatment (Assink et al., 2018). Another research focused on the effect of depressed mothers who have been abused in their childhood on parenting stress and hassles (Zalweski et al., 2013). Externalizing behaviors, the

irritability of an ill child and emotional withdrawal tested the maternal ability to stay calm during parenting tasks. Depressed mothers with a history of abuse show more emotional reactivity in these situations, thus appear to experience more stress.

Socioeconomic Status (SES). Previous research suggests an association between experienced childhood traumas and SES-related conditions such as occupational functioning and educational achievement (Gilbert et al., 2009; Pereira et al., 2017; Hussy, Chang, & Kotch, 2006). As previously described, only 41% of the maltreated children in the research from Gilbert et al. (2009) remained in employment within a period of five years and usually work the low- or semi-skilled jobs. Furthermore, maltreatment is related to lower SES overall, more sickness absence from work and financial

insecurity (Pereira et al., 2017). Moreover, maltreated children often receive special education and have lower educational achievements. Font and Maguire-Jack (2016) examined whether the association between adverse childhood experiences (ACEs), for example child maltreatment, and health outcomes was mediated by socioeconomic conditions (educational level, income level and marital status). The results showed that multiple ACEs are associated to a lower level of income and educational achievement. However they also found that ACEs may negatively influence the parents’ relationship, causing more domestic violence for example, which lessens children’s educational achievement in adulthood. Parents who finished less than a high school education, graduated high school or completed college, were more likely to report maltreatment, particularly physical neglect (Hussy et al., 2006). Thus, childhood traumas are associated to lower SES (Gilbert et al., 2009; Wade et al., 2016). Furthermore, a research conducted by Emmen et al. (2013) focused on the relation between SES, family stress and positive parenting in ethnic minority families. The results showed that lower SES was related to more family distress and hassles, which led to less positive parenting. Due to the association of SES between both childhood maltreatment and daily parenting hassles, SES will be tested as a moderator in the current research.

Aim and research question

The aim of the current research is to contribute to clarifying the relationship between child maltreatment and daily parenting hassles. The study of Santorelli, Woods, Carlin, Marsic, and Kaslow

(2012) examined the mediating effect of attachment on the association between childhood maltreatment and daily life hassles in a low-risk sample. Their results emphasized an indirect relationship between childhood maltreatment and daily life hassles. Experiencing child maltreatment can cause maladaptive attachment patterns which in turn can cause more stressful experiences in daily life. To elaborate the role of socioeconomic status this research will focus on the moderating effect of SES on the association between childhood maltreatment and daily parenting hassles. Accordingly, interventions can be developed for adults who have experienced trauma and specific SES groups in order to prevent daily parenting hassles and its consequences for the child and parent itself (Pereira et al., 2012; Gülseven et al., 2017). Besides, childhood traumas can be considered as major life events, and further research is needed to investigate how these events matter in adult/parental life (McDaniel & Slack, 2005).

The overall research question in this study is: “To what extent can experienced maltreatment in childhood predict daily hassles in parenthood for parents of one-and-half to two-year-olds in a Dutch sample?” Based on this overall research question, three sub-questions are created. First, is there an association between experienced child maltreatment and daily parenting hassles? Accordingly, it is hypothesized that childhood maltreatment would be associated with more daily parenting hassles in parenthood. Secondly, is there is a difference in daily parenting hassles according to SES? It is expected that daily parenting hassles occur more often when the SES is lower. Thirdly, this paper will examine whether the association between child maltreatment and daily parenting hassles is moderated by SES (Figure 1). It is expected that the association between experienced maltreatment in childhood and daily parenting hassles is stronger when there is a low SES level. In summary, the following hypotheses were examined.

Hypothesis 1: Childhood maltreatment is associated to more daily parenting hassles for parents of one-and-half to two-year-olds in a Dutch sample.

Hypothesis 2: Daily parenting hassles occur more often when the SES is lower for parents of one-and-half to two-year-olds in and a Dutch sample.

Hypothesis 3: The association between experienced childhood maltreatment and daily parenting hassles is stronger when there is a lower SES level for parents of one-and-half to two-year-olds in a Dutch sample (Figure 1).

Figure 1. Moderator hypothesis

Method Design

The current study selected data from a larger quantitative study at the University of Leiden called ‘Rondom het Jonge Kind’ (RJK). This data was collected through questionnaires and coding tasks of the mother and infant. These questionnaires and tasks were administered during home visits and were used to collect information of participants on different kinds of subjects. Respondents participated in two types of measurements, a pre- and posttest. Both the pre- and posttest consisted out of two home visits. The first one focused on the parent-child interactions. During the home visit the parent filled out various questionnaires, including a childhood trauma questionnaire and questionnaire assessing daily parenting hassles. The second home visit was without the infant, only with the parent. During this visit the parent had to take care of a realistic baby doll, complete some tasks at a laptop and tablet, and fill out various questionnaires. The posttest was almost identical to the pretest. It also consisted of one parent-child visit and one parent visit. During the posttest, the daily parenting hassles questionnaire was assessed once again. In between the pre- and posttest, a random controlled trial (RCT) was used to randomly assign the respondents to an intervention group or a control group. The intervention group received three phone calls and four home visits, while the control group received seven phone calls. The main goal of the intervention was to decrease chaos inside the home

environment.

The first instrument that was used during this study was the Childhood Trauma Questionnaire Short Form (CTQ-SF). This questionnaire was completed by the participating parent on a laptop, during the pretest. The second instrument, Parenting Daily Hassles (PDH) was assessed during both the pretest and posttest. This questionnaire was handed to the participating parent and asked if they could fill it out before the next home visit. The goal of the intervention was to decrease chaos inside the home environment, thus the PDH was assessed before and after the intervention to measure if the hassles had decreased.

Participants

Following approval by the Research Ethics of the Faculty of Child and Education Sciences, 100 parents and their infants were recruited through the municipalities in and around Leiden. 48 boys and 44 girls (𝑀𝑀_{𝑎𝑎𝑔𝑔𝑔𝑔}= 20.22, 𝑆𝑆𝑆𝑆 = 7.73) participated together with their mother (𝑁𝑁 = 88) or father (𝑁𝑁 = 12) (Table 1 and 2). The participating parents had a mean age of 36 years old (𝑆𝑆𝑆𝑆 = 4.30). Out of the 100 participants, 73 graduated from HBO or University (73%). The municipalities were contacted to provide contact information from families with children of approximately one-and-half years old. These families received a link for digital screening questionnaires. By the means of their scores on the CHAOS-questionnaires, families were screened on how much chaos occurred inside their household. Those families who proved to have sufficient chaos inside their household were

invited to participate in the intervention study. They each received an invitation to participate along with information about the research. Not only the families who replied to these letters later received a phone call, also the families who filled in their phone number on the questionnaire but had not yet replied to the letter were contacted. Specifically, most of the families decided to participate after they were contacted by the researchers. During the phone call the first appointment was scheduled and further questions about the study could be discussed.

Table 1

Frequencies categorical background variables

Categories N Missing Total

Genderchild Boy 48 8 100

Girl 44

Genderparent Mother 88 0 100

Father 12

Highest grade/yearparent VMBO/MAVO/VBO/LTS/LHNO 1 0 100

VWO 1

MBO/MTS/MEAO 25

HBO/HTS/HEAO 32

University 41

Table 2

Frequencies continues background variables

M SD N

Agechild * 20.22 7.73 86

Ageparent** 35.96 4.30 100

*Age_child in months **Age_parent in years

Instruments

Childhood Trauma Questionnaire Short Form (CTQ-SF). This research used the CTQ-SF, an 28-item self-report questionnaire to evaluate the experiences of childhood trauma (Thombs,

Bernstein, Lobbestael, & Arntz, 2009). The questionnaire consisted of five dimensions (a) physical abuse, (b) physical neglect, (c) sexual abuse, (d) emotional abuse, and (e) emotional neglect. Each dimension consisted of five items and a five-point Likers scale was used (1=never true; 5=very often

true). Therefore, each dimension had a score range of 5 to 25. This score range was divided into four categories (1) none to low trauma exposure, (2) low to moderate trauma exposure, (3) moderate to severe trauma exposure, and (4) severe to extreme trauma exposure. An example of one of the items is “During my childhood I knew that there was someone who would take care of me and protect me”. All subscales showed reasonable Cronbach’s alphas in previous studies, 𝛼𝛼 = .91 for physical abuse, 𝛼𝛼 = .63 for physical neglect, 𝛼𝛼 = .95 for sexual abuse, 𝛼𝛼 = .89 for emotional abuse and 𝛼𝛼 = .91 for emotional neglect (Thombs, Bernstein, Lobbestael, & Arntz, 2009; Liebschutz et al., 2018). An overall Cronbach’s alpha of 𝛼𝛼 = .89 was found for the CTQ in the current study.

Parenting Daily Hassles (PDH). PDH is a questionnaire operated to assess the amount of hassles experienced in daily parenting events and parent-child interactions (Crnic & Greenberg, 1990). The questionnaire consisted of a 5-point scale to rate the intensity of the hassle in certain events (1=no hassle; 5=big hassle). This intensity scale has a high Cronbach’s alpha in Crnic and Greenberg’s study (2009), 𝛼𝛼 = .90. The original questionnaire consisted of 20 items divided in frequency and intensity of the hassles, however the questionnaire that was used in the current study consisted of 46 and only measured intensity (Crnic & Greenberg, 1990). The PDH questionnaire in the current study also proved to have a Cronbach’s alpha 𝛼𝛼 = .90. An example of one item from the questionnaire in the current study is “My children do not listen or do as told without complaining”.

SES. There are different ways to measure SES. For example, educational level of the participating parent (Greitemeyer & Sagioglou, 2016). The research by Cohen et al. (2008) utilized educational level as a measure of SES. This level was estimated by asking the participants the following question “What is the highest grade or year of school you have completed?”. In the current study, educational level of the parent was an indicator for SES as well, in fact the respondents were asked the same question as the ones in the research of Cohen et al. (2008). The answer possibilities ranged from 1 = primary school to 8 = university graduate.

Analysis

In this study, childhood traumas experienced by the parent was the independent variable and the daily parenting hassles the dependent variable. SES was the moderator. First of all, univariate and bivariate analyses were conducted to check various assumptions. The outliers and missing values were examined and afterwards the assumption of normality was checked for all variables. To examine whether the experience of childhood traumas was a significant predictor of daily parenting hassles, a regression analysis was conducted. Before conducting this regression analysis the linearity between the predictor and response variable was checked. Next, an one-way ANOVA was conducted to analyze if there is a difference in daily parenting hassles according to SES. The homogeneity of variance was checked for the variables daily parenting hassles and SES. This is important because the samples are obtained from populations with equal variances. Thus, the variability of scores for DPH and SES should be similar. Finally, a multiple regression was operated to examine whether SES

moderated the relationship between childhood traumas and daily parenting hassles. An interaction term was created for childhood traumas and SES, and was added into the regression analysis. This interaction term was created by multiplying the centered CTQ variable and a dummy SES variable. To create this dummy variable the three existing SES values were changed into 0 = low and middle SES and 1 = high SES. The interaction term was created to investigate whether the relationship between childhood traumas and daily parenting hassles would differ when the SES was added to the model. This relationship was examined through the explained variance, regression coefficient and a scatterplot. Again, before conducting this analysis the linearity between the predictors and the

response variable was checked, but also the normality of residuals and homoscedasticity was checked. All analyses were conducted through the program Statistical Package for the Social Sciences 25 (SPSS) and operated a significance level of . 05.

Results Univariate data exploration

Missing data. Missing scores were excluded pairwise. Thus, if a respondent had a missing score on either the response or predictor variable, the respondent was excluded from the analyses. After inspecting the CTQ data, all missing items on this questionnaire were not at random. This means that a respondent had a missing score on every item of the CTQ. For the DPH however, missing scores were at random. Some respondents only had one or two missing scores which could indicate that they accidentally forgot to answer an item. In total, 23 missings were deleted from the sample, 𝑁𝑁 = 77.

Outliers. Three outliers were found for the dependent variable ‘DPH’ (Figure 1). Respondent 4609 seemed the most isolated from all other scores. Therefore, this respondent was excluded in the descriptive analysis to check if more normality would be assessed without this respondent. Based on the skewness (4.46) and kurtosis (3.64) it could be concluded that the distribution of the variable DPH was skewed to the right and peaked (Table 3). This suggested that most scores were clustered at the lower values. The peakedness of this distribution indicated that the scores were clustered in the center with thin tails. After excluding respondent 4609, the distribution was still skewed to the right (3.24), however not as extreme as before (4.53). The kurtosis decreased from (3.64) to (1.24) after excluding, implying a less peaked distribution. Even though the mean does not differ drastically with (2.08) or without (2.05) the outlier, due to the decrease of both the kurtosis and skewness outlier 4609 was excluded from the sample. As Figure 1 displayed, respondents 3621 and 6207 were also outliers. However, excluding those two respondents did not affect changes with regards to the kurtosis or skewness. Therefore only outlier 4609 was excluded. After exploring the distribution, without respondent 4609, the descriptive statistics indicated an improvement in the normality of the distribution (Figure 2).

The boxplot created for the independent variable ‘CTQ’ displayed five outliers (Figure 3). Respondent 6239 appeared to be the furthest away from the overall scores, slightly isolated from the other outliers. To check if there would be a decrease in skewness of the sampling distribution after excluding respondent 6239, a descriptive analysis was executed. The sampling distribution of the variable ‘CTQ’ proved to be skewed to the right (4.76) and slightly peaked (2.00) (Table 3). The descriptive statistics after excluding 6239 showed that the sampling distribution was still skewed to the right (4.42), however less skewed than before (4.76). The peakedness decreased to (1.52),

meaning the peakedness of the sampling distribution slightly assessed more normality. After excluding respondent 6239 a more equally divided distribution curve was found, as displayed in Figure 4.

Descriptive statistics. To summarize, the final sample of ‘DPH’ had a mean score of 2.05 with a standard deviation of .51 (Figure 2). The minimum score on the ‘DPH’ was 1.28 and the maximum score 3.57. The skewness was 3.25 and the kurtosis 1.24. The final overall mean score on the ‘CTQ’ was 1.60 with a standard deviation of .38 (Figure 4). The minimum score on the ‘CTQ’ was

1.140 and the highest score in this sample was 2.64. Its skewness was 4.42 and the kurtosis 1.52. Table 3 described the descriptive statistics before and after excluding the outliers.

The categorical variable ‘SES’ was created based on the completed highest grade or year of school of the participating parent. Table 1 showed the distribution of these grades or years. In order to create relatively equal groups, three categories were made: lower SES, middle SES and higher SES. Lower SES consisted of VMBO/MAVO/VBO/LTS/LHNO, VWO and MBO/MTS/MEAO. The parents who had completed HBO/HTS/HEAO were ranked into the middle SES group and the parents who had completed university into the higher SES group. The lower SES group consisted of 21 participants, the middle SES of 22 and the higher SES group of 32. The group sizes were relatively equal to one another, however the group size of the categories ‘lower SES’ and ‘higher SES’ showed to be more distributed.

Figure 1. Boxplot variable ‘DPH’

Table 3

Descriptive statistic of the continues variables

N Total Mean Score

Std. Min. Max. Skew. Std. Skew Kurtosis Std. Kurt DPH 77 2.08 .56 1.28 4.13 1.22 4.46 1.97 3.64 DPH* 75 2.05 .51 1.28 3.57 .89 3.20 .62 1.14 CTQ 77 1.62 .40 1.14 2.86 1.30 4.76 1.08 2.00 CTQ** 75 1.60 .38 1.14 2.64 1.22 4.42 .83 1.52 Valid N (pairwise) 75

*DPH descriptive statistics after excluding the outlier **CTQ descriptive statistics after excluding the outlier

Figure 2. Histogram ‘DPH’ after excluding outlier

Figure 3. Boxplot variable ‘CTQ’

Bivariate data exploration

Multiple assumptions were checked before conducting the multiple regression. First, the sample size should be larger than 𝑁𝑁 > 50 + 8𝑚𝑚, where 𝑚𝑚 was the number of independent variables (Pallant, 2016). In this research there were two independent variables, the CTQ and SES level. Thus, the research sample should be bigger than 𝑁𝑁 > 50 + 8 ∗ 2 = 𝑁𝑁 > 66. This assumption was satisfied since the sample size is 𝑁𝑁 = 75.

Next, the multicollinearity was checked conducting a correlation analysis (Table 4). This analysis referred to the relationship between the independent variables, CTQ and SES level. Multicollinearity occurs when the independent variables are highly correlated (𝑟𝑟 ≥ .90) (Pallant, 2016). A negative correlation was found between the two independent variables 𝑟𝑟(73) = −.008. Thus, no multicollinearity was found between CTQ and SES. Figure 5 showed the distribution of SES levels by daily parenting hassles (DPH). The means of the three categories were relatively close to each other, specifically the middle and higher-SES group. As displayed in Figure 5, the mean of the lower-SES group on DPH was higher. The lower-SES group showed a broader range than the middle-SES and higher-middle-SES. Figure 5 did not show any outliers. The normality of the sampling distribution was checked with a scatterplot (Figure 6). Ideally, the residuals in the scatterplot would be

rectangularly distributed, with most of the scores concentrated around the zero-point. The residuals in Figure 6 were distributed rectangular, no clear systematic pattern was detected. Besides, no outliers or extreme scores were detected (Figure 6). The Mahalanobis distance was also checked for outliers. Since two independent variables were included in the analyses the critical value was 13.82. The Mahalanobis distance in the SPSS output showed that the maximum value in this data file was 9.21, not exceeding the critical value and no outliers needed to be excluded. A small negative linear relationship was found between SES and DPH (Table 4).

Table 4

(Multicol)linearity between the predictor, response variable and moderator

Total DPH Total CTQ SES

Total DPH Pearson correlation 1 .099 -.233 Sig. .397 .044* Total CTQ Pearson correlation .099 1 -.008 Sig. .397 .949 SES Pearson correlation -.233 -.008 1 Sig. .044* .949 N = 75 *𝑝𝑝 < .05

Figure 6. Scatterplot residuals

Hypothesis 1: Regression

The first hypothesis checked if childhood maltreatment was associated with daily parenting hassles. A positive relationship was expected to be found. More experiences with childhood maltreatment was expected be associated to more daily parenting hassles. Based on the regression analysis, it could be concluded that CTQ score (predictor) did not significantly predict DPH score (outcome variable), 𝐹𝐹 = .73, 𝑝𝑝 = .397 (Table 5). CTQ only explained a small but not significant proportion of the variable DPH, 𝑅𝑅2= .01 (The regression coefficient for CTQ appeared to be . 13, meaning that if CTQ increased with 1 point, the predicted DPH-score would increase with . 13. This result was, however, not significant. Hence, the increase was only found for the sample and therefore cannot be generalized to the population.

Table 5

Regression statistics of the dependent variable DPH

Model Sum of Squares df Mean Square F Sig. 1 Regression .19 1 .19 .73 .397 Residual 19.07 73 .26 Total 19.26 74

Hypothesis 2: One-way ANOVA

An one-way ANOVA was conducted to test the second hypothesis, that daily parenting hassles occurred more frequently when the parents had a lower SES. First, the Levene’s test for

homogeneity of variances was checked. This test analyzed whether the variance of scores were the same for each of the three SES groups. A significance level greater than . 05 was obliged in order to not violate the assumption of homogeneity of variance. This assumption was however violated in the research, since 𝑝𝑝 = .043; 𝑝𝑝 < .05. Due to this violation and the small SES groups a non-parametric Kruskal-Wallis test was conducted. No significant main effect of SES on parenting hassles was found, 𝐻𝐻 = 2.32, 𝑝𝑝 = .31.

Hypothesis 3: Multiple Regression

To check whether the association between childhood maltreatment and daily parenting hassles was stronger when parents had a lower SES, a multiple regression was conducted. The total model which included the interaction term, explained a small portion of variance in DPH and was not significant, 𝑅𝑅2= .22, 𝑝𝑝 = .309. However, the strength of the relationship was weaker without the interaction term, 𝑅𝑅2= .19, 𝑝𝑝 = .262. No significant linear relationship was found between DPH and one of the independent variables, 𝐹𝐹(3,74) = 1.22, 𝑝𝑝 = .309. The unique contribution of the CTQ was not significant 𝛽𝛽 = .18, 𝑝𝑝 = .235. The regression coefficient (B) of CTQ was . 24, meaning that if CTQ increased with one point, the predicted DPH-score increased with . 24 (Table 6). Nevertheless, this result was not significant, thus was only found within this sample and cannot be generalized to the population. SES also proved to not have a significant unique contribution to DPH and a smaller beta: 𝛽𝛽 = −.17, 𝐵𝐵 = −.17 𝑝𝑝 = .156. Thus, if SES increased with one point, the predicted DPH-score decreased with −.17. However, since this result was not significant, it cannot be generalized to the population.

Figure 7 shows the plot with the centered CTQ variable on the x-axis and the total DPH on the y-axis both marked by the SES dummy variable. This scatterplot seemed to indicate a steeper

regression line for the lower and middle SES group, which meant that with an increased CTQ score, parents with a low or middle SES improved more on the DPH score. The regression coefficient (B) for the interaction term was −.30. Since high SES had the higher score on SES, the line for high SES was -.30 steeper than for lower and middle SES. Thus, low or medium SES and a higher CTQ score led to more hassles, while CTQ had no effect on daily parenting hassles for the high SES group.

Nevertheless, this result proved to be not significant, thus cannot be generalized to the population and was only found for the sample within this study.

Table 6 Coefficients moderation Model Unstandardized B Coefficients Std. Error Standardized Coefficients Beta t Sig. 1 (constant) 2.12 .08 27.34 .000 Centered CTQ .24 .20 .18 1.20 .235 SESdummy -.17 .12 -.17 -1.43 .156 CTQ * SES -.30 .31 -.15 -.97 .337

Discussion

The aim of this research was to investigate the association between experienced child maltreatment and daily parenting hassles (DPH), and the possible moderating effect of SES on this association. First of all, the association between childhood maltreatment (CTQ) and daily parenting hassles (DPH) was investigated. The results proved that the CTQ score did not predict the DPH score. This result

contradicts the indirect relationship that was found between childhood maltreatment and daily hassles in earlier research (Santorelli et al., 2012). Gülseven et al. (2017) did find an association between daily parenting hassles and child maltreatment. However, their research included daily parenting hassles as the independent variable and child maltreatment as the dependent variable.

Next, the second hypothesis was tested: whether lower SES was associated to more daily parenting hassles. Foremost, all three SES groups scored the same on the DPH questionnaire. Hence, the group differences in SES did not influence the frequency of experiencing daily parenting hassles.

Thirdly, the moderating effect of SES on the association between childhood maltreatment and daily parenting hassles was investigated. This moderating effect was however not found, as the interaction did not significantly predict the DPH outcome. Thus, socioeconomic status did not significantly influence the association between childhood maltreatment and current daily parenting hassles. This could be due to the discrepancy between the average SES level within this study and the average level within the Dutch population. The current study describes HBO as the average SES group, whereas in the actual population, vmbo/MBO was seen as the average SES level (CBS, 2019). Nonetheless, the model which included the interaction term explained a bigger percentage of DPH than the model that only included the main effect of CTQ and SES. Both explained variances were not significant, thus no conclusions can be made about the results. The scatterplot in Figure 7 seemed to indicate the effect of SES on the DPH scores. Respondents in the lower or middle SES group with more childhood maltreatment experiences had a higher DPH score, whereas respondents from the high SES group with more maltreatment experiences showed a more stable DPH score. However, this difference was not significant and thus no conclusions can be made about the scatterplot. Finegood et al. (2017) highlighted the role of socioeconomic factors in parents’ perceived daily hassles, instead of child behavior that could provoke this. Parents who experienced deep poverty were at an increased risk of experiencing negative life events and depression, but also reduced coping skills necessary to deal with their daily hassles. However, it is important to emphasize that not all low-income families experience more hassles and stress. Specifically, their experience of financial hardship is related to more daily parenting hassles, even after controlling for higher versus lower SES (Finegood et al., 2017). The discovery that the experience of financial hardship outweighs the actual financial status of families may explain the non-significant difference between the SES groups on the DPH

questionnaire. It should be noted that due to non-significant findings, these conclusions cannot be made nor generalized to the population and are solely based on the current sample.

Implications

The findings from this research could have an implication for prevention or intervention programs applicable to low educated families. The interaction between CTQ and SES seemed to explain a bigger percentage of DPH in this study. Although this was not a significant result, this could indicate the importance of a family’s socioeconomic status. Besides the importance of SES, previous research proved the negative consequences of frequent parenting hassles on parent-child interactions. Therefore, an intervention with a focus on educational opportunities for the child, in order to prevent them from having financial insecurities later in life, as well as on parent-child interactions would be preferred. An example of a well substantiated intervention for families with a lower SES is Instapje (Ince, 2017). This is a family-based stimulation program developed to increase educational

opportunities for children between the age of 1 and 4. The families who participate in this intervention receive 26 home visits, in which they will receive tools and develop skills to stimulate their children. In an intervention like Instapje the knowledge about the probable ability of socioeconomic status to predict parenting hassles can be of great use. Instapje involves a main focus on improving child factors, however adding a focus on parent-child interactions could create an intervention which improves both child factors and daily parenting hassles.

Limitations

Various limitations of this research can be described. First of all, after checking for missing data and excluding outliers the sample size was only 75. Due to a smaller sample the research has less power (Anderson, Kelley, & Maxwell, 2017). This makes it harder to find significant outcomes, even if these are present in the population. It could be expected that with more power SES would moderate the association between childhood maltreatment and daily parenting hassles. However, no causal conclusions can be made about the significant result, since no experimental method was used in this research.

Secondly, essentially high educated parents participate in the research. Therefore, the results of this research are less generalizable to other groups with, for example, lower educational levels. In this research a vast group of parents finished HBO or University (73%), while according to Centraal

Bureau Statistiek the biggest population group within the Netherlands are havo, vwo or mbo2-4

graduates (CBS, 2019). As a lower SES is associated to daily parenting hassles, they can be described as the high-risk group (Finegood et al., 2017). Thereby, interventions would preferably be created for these high-risk groups. However, due to the overrepresentation of highly educated parents in this study, the results that were found cannot be generalized to the high-risk groups. This hinders the development of interventions for those low-SES groups, while they are necessary. Due to the larger amount of high educated parents, the created SES groups are not as equally divided as preferred, as both the low and middle SES groups were smaller than the high SES. The effect of these unequally divided groups is however limited by conducting the Kruskal-Wallis test. Besides, only children from

Leiden and cities near Leiden are included in this research. Future research could include children from all over the Netherlands to increase the generalizability of the results.

Thirdly, the daily parenting hassles questionnaire could be influenced by negative experiences on the day of the home visit and therefore be biased. If the parent in question experienced a stressful situation with their partner or child previous to the questionnaire, the parent could answer the questions more negatively. As a result, they would receive a higher DPH score than their true score. On the other hand, if the parent has experienced one or more positive moments before filling in the questionnaire, their observed score might be lower than their true score.

Lastly, a gender discrepancy within the sample was found. 88 mothers participated whilst there were only 12 fathers. Traditionally, mothers are seen as the primary caregivers, while fathers are the main provider. However, during the second half of the 20th _{century an important transformation} took place. Mothers started participating in the labor market more frequently. Therefore fathers were expected to take-on a bigger role in parenting (Guerrina, 2010). Yet even after this shift, the image of a mother as the primary caregiver maintained. This existing distinction between maternal and paternal roles causes fathers to be discounted for within child welfare and scientific studies (Brown, Callahan, Strega, Walmsley, & Dominelli, 2009). However, gender differences between mothers and fathers on child development do exist. Previous research focused on the effect of maternal versus paternal abuse on the risk of depression in adolescence. Differences between mothers and fathers were found. While maternal abuse was indirectly associated to both daughters’ and sons’ levels of depression, paternal abuse only showed an indirect relationship with sons’ depressive symptoms (Moretti & Craig, 2013). Hence, an equally divided sample of mothers and fathers would have been preferred. Nevertheless, this research also consists of strong points.

All parents with children around the age of one-and-half (in and around Leiden) were invited to participate in this research. The first three years of a child’s life are to be crucial to their

development. Maguire-Jack, Berger, and Gromoske (2012) found physical punishment, such as spanking, at the age of one predicts internalizing and externalizing behavior problems at the age of three. Therefore, it is important to develop accurate interventions for families to prevent later

developmental issues for the child or parent. Especially parenting stress can have a major influence on the child’s temperament and behavioral development (Potapova, Gartstein, & Bridgett, 2014). For example, more stress can cause the infant to experience more fearfulness, increased crying and negative affectivity. Besides, questionnaires were used instead of observations that included coding. Specifically, the questionnaires applied in this research had a relatively high level of reliability ≥ .81, except for the physical neglect dimension in the CTQ-SF (≥ .63).

Conclusion

This research was based on the research question “To what extent can experienced

two-year-old’s in and around Leiden?”. No association between childhood maltreatment and daily parenting hassles was found, neither when socioeconomic status was added to the model. Additionally, no differences were found between the socioeconomic status groups on daily parenting hassles. The tested moderation effect proved to not be significant. Concluding, none of the hypothesized results were found. Nonetheless, a bigger research sample in future research may stimulate a more definite understanding of the relationship between childhood experiences, socioeconomic status and daily parenting hassles in adulthood.

References

Alink, L., Pannebakker, F., Euser, S., Bakermans-Kranenburg, M., Vogels, T., & Van IJzendoorn, R. (2010). Kindermishandeling in Nederland anno 2010: De twee nationale prevalentiestudie

mishandeling van kinderen en jeugdigen (NPM-2010). Leiden: Casimir Publishers

Anderson, S. F., Kelley, K., Maxwell, S. E. (2017). Sample-size planning for more accurate statistical power: A method adjusting sample effect sizes for publication bias and uncertainty. Psychological Science, 28, 1547-1562. doi:10.1177/0956797617723724 Assink, M., Spruit, A., Schuts, M., Lindauer, R., Van der Put, C. E., & Stams, G-J. J. M.

(2018). The intergenerational transmission of child maltreatment: A three-level meta-analysis. Child Abuse & Neglect, 84, 131-145. doi:10.1016/j.chiabu.2018.07.037

Ay, R., & Erbay, L. G. (2018). Relationship between childhood trauma and suicide probability in obsessive-compulsive disorder. Psychiatry Research, 261, 132-136.

doi:10.1016/j.psychres.2017.12.054

Brown, L., Callahant, M., Strega, S., Walmsley, C., & Dominelli, L. (2009). Manufacturing ghost fathers: the paradox of father presence and absence in child welfare. Child and Family Social

Work, 14, 25-34. doi:10.1111/j.1365-2206.2008.00578.x

DeSantis, S. M., Baker, N. L., Back, S. E., Spratt, E., Ciolino, J. D., Moran-Santa, M. M., Dipankar, B., & Brady, K. T. (2011). Gender differences in the effect of early life

trauma on hypothalamic-pituitary-adrenal axis functioning. Depression and Anxiety, 28, 383-392. doi:10.1002/da.20795

CBS (2019). Bevolking; hoogstbehaald onderwijsniveau en onderwijsrichting. Retrieved from https://opendata.cbs.nl/statline/#/CBS/nl/dataset/82816ned/table?dl=19D07

Cohen, S., Adler, N., Alper, C. M., Doyle, W. J., Treanor, J. J., & Turner, R. B. (2008). Objective and subjective socioeconomic status and susceptibility to the common cold. Health Psychology,

27, 268-274. doi:10.1037/0278-6133.27.2.268

Crnic, K. A., & Greenberg, M. T. (1990). Minor parenting stresses with young children. Child

Development, 61, 1628-1637. doi:10.1111/j.1467-8624.1990.tb02889.x

Emmen, R. A. G., Malda, M., Mesman, J., Van IJzendoorn, M. H., Prevoo, M. J. L., & Yeniad, N. (2013). Socioeconomic status and parenting in ethnic minority families: Testing a minority family stress model. Journal of Family Psychology, 27, 896-904. doi:10.1037/a0034693 Fenerci, R. L. B., DePrince, A, P. (2018). Intergenerational transmission of trauma: Maternal trauma

related cognitions and toddler symptoms. Child Maltreatment, 23, 126-136. doi:10.1177/1077559517737376

Finegood, E. D., Raver, C. C., DeJoseph, M. L., & Blair, C. (2017). Parenting in poverty: Attention bias and anxiety interact to predict parents’ perceptions of daily parenting hassles. Journal of Family Psychology, 31, 51-60. doi:10.1037/fam0000291

Font, S. A., & Maguire-Jack, K. (2016). Pathways from childhood abuse and other adversities to adult health risks: The role of adult socioeconomic conditions. Child Abuse & Neglect, 51, 390-399. 10.1016/j.chiabu.2015.05.013

Gilbert, R., Widom, C. S., Browne, K., Fergusson, D., Webb, E., & Janson, S. (2009). Burder and consequences of child maltreatment in high-income countries. The Lancet, 373, 68-81. doi:10.1016/S0140- 6736(08)61706-7

Goodman, E., Adler, N. E., Kawachi, I., Frazier, A. L., Huang, B., & Colditz, G. A. (2001). Adolescents’ perceptions of social status: Development and evaluation of a new indicator.

Pediatrics, 108, 1-8. doi:10.1542/peds.108.2.e31

Greitemeyer, T., & Sagioglou, C. (2016). Subjective socioeconomic status causes aggression: A test of the theory of social deprivation. Journal of Personality and Social Psychology, 111, 178-194. doi:10.1037/pspi0000058

Guerrina, R. (2010). The reproduction of mothering. In A. O’Reilly, J. G. Golson (red.), The

encyclopedia of motherhood. London, England: Sage.

Gülseven, Z., Carlo, G., Streit, C., Kumru, A., Selcuk, B., & Sayil, M. (2017). Longitudinal relations among parenting daily hassles, child rearing, and prosocial and aggressive behaviors in Turkish children. Social Development, 27, 45-57. doi:10.1111/sode.12247 Hussey, J. M., Chang, J. J., & Kotch, J. B. (2006). Child maltreatment in the United States:

Prevalence, risk factors, and adolescent health consequences. Pediatrics, 118, 933-943. doi:10.1542/peds.2005-2452

Ince, D. (2017). Databank effectieve jeugdinterventies: Beschrijving Instapje. Utrecht: Nederlands Jeugdinstituut. Gedownload van www.nji.nl/jeugdinterventies

Lieberman, A. F., Chu, A., Van Horn, P., Harris, W. W. (2011). Trauma in early childhood: Empirical evidence and clinical implications. Development and Psychopathology, 23, 397-410. doi:10.1017/S0954579411000137

Liebschutz, J. M., Chen, C. A., Richardson, M. A., Heeren, T. C., Cabral, H. J., Buchanan Howland, K., … Rose-Jacobs, R. (2018). Childhood Trauma Questionnaire (CTQ) correlations with prospective violence assessment in a longitudinal cohort. Psychological

Assessment, 30, 841-845. doi:10.1037/pas0000549

Maguire-Jack, K., Gromoske, A. N., & Berger, L. M. (2012). Spanking and child development during the first 5 years of life. Child Development, 83, 1960-1977.

doi:10.1111/j.1467-8624.2012.01820.x

McDaniel, M., & Slack, K. S. (2005). Major life events and the risk of a child maltreatment investigation. Children and Youth Services Review, 27, 171-195.

McElvaney, R., Greene, S., & Hogan, D. (2014). To tell or not to tell? Factors influencing young people’s informal disclosures of child sexual abuse. Journal of Interpersonal Violence, 29, 928-947. doi:10.1177/0886260513506281

McRea, L., Daire, A. P., Abel, E. M., & Lambie, G. W. (2017). A social learning perspective on childhood trauma and same-sex intimate partner violence. Journal of Counseling &

Development, 95, 332-338. doi:10.1002/jcad.12147

Moretti, M. M., & Craig, S. G. (2013). Maternal versus paternal physical and emotional abuse, affect regulation and risk for depression from adolescence to early adulthood. Child Abuse &

Neglect, 37, 4-13. doi:10.1016/j.chiabu.2012.09.015

Pallant, J. (2016). SPSS survival manual (6th _{ed.). New York, NY: Two Penn Plaza.}

Pereira, S., Li, L., & Power, C. (2017). Child maltreatment and adult living standards at 50 years old.

Pediatrics, 139, 1-10. doi:10.1542/peds.2016-1595

Pereira, J., Vickers, K., Atkinson, L., Gonzalez, A., Wekerle, C., & Levitan, R. (2012). Parenting stress mediates between maternal maltreatment history and maternal sensitivity in an

community sample. Child Abuse & Neglect, 36, 433-437. doi:10.1016/j.chiabu.2012.01.006 Potapova, N. V., Gartstein, M. A., & Bridgett, D. J. (2014). Parental influences on infant

temperament: Effects of father internalizing problems, parenting-related stress, and temperament. Infant Behavior Development, 37, 105-110.

doi:10.1016/j.infbeh.2013.12.014

Santorelli, N., Woods, A., Carlin, E., Marsic, A., & Kaslow, N. J. (2012). Attachment mediates the childhood maltreatment-daily hassles link in low-income, suicidal African American women.

Journal of Aggression, Maltreatment & Trauma, 21, 739-757.

doi:10.1080/10926771.2012.703290

Schwartz, S., Dohrenwend, B. P., & Levav, I. (1994). Nongenetic familial transmission of psychiatric disorders? Evidence from children of holocaust survivors. Journal of

Health and Social Behavior, 35, 385-402. doi:10.2307/2137216

Schwerdtfeger, K. L., Larzelere, R. E., Werner, D., Peters, C., & Oliver, M. (2013). Intergenerational transmission of trauma: The mediating role of parenting styles on

toddlers’ DSM-related symptoms. Journal of Aggression, Maltreatment & Trauma, 22, 211 -229. doi:10.1080/10926771.2013.743941

Steketee, M., Aussems, C., & Marshall, I. H. (2019). Exploring the impact of child maltreatment and interparental violence on violent delinquency in an international sample. Journal of

Interpersonal Violence, 0, 1-31. doi:10.117/0886260518823291

Thombs, B. D., Bernstein, D. P., Lobbestael, J., & Arntz, A. (2009). A validation study of the Dutch Childhood Trauma Questionnaire – Short Form: Factor structure, reliability, and known groups validity. Child Abuse and Neglect, 33, 518-523. doi:10.1016/j.chiabu.2009.03.001

Ungar, M., Ghazinour, M., & Richter, J. (2013). Annual research review: What is resilience within the social ecology of human development. Journal of Child Psychology and Psychiatry, 54, 348-366. doi:10.1111/jcpp.12025

Wade, R., Cronholm, P. F., Fein, J. A., Forke, C. M., Davis, M. B., Harkins-Schwarz, M., Pachter, L. M., & Bair-Merritt, M. H. (2016). Household and community-level adverse childhood experiences and adult health outcomes in a diverse urban population. Child Abuse

& Neglect, 52, 135-146. doi:10.1016/j.chiabu.2015.11.021

Yun, I., Ball, J. D., Lim, H. (2011). Disentangling the relationship between child maltreatment and violent delinquency: Using a nationally representative sample.

Journal of Interpersonal Violence, 26, 88-110. doi:10.1177/0886260510362886

Zalewski, M., Cyranowsky, J. M., Cheng, Y., Swartz, H. A. (2013). Role of maternal childhood trauma on parenting among depressed mothers of psychiatrically ill children. Depression and