Family Factors and Parenting in Ukraine

Abstract

The present study aimed to estimate the use of positive and negative parenting practices in Ukraine and explore relationships between parenting practices, intimate partner violence (IPV), alcohol use, and sociodemographics. Parents of children (N=320) ages 9–16 from three Ukrainian regions answered questions from the Alabama Parenting Questionnaire (APQ), the Revised Conflict Tactics Scales (CTS-R), Family Adaptability and Cohesion Evaluation Scale (FACES), and the Alcohol Use Section of the Drinking and Drug History and Current Use Patterns questionnaire. Ukrainian parents who reported lower use of alcohol, balanced family functioning and lower intimate partner violence were more likely to use positive parenting and less likely to use negative parenting practices. Parents with lower education were more likely to use negative parenting practices. Furthermore, alcohol use, IPV, parent education and higher family income were significantly and indirectly related with positive and negative parenting scores. The model explained 61% of variance in the positive parenting, 67% in the negative parenting.

Parenting, the act of support and development of children, has been well-studied in developed countries (Ainsworth, Blehar, Waters, & Wall, 1978; A. R. Anderson & Henry, 1994; S. A. Anderson, Sabatelli, & Kosutic, 2013; Belsky, 1984; Darling & Steinberg, 1993). However, very little is known about parenting in Ukraine. In a recent study, the risk for externalizing behaviors was higher for Ukrainian children if their parents did not use positive parenting, were not providing adequate monitoring and relied on corporal punishment (Burlaka, 2016). This study aims to extend understanding of parenting in Ukraine by testing the association of parent- and family-level factors with two clusters of parenting behaviors: (1) parenting practices associated with optimal development and socialization of the child (positive parenting); and (2) parenting practices associated with inconsistent discipline and increased risk for child abuse and neglect (negative parenting).

The ecological systems theoretical framework posits that children grow in environments that influence their growth and individual development (Bronfenbrenner, 1981). The family systems theory regards parenting as one of the family’s basic tasks (Anderson, Sabatelli, & Kosutic, 2013). Positive parenting can propel children to mastery of interpersonal relations and successful career. Belsky (1984) proposed that the child’s self-esteem, academic achievement and ability for self-control are positively correlated with parental warmth and consistent discipline. To reinforce child’s desired behaviors, rules, routines and predictable schedules should be key targets in effective parenting (Webster-Stratton, 2012). Webster-Stratton (2012) argued there is a great benefit in structured and positive parent-child interactions that may contribute to emotional and behavioral stability of children across all ages. Additionally, parental involvement with the child’s daily activities, such as attending parent-teacher conferences and attending extracurricular activities, is related with children’s higher academic achievement (Spera, 2005; Stright & Yeo, 2014). Even for high-risk families, effective parenting is a key component that can mediate and moderate the impact of adversity on child behavior outcomes and help raise resilient children (Burlaka, 2016, 2017; Masten, 2001).

In contrast, negative parenting can contribute to the early onset of aggressive and defiant behaviors that may continue into adulthood and contribute to other mental health problems, such as substance abuse (Dubow, Boxer, & Huesmann, 2008; Kawabata, Alink, Tseng, van IJzendoorn, & Crick, 2011). Children independent decision-making and responsibility is developed through consistent parenting—clearly communicated behavior expectations followed up with natural and logical consequences when children make mistakes or violate rules (Webster-Stratton, 2012). In addition to the lacking consistency, negative parenting is associated with child abuse and neglect. Neglectful parents fail to provide proper care, do not encourage school attendance, provide very little supervision, fail to establish norms or provide emotional and practical support to their children. Lack of child monitoring increases the risk of conduct problems, rule breaking and aggression (Burlaka, 2016; Pettit, Laird, Dodge, Bates, & Criss, 2001; Snyder, Schrepferman, Bullard, McEachern, & Patterson, 2012). Additionally, the use of physical punishment as a logical consequence of misbehavior can result in physical child abuse, the “nonaccidental injury inflicted on a child” (Crosson-Tower, 2013, p. 180). The corporal punishment is also linked with increased externalizing psychopathology in Ukraine and globally (Burlaka, 2016; Gershoff, 2002, 2010, 2013; Ma, Han, Grogan-Kaylor, Delva, & Castillo, 2012). The UN Committee on the Rights of the Child urged member States to address “any level of legalized violence against children” (UN Committee on the Rights of the Child, 2007, p. 6). Although Ukraine has ratified the Convention on the Rights of the Child (United Nations, 2015), its implementation has not been very successful. In a recent study, 75% of Ukrainian parents reported spanking, slapping and hitting children with objects (Burlaka, 2016).

Parent and Family Characteristics Related to Parenting Behavior

Several sociodemographic characteristics have been found to influence the quality of parenting. For example, lower income, unstable work and high debt may contribute to caregiver depressive symptomatology and impact the ability to provide nurturing, warm and involved parenting (Conger et al., 1992; Dovgopol, 2009). Chumak and Tkachenko (2008) also proposed that constructive and respectful relations with the child largely depend on family socioeconomic status, the psychological climate in the home, and education of parents. Furthermore, older parents may be more mature and psychologically stable, and able to provide quality care for children (Belsky, 1984). In fact, older mothers may feel more pleasure from spending time taking care of their children (Ragozin, Basham, Crnic, Greenberg, & Robinson, 1982). Interestingly, Fergusson and Woodward (1999) found that children who were born to mothers under 30 years of age had 1.5 to 8.9 times higher risk to develop conduct and mental health problems.

Additionally, alcohol use can significantly impact the ability to parent children, particularly in the context of Ukraine, one of the top drinking nations in the world (World Health Organization, 2014). Misuse of alcohol is an important factor that can set the stressful dynamics in the entire family (Loukas, Fitzgerald, Zucker, & von Eye, 2001) and limit parents’ personal psychological resources and the ability to parent (Dovgopol, 2009; Iakunchykova & Andreeva, 2012; Levendosky & Graham-Bermann, 2001). In a world mental health (WMH) survey in Ukraine (Webb et al., 2005), 8.5% of Ukrainian women reported heavy alcohol use in the past 12-month.

The increased frequency of alcohol use among Ukrainian mothers is associated with lower spirituality and the use of escape-avoidant coping strategies, confrontation and less frequent use of positive reappraisal of stressful life situations (Churakova, Burlaka, & Parker, 2017), which may have negative “trickle down” effects on children, ultimately leading to less successful child adjustment. Children of parents with alcoholism had increased internalizing behavior symptomatology that was mediated by negative parenting (Burstein, Stanger, Kamon, & Dumenci, 2006) as well as externalizing psychopathology that was mediated by parent-child conflict as well as by family conflict (Loukas et al., 2001).

According to conflict theory, conflicts are inevitable and family members use various tactics to resolve them, ranging from peaceful negotiations to physical violence (Straus, Hamby, Boney-McCoy, & Sugarman, 1996). In a recent study, 58% of Ukrainian women reported being physically assaulted during last year, 34% were victims of sexual coercion, and 44% reported being injured as a result of intimate partner violence (IPV; Burlaka, Grogan-Kaylor, Savchuk, & Graham-Bermann, 2017). In that study, mothers who had lower education were more likely to experience IPV. The IPV is a major issue that can affect parenting ability (Levendosky & Graham-Bermann, 2001).

Balanced family cohesion and flexibility represent other basic family systems processes frequently associated with better psychosocial functioning of the parents and higher quality of parent-child interactions (Anderson et al., 2013; Huth-Bocks & Hughes, 2008). A balanced cohesion exists in families where members are neither overly enmeshed nor disengaged in relationships with each other while families with a balanced flexibility have a healthy “amount of change in its leadership, role relationships, and relationship rules” (Olson, Russell, & Sprenkle, 1983, p. 519). Poor emotional bonding and problem power and role structures in the family were related to children’s criminal, addictive behaviors and increased mortality in the U.S., Norway, Sweden, Italy and in former Soviet Union republics (Ogden & Halliday-Boykins, 2004; Olsson, 2010; Redmond & Spooner, 2009; Schaeffer & Borduin, 2005; Tafà & Baiocco, 2009).

The present study

Given recent data suggesting a link between child externalizing behaviors and lower use of positive parenting, poor monitoring, and corporal punishment, it is critical to investigate specific family processes associated with parenting in the Ukrainian families. The current study aimed to evaluate the direct and indirect effects of parent alcohol use, IPV, family functioning, and sociodemographic characteristics on parenting. We hypothesized that the use of positive parenting practices will be significantly associated with older parent age, higher parent education and income, lower parent involvement with alcohol, less IPV, and balanced family cohesion and flexibility. It was also hypothesized that use of negative parenting practices will be significantly associated with younger parent age, lower parent education and income, higher parent involvement with alcohol, more frequent IPV, and unbalanced family cohesion and flexibility.

Method

Participants

This study used a cross-sectional, community-based sample of parents-child dyads. Both children and their parents were interviewed, however, only parent data is used in the present study. The study was conducted in three Ukrainian regions in order to understand contextual and developmental factors associated with the onset of mental health and alcohol use disorders. Participants were invited to the study through fliers and personal invitations if they had a school-age child. School psychologists conducted face-to-face interviews on safe school premises and in participants’ homes. Families were recruited to the study to form an age- and sex- clustered sample of children. Recruitment was terminated when the sample reached a total of 320 parent-child dyads with 40 children (50% boys) in each age year, from 9 to 16. All measures were administered in the Ukrainian language, the official language in the country. The measures were translated from English to Ukrainian by a professional translator and then back-translated into English. The process was supervised by the first author who was trained as Ukrainian social worker and holds a U.S. doctorate in social work and clinical psychology. After translation, a feedback from a small group of Ukrainian children and a group of Ukrainian social workers, psychologists and teachers was used to confirm cultural relevancy and linguistic aspects of the instruments. Data were collected in accordance with Ukrainian law and standards for responsible conduct of research by the Ukrainian Methodological Psycho-medico-pedagogical Center of the Department of Education. The Center obtained informed consents and assents, interviewed participants and provided data without personal identifiers to authors for analyses. The notice of determination of non-regulated status was subsequently issued by the University of Michigan Institutional Review Board.

The sample included two adoptive parents, six grandmothers, 18 fathers and 294 mothers. The mean age of the parents was 37.79 years (SD = 6.52), 92% of respondents self-identify as Ukrainian, and others are Russian, Roman, Polish and Armenian. Nineteen percent of participants completed 11 grades of school or less, 49% had vocational training, 6% had some college education and 26% had a university degree. Sixty percent were married or lived with a partner and others were single, widowed, divorced or married and living separately.

Measures

Demographic characteristics

As part of this study, parents answered questions about their age (measured in years), level of education (measured in total number of years spent at educational establishments), ethnicity and family income (measured in US$).

Parenting practices

The Alabama Parenting Questionnaire (APQ; Frick, Christian, & Wootton, 1999; Frick, 1991) has been widely used for the purpose of measuring parenting skills in parents of children 6–18 years of age (Dadds, Maujean, & Fraser, 2003; Essau, Sasagawa, & Frick, 2006; Hawes & Dadds, 2006). The APQ consists of 42 items assessing positive and negative parenting practices grouped within six areas: (a) involvement, (b) positive parenting, (c) poor monitoring/supervision, (d) inconsistent discipline, (e) use of corporal punishment, and (f) use of discipline practices other than corporal punishment. Parents rated the frequency with which the parenting practices typically occurred in their home on a 5-point scale ranging from 0 (never) to 4 (always). Two subscales of the measure, Involvement and Positive Parenting form the APQ Positive Parenting Scale (Cronbach’s alpha = .95) while Inconsistent Discipline, Poor Monitoring/Supervision and Use of Corporal Punishment form the APQ Negative Parenting Scale (Cronbach’s alpha = .85).

Intimate partner violence

The revised version of the Conflict Tactics Scales (CTS2; Straus, Hamby, Boney-McCoy, & Sugarman, 1996; Straus, 1979) was used in this study to understand the relationship between parenting and IPV. The CTS2 consists of 39 items measuring the extent of psychological aggression, physical assault, sexual coercion, and injury as well as use of negotiation to address conflicts in dating, cohabitating, or marital relationships (Straus, Hamby, Boney-McCoy, & Sugarman, 1996; Straus, 1979). Answering the CTS2 questions, parents reported on Psychological aggression, Physical Assaults, and Sexual Coercion tactics used by their partner during last year on a 7-point Likert scale ranging from 0 (never) to 6 (more than 20 times). In the present study, alpha coefficients for the Psychological aggression was .87, for Physical Assault, α = .91, for Sexual Coercion, α = .82. The CTS2 Total Violence alpha was .95.

Family cohesion and flexibility

Two key concepts relating to family functioning—family cohesion and family flexibility—were measured using the Family Adaptation and Cohesion Scales-IV (FACES; Olson & Gorall, 2003). The family cohesion subscale measures the agreement of family members with such statements as “Our family seldom does things together,” “Family members are involved in each others lives”, while family flexibility is measured by such questions as “We have clear rules and roles in our family” and “When problems arise, we compromise” on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Circumplex ratio scores were created for the cohesion and flexibility subscales to capture both balanced and unbalanced family functioning. Alpha reliability coefficient reported in the FACES-IV validation study was .89 for Balanced Cohesion scale and .84 for Balanced Flexibility scale (Olson, 2011). In this study, the alpha for internal consistency of items in the Balanced Cohesion scale is .89 and .81 in the Balanced Flexibility scale. The Total Circumplex Ratio was used in the model, which is the average of the Cohesion Ratio and Flexibility Ratio (Olson, 2011). Scores below one indicate beliefs that family functioning is unbalanced. Conversely, the scores above one suggest that system is perceived as balanced. The reliability alpha of the Total Circumplex Ratio was .92.

Frequency of alcohol use

The Alcohol Use Section of the Drinking and Drug History and Current Use Patterns questionnaire (Zucker, Fitzgerald, & Noll, 1990) was used in this study to assess the frequency of alcohol consumption among parents in this sample. Parents answered questions about the frequency with which they consumed beer, wine (or a punch containing wine) and liquor on 12-point scale (0 = never, 1 = less than once a year, 11 = 3 or more times a day). The drinking frequency score was then converted to approximate the number of drinking occasions during the year. That is, the answer “once a month” was recoded as 12, to indicate that a person drank one time a month×12 months. By the same token, the answer “two times a day” was recoded as 730. Thus, the drinking frequency count for every alcoholic beverage ranged from 0 to 1460. Subsequently, a variable of Global Annual Alcohol Use Frequency was generated, which is a sum of reported frequencies of use of individual alcoholic drinks. The Cronbach’s alpha for the Global Annual Alcohol Use was .85.

Overview of the Analyses

Structural equation modeling (SEM) was the principal analytic technique used in this study (Acock, 2013). SEM analyses combine measurement models with path analyses and allows use of multiple indicators for latent variables. This approach permits isolation of random measurement error from latent variables, which increases explanatory power. The maximum likelihood estimation was used to fit the model where an observed information matrix was utilized to compute the variance-covariance matrix of the estimators and standard errors. The model was fit with Stata 13 software package (StataCorp, 2015). The guidelines suggested by Raykov, Tomer, and Nesselroade (1991) were followed in reporting SEM results in this study. In particular, the goodness of fit was assessed using the Comparative fit index (CFI), Tucker-Lewis index (TLI), and the root mean squared error of approximation (RMSEA). The recommended cutoff value for CFI and TLI is .95 while for RMSEA, the cutoff of .05 represents a good fit and less than .08 – a reasonably close fit (Acock, 2013). A series of analyses were performed to test indirect relationships in the model. These tests were performed using the delta method-based nlcom command in Stata (Acock, 2013; Phillips & Park, 1988; StataCorp, 2015). There were about 6% of missing data and multiple imputation technique was used to impute the missing values (Molenberghs & Kenward, 2007).

Results

Results suggested that parents in Ukraine frequently praised and rewarded positive behaviors of their children. The mean score was 2.90 (SD = .86) on positive parenting and 2.09 (SD = .98) on the Involvement scale. As for negative parenting scales, parents scored 1.65 (SD = .63) on Inconsistent Discipline, 1.56 (SD = .69) on Poor Monitoring Scale and 1.51 (SD = .91) on Corporal Punishment. Comparison of the Global Positive Parenting scale (M = 2.49, SD = .84) with the Global Negative Parenting scale (M = 1.57, SD = .58) revealed that overall, Ukrainian parents reported significantly higher scores on positive parenting, t(319) = 13.10, p < .001. Majority of parents (92.5%) used alcohol and 84% experienced psychological or physical violence, or sexual assault from their intimate partners during past year. Additional demographic information is presented in Table 1.

Table 1.

Correlations, Means, Standard Deviations, and Chronbach Alpha Coefficients of the Major Study Variables Displayed in Figure 1 (N = 320)

Variable	Positive parenting	Negative parenting	Alcohol use	IPV	Family functioning	Income	Age	Mean	Standard Deviation
Positive parenting	(.95)							2.49	.05
Negative parenting	−.54	(.84)						1.57	.03
Alcohol use	−.50	.35	(.85)					172.73	261.85
IPV	−.47	.52	.39	(.95)				65.86	88.43
Family functioning	.58	−.53	−.35	−.49	(.92)			.35	.11
Income	.22	−.24	−.21	−.27	.30	-		405.73	254.81
Age	.01	−.05	−.10	−.03	.05	−.11	-	37.82	6.50
Education	.29	−.38	−.29	−.40	.32	.43	.05	13.26	2.10

Measurement Model

The first latent variable, frequency of Parental Alcohol Use, was indicated by frequency with which beer, liquor and wine were used during last 12 months (Figure 1).

Figure 1.

Structural equation model representing effects of parental alcohol use, family functioning, IPV and sociodemographic characteristics on positive and negative parenting. Large ellipses represent latent constructs. Bolded lines represent statistically significant standardized path coefficients at .05 or above. Curved lines represent correlations among errors. Small circles represent residual variances.

Parent experiences of IPV were indicated by three subscales that were created from the measure of IPV described earlier. Flexibility and cohesion scales served as the two indicators for the Family Functioning latent variable. Positive parenting latent variable was indicated by APQ Positive Parenting and parent involvement scales. Finally, corporal punishment, inconsistent parenting and poor monitoring scales served as indicators of the APQ Negative Parenting latent variable. Table 1 includes Cronbach’s alpha reliabilities, means, standard deviations, and intercorrelations among latent variables. Figure 1 shows loadings of observed variables on latent indicators. The measurement model also included a covariance between all latent variables in the model. Because measurement model demonstrated a good fit to the data: χ² (55, N = 320) = 94.16, p < .001, CFI = .982, TLI = .974, RMSEA = .047, it was subsequently used to test a following structural model.

Structural Model: Linking Sociodemographic Variables, Alcohol Use, IPV, Family Functioning with Positive and Negative Parenting

In the first step, the relationship between predictor and outcome variables was analyzed. Table 1 illustrates significant correlations of parent alcohol use, IPV experiences, family functioning, parent education and income with positive and negative parenting practices. Parent age was not significantly related to outcome variables and, therefore, was not used for further analyses (Baron & Kenny, 1986). Parent education and income that represent observed sociodemographic covariates were included in the subsequent path analyses. Figure 1 shows results of these analyses. The straight lines represent paths with the values along the lines showing the estimated strength of influences. These values are standardized path coefficients (betas). This model provided a good fit for the data, χ² (76, N = 320) = 116.15, p < .01, CFI = .98, TLI = .98, RMSEA = .04.

Results suggest that lower parental involvement with alcohol, balanced family functioning, and fewer experiences of partner violence had a direct significant effects on more frequent use of positive parenting practices and less frequent use of negative parenting. Balanced family functioning had a strongest effect on the positive parenting, β = .43, z = 6.49, p < .001 and on the negative parenting, β = −.41, z = −5.55, p < .001. Figure 1 also shows that parents with lower education were more likely to use negative parenting practices, β = −.16, z = −2.72, p < .01. The direct relationships between parent education and use of positive practices was not significant. Likewise, family income had no direct effect either on positive or negative practices.

Results also indicated that lower alcohol use, fewer incidences of IPV, better parent education and higher family income were significantly and indirectly related with higher positive parenting score and with lower negative parenting score. Parent education had a strongest indirect effect on positive parenting, β = .21, z = 4.08, p < .001 and alcohol use had the strongest indirect effect on negative parenting, β = .26, z = 5.67, p < .001. Alcohol use also had the strongest total standardized indirect effect on positive parenting, β = −.59, z = −6.84, p < .001. Parents experiences of domestic violence had the strongest total standardized indirect effect on negative parenting, β = .50, z = 6.48, p < .001. The model explained 61% of variance in the positive parenting, 67% in the negative parenting.

Mediation analyses

Results of mediation analyses suggested multiple significant indirect chains of influence on parenting. Income had a significant standardized indirect effect mediated by family functioning on positive β = .06, p < .05 and negative β = −.06, p < .05 parenting. Next, there was a trend-level indirect effect of parent education mediated by IPV on positive parenting and a statistically significant indirect effect on negative parenting, β = −.10, p < .001. There was a significant indirect effect of alcohol use, medicated by IPV, on positive β = −.04, p < .05 and negative β = .11, p < .001 parenting. Additionally, there was a significant indirect effect of alcohol use, medicated by family functioning, on positive β = −.10, p < .001 and negative β = .10, p < .001 parenting. Also, there was a significant indirect effect of IPV, medicated by family functioning, on positive β = −.17, p < .001 and negative β = .16, p < .001 parenting.

There was a trend-level effect of parent education mediated by alcohol use on positive parenting and non-significant effect on negative parenting. Likewise, there was no significant indirect effect of parent education on parenting mediated by family functioning. Next, there was no significant indirect effect of income, mediated by IPV, on positive parenting, however, there was a trend-level of significance in effect on negative parenting. Finally, the indirect relationship between family income, mediated by parent drinking, on parenting was not supported by the data.

Discussion

Majority of Ukrainian parents reported use of positive parenting practices with their children in this study. Parents with higher education tended to drink less alcohol and have more balanced family life which was associated with greater use of positive parenting. This finding is consistent with the current trend among progressive Ukrainians to use Internet resources to look for the best strategies to raise competent and well-behaved children who would succeed in the highly competitive global economy. Modern psychotherapy and parenting programs are also now available practically in every Ukrainian town or city. Perhaps, wealthier parents in this study had more resources to get access to these interventions. In this study, parents with higher income reported fewer IPV experiences and higher quality of family life and both lower IPV victimization and balanced family functioning were linked with more frequent use of positive parenting.

Results of the measurement analyses suggest that the positive parenting construct was equally loaded by positive disciplining and engaged parenting. However, negative parenting construct was primarily driven by potentially abusive/neglectful parenting behaviors, spanking and poor child monitoring with inconsistent parenting having a weaker loading on the construct. Without a proper training in parenting, parents might not know what constitutes the successful childrearing and their choice of interaction with children may depend on other factors, such as personality and life circumstances. This can shed the light on other results of the analyses tested in the model (see Figure 1).

The results provide partial support for hypothesized relationships between parent sociodemographic variables and parenting practices. Prior research linked parent education with the ability to provide higher quality of childcare and better parenting (Belsky, 1984; Chumak & Tkachenko, 2008). Our results suggest that higher parental education had a direct relationship with lower use of negative parenting. In this study, parents with better education had higher income and, therefore, could experience lower financial strain compared to less educated and impoverished parents. Being under pressure, some Ukrainian parents might fail to refrain from becoming angry and hitting the child. Lower education can also force parents to accept lower paying and irregular jobs and they might need to spend more hours away from their children and have less opportunities to monitor their children’s behavior and whereabouts. In this model, parent education also had a significant indirect effect on negative parenting mediated by IPV suggesting that parents with lower education are at increased risk for domestic violence and such abusive relationships may contribute to increase of negative parenting. This result is consistent with prior research in Ukraine (Burlaka et al., 2017) and in the United States (Levendosky & Graham-Bermann, 2001) and was expected.

Contrary to hypothesis, higher parent education was not associated with increased positive parenting. Regardless of academic background, parents equally well rewarded good behavior of their children and stayed engaged in school and out-of-class activities. It is possible that many parents find it natural to praise their children. High parent involvement, however, can be explained by traditionally strong relationship between Ukrainian schools and families. Ukrainian teachers, for example, make regular house calls, home visits, assessment of family functioning, they reach out to parents of children who show signs of academic failure or discipline problems, refer families to various governmental bodies and interact with low enforcement agencies in crises situations. This system may partly explain the high scores on parent involvement.

In this study, parent income and age were not significantly related with either positive or negative parenting. In Ukraine, majority of families live in poverty, therefore, it is possible that with the sample size used in this study, statistical analyses could not detect effects of income on parenting behavior. However, income had a significant indirect association with negative parenting mediated by family functioning. These results suggest that parents with lower income experience difficulties with emotional closeness, rules and role structure in their homes and these difficulties appear to contribute to poor child monitoring, more frequent use of spanking and inconsistent parenting. Furthermore, the family income was measured by a question that asked about total monetary amount received by the family in the last month. This method may not be sensitive to seasonal variations of income typical for rural Ukrainian families involved in agricultural and animal production. Also, this question does not take into account income from informal sources, such as exchanges of commodities, and informal payments that some participants may not be willing to admit. In the future studies, it might be better to assess income using indirect information about family wealth. For example, instead of asking parents how much money they earn, parents could be asked questions about their families’ buying and recreation capability.

The finding that the age had a weak relationship with other study variables can be partially explained by the fact that this study was limited to parents of children aged 9–16 years and as in the case with income, there was not enough statistical variation in the age variable, which can explain these insignificant results. Indeed, majority of parents in this study were 32–40 years of age, which means that they already moved beyond the challenges of young adulthood age and have not encountered health, social and economic challenges related to transition to old age. For more representative results, future studies might need to estimate parenting of children across all age groups.

The results of this study suggest fully supported the hypothesis that alcohol has a strong negative impact on parenting. In addition to direct relationship with decreased use of positive parenting and increased use of negative parenting, higher frequency of alcohol use influenced parenting through indirect paths. In particular, increased alcohol use was associated with more conflicts in the homes and more unbalanced family functioning, which in turn resulted in lower use of positive parenting and higher use of negative parenting practices. Indeed, parental alcoholism can disturb the healthy balance in the family, weaken emotional connection between family members or lead to codependent and overly reactive behaviors. Also, alcohol can lead to chaotic organization of the home environment and disturbance of the roles. For example, it is not uncommon for alcoholic families that children assume adult roles, become breadwinners, and help raise younger siblings. Furthermore, parent alcohol abuse can undermine parents’ ability to set and maintain routines and schedules. Intoxicated parents may have impaired ability to interact with their children, understand and adequately respond to children’ emotions and communication cues, engage in children’s life, help them with school and extracurricular activities, and provide adequate supervision and monitoring. These effects can be a result of parents’ impaired memory and attention, emotional dysregulation and the fact that parents’ time and resources have been directed away from their children and toward activities related to securing, consuming and recovering from alcohol use. These results were expected based on previous research in developed countries (Anderson et al., 2013).

Furthermore, IPV had significant direct and indirect association with parenting in this study. Parents who experienced more violence reported lower scores on positive parenting and higher scores on negative parenting. Also, higher intensity of IPV was related to less balanced family cohesion and flexibility. Unbalanced family functioning mediated the impact of IPV on parenting and also was directly related with lower positive parenting and higher negative parenting.

Results of this study indicate a strong relationship between alcohol use and IPV. The chosen direction from alcohol use to the IPV was an attempt to satisfy a requirement of temporal precedence for SEM models (Baron & Kenny, 1986). Most Ukrainians start using alcohol before they develop relations with intimate partners. Also, the measure of domestic violence used in this study only asks about IPV experiences during last years. It is possible that such experiences were the first experiences of abuse. Alcohol use, however, is a developmental problem with chronic and progressive features (Hicks & Zucker, 2014) with up to 52% risk for the disorder present already at birth (McLellan, Lewis, O’Brien, & Kleber, 2000). Active involvement and experimentation with alcohol usually begins during adolescence with and under pressure from peers (Burlaka, 2017). Adolescents who drink more alcohol usually have more symptoms of conduct disorder as both problems are steered by a common underlying mechanism of neurobehavioral disinhibition (Tarter et al., 2003). It is, therefore, possible that participants who reported heavier alcohol use in this study had been socialized in peer environments where both extensive alcohol use and interpersonal violence were common. These socialization mechanisms may have contributed to the choice of intimate partners and increased the risks for co-occurring violence and substance abuse. However, the other scenario is also possible in which primary IPV victimization increased the risk for subsequent alcohol use (Devries et al., 2014). This explanation is consistent with prior research with young Ukrainian adults who used alcohol as one of the main tools for coping with psychological problems (Burlaka, Churakova, Aavik, Staller, & Delva, 2014) in the absence of adequate mental health services (Burlaka, Churakova, Aavik, & Goldstein, 2014).

Conclusions, limitations and recommendations

In spite of the fact that this study provides rare and rich insight into relationship between family factors and parenting in Ukraine, its cross-sectional design requires caution when interpreting the results. Furthermore, these findings are based on parents’ recall of parenting practices and co-occurring family factors. Notably, majority of parents in this study were women. Several participants (2.5%) were parent figures rather than biological parents. Future studies that test the relationship between family factors and child outcome should be careful to test genetic influences and environmental on child outcomes. However, in this study we utilized only one level of analysis to test within parent variability in the relationship between family processes and parenting. Because the present study did not account for child behavior outcomes it is not clear to what degree harsh parenting practices were associated with child conduct problems. Additionally, these findings speak to the processes within families who have children from 9 to 16 years of age; therefore, it is not clear whether the same processes would affect parenting of children in other age groups. Other limitations of the present study include data from only three out of 27 Ukrainian regions and that the study used participants’ self-reports. Perhaps, the use of corroborative sources, such as reports from other family members, and real-time observations of parenting behaviors could have strengthened the results. Finally, experimental designs of the future studies could help clarify causal relationships between parenting and parent-level and family-level variables. We also encourage future researchers to use qualitative and mixed method designs to explore cultural aspects of parenting in Ukraine.

In spite of the limitations of a cross-sectional study, this research makes an important contribution to the very limited literature on relationship between alcohol use, violence, family system characteristics and parenting in Ukraine. It is crucial that practitioners and policymakers are aware of these patterns that increase the risk for inadequate parenting. Although not causal, these risk markers can be very useful for development of effective and culturally sensitive interventions with parents in Ukraine and in other countries with similar social and historical backgrounds. It is important to make these interventions affordable for Ukrainian parents who mostly live in poverty. Through these interventions, Ukrainian parents should be encouraged to avoid corporal punishment, to supervise and actively participate in the lives of their children and reward their good behaviors. Moreover, practitioners who work with parents should acknowledge the potential adverse impact of frequent alcohol use, unbalanced family systems and IPV on the parenting.