Deciding on Discipline: The Importance of Parent Demeanor in the Transmission of Discipline Practices

: Although child abuse is a social problem in the United States, many cases go unreported because there is not a consensus as to what disciplinary actions are deemed abusive. Thus, it is paramount to understand the demarcation between physical punishment and physical abuse among parents and their use of certain forms of discipline. This study examines how discipline experienced by adolescent respondents affects their choice of discipline practices in adulthood. A random sample of residents was selected from three South Carolina counties using the 2016 state voter registration list. Respondents were mailed a survey asking questions pertaining to their disciplinary practices and experiences. Analyses were conducted using the ordinary least squares regression. Those who experienced abusive discipline as a child were signiﬁcantly less likely to report that they use the same discipline techniques as their parents. However, adding parenting traits into the model revealed a mediation effect. Abusive discipline no longer plays a signiﬁcant role in how respondents discipline their own children once the perceived demeanor of their parent is taken into consideration. These ﬁndings suggest that disciplinary techniques are less important than a parent’s attitude when correcting their children’s behavior. Implications for the current research, limitations, and directions for future research are discussed.


Introduction
Child abuse is a major social problem in the United States, even though the U.S. is one of the most advanced nations in the Western world. In 2015, Child Protective Services (CPS) received four million referrals for alleged maltreatment of children and approximately 2.2 million of these referrals (4.1 million children) were screened by CPS and further action was deemed necessary (Child Welfare Information Gateway 2017). In addition, in 2015, approximately 683,000 children were victims of abuse for a rate of 9.2 per 1000 children. These numbers reveal a 3.8% increase in child abuse since 2011 (U.S. Department of Health and Human Services 2015). The rate of physical abuse specifically is 4.4 per 1000 children with children between the ages of six and eight having the highest rates of this form of abuse (Sedlak et al. 2010). What is more alarming is that although these numbers are high, they actually underreport child abuse (Rodriguez and Sutherland 1999). Some cases are not reported because many parents in the U.S. do not see certain discipline techniques involving physical punishment as inappropriate or abusive.
There is a fine line between physical abuse and physical punishment, and precipitating factors seem to overlap (Straus 1983). As a result, knowing the elements that impact parents' choice of discipline can contribute to understanding the dynamics of abuse. Research has shown that factors such as the parent's own childhood experiences with discipline influence their parenting practices (Kaufman and Zigler 1993). In fact, the intergenerational transmission of violence hypothesis, which states that individuals exposed to abuse are more likely to perpetrate abuse themselves, is established in the literature (Craft and Serovich 2005). Moreover, experiencing harsh and abusive parenting during childhood is a risk factor for parents treating their own children negatively (Pears and Capaldi 2001). Finally, several studies have shown an intergenerational transmission of parenting attitudes and behaviors. For example, parents who take a harsh discipline approach to childrearing tend to transmit these beliefs and practices to their children (Seay et al. 2016).
Research has also revealed several other factors related to child abuse, including educational attainment, gender, and religiosity (Gershoff et al. 1999). Research has shown that female parents and those with lower levels of education are more likely to use physical discipline than male parents and those with higher educational attainment (Jackson et al. 1999). Gershoff et al. (1999) found that parents with conservative Protestant values are more likely to use corporal punishment compared to parents with other religious values.
Given what we know regarding the intergenerational transmission of violence and the social factors associated with discipline, this exploratory study seeks to examine how childhood experiences influence parental beliefs and disciplinary practices in adulthood for a small sample of respondents in three counties from the Pee Dee Region in Northeast South Carolina. Specifically, this paper evaluates the types of discipline respondents received as children, the perceived demeanor of their parents during discipline, and how these factors influence respondents' perceptions of how they discipline(d) their own children. Demographic characteristics were also examined to explore correlates of parenting beliefs and discipline.

Intergenerational Transmission of Violence, Parental Beliefs, and Discipline
The intergenerational transmission of violence refers to the notion that experiencing abuse or witnessing violence as a child leads to adult perpetration of violence toward their spouses and partners and their own children (Widom and Wilson 2015). Social Learning Theory is one major perspective used to explain how patterns of abuse and violence occur across generations. According to this theory, children partially learn behaviors by observing and imitating others and such behaviors are reinforced through social interactions. Parents provide a highly influential model of behavior and parents who are violent or aggressive, whether with each other or in their disciplinary choices, provide a social context where violence is normal, appropriate, and used to resolve conflict (Bandura 1973). Research supports a social learning model to explain the intergenerational transmission of violence. Baron and Richardson (1994) reveal that exposure to corporal punishment during childhood was related to violent behavior in adulthood. Additionally, Simons and Wurtele (2010) find that children who experience physical punishment tend to view aggression as an appropriate means for resolving conflict. Therefore, parents who resort to violence and corporal or physical punishment to solve problems and conflict may set the stage for their children's behaviors both in child and adulthood.
There is a vast amount of literature that investigates the intergenerational transmission of violence to examine patterns of abuse across generations. In fact, research generally supports the notion that parents who experience abuse during childhood are at an increased risk for perpetuating child abuse (Craft and Serovich 2005;Heyman and Slep 2002;Pears and Capaldi 2001). However, the extent of that risk is not clear in the literature. For instance, Thornberry et al. (2013) find that parents who were exposed to abuse as children were 2.6 times more likely to abuse their own children than parents who were not exposed to abuse as children. Other research reveals a weak to moderate risk for the intergenerational transmission of abuse (Leve et al. 2015;Thornberry et al. 2012). Many studies are critiqued, however, due to methodological and/or sample limitations. Research conducted with a stronger methodology or more representative samples reveal inconsistent support for the intergenerational transmission of abuse/violence hypothesis (Thornberry et al. 2012). Nonetheless, more recent research reveals support for this thesis. In a sample of adolescent mothers, Putnam-Hornstein et al. (2015) find that a history of abuse or maltreatment predicted later abuse and involvement by Child Protective Services. Bartlett et al. (2017) also expose a relationship between experiencing abuse as a child and committing subsequent child abuse after becoming a parent; whether or not a report of abuse or maltreatment was substantiated, type of abuse, and perpetrator type influenced this relationship. Specifically, children of adolescent mothers with a history of abuse were 50% more likely to experience maltreatment. Moreover, the risk of child abuse increased more than 300% when mothers have at least one report of experiencing multiple types of abuse.
In addition to the possibility of violence and abuse being passed down to the next generation, the literature reveals that both supportive and harsh disciplinary practices may also be transmitted from parent to child. Older research shows that children whose parents used discussion to resolve conflict tend to also use that same method as adults (Jorgenson 1985;Steinmetz 1977). Moreover, Chen and Kaplan (2001) find that supportive and warm discipline transcends generational lines and those who experience this type of parenting style as children are more likely to adopt such beliefs and practices with their own children. This intergenerational transmission of positive parenting was also found among mothers who experienced a trusting relationship with their parents, a positive family environment, and were not exposed to an authoritarian parenting style as children . Finally, Shaffer et al. (2009) collected longitudinal data on a sample of children when they were 10 years old (time 1) and approximately 20 years later (time 2) when the children were grown and had become parents. Results show that overall parenting techniques were transmitted between parent and child, as the individuals in the study tended to employ the same parenting practices they experienced as children.

Parental Personality Traits
Parental personality traits or demeanor is related to disciplinary and abuse outcomes. In fact, scholars suggest that parental traits or demeanor play a significant role in how parents interact with their children, the quality of the parent-child relationship (Belsky and Barends 2002;Vondra et al. 2005), and how discipline is administered (Socolar et al. 2007). Several traits have been suggested to influence both harsh discipline and child maltreatment. Difficulties controlling anger is found to be a key characteristic of an abuser (Ammerman 1990;Rodriguez and Green 1997) and research also shows that the level of physical discipline given to a child depended on how angry a parent was with their child's behavior (Peterson et al. 1994). Findings based on a meta-analysis on the risk factors for child abuse reveal a large effect size between parental anger and physical abuse. In addition, the quality of the parent-child relationship is found to be a significant factor for both physical abuse and neglect (Stith et al. 2009). Moreover, research reveals several other negative traits that influence child maltreatment, specifically physical abuse and neglect. These traits include, but are not limited to, low parental involvement, low father warmth, maternal alienation, dissatisfaction, and hostility (Brown et al. 1998).
In addition to traits predicting discipline practices, there is some evidence that parental dispositions may be transmitted to their children. Patterson et al. (1989) suggest that children learn to use aggressive behavior toward others as a result of being exposed to an angry and hostile parenting demeanor. Moreover, research by Chen and Kaplan (2001) revealed that children exhibit the same demeanor as their parents when disciplining their own children later in life. Together, these findings suggest support for the intergenerational transmission of parenting traits. It stands to reason, therefore, that if children are internalizing and demonstrating the same traits as their parents, they may follow in their parents' footsteps regarding their own disciplinary practices. Thus, it is vital that parental personality traits or parental demeanor are examined in the present research. Learning violence or abuse and disciplinary choices from one's own experience as a child are only two pieces of the larger parent-child relationship puzzle.

Other Factors That Influence Parental Beliefs and Discipline Choices
Several other social factors contribute to child maltreatment and influence parental beliefs, along with the type of discipline they choose to utilize on their children. These social factors or characteristics include the religiosity, racial background, age, and gender of the parent (Wolfner and Gelles 1993). Religion is a predictor of discipline style as individuals from various religious affiliations cite the Bible as validation for using corporal punishment (Carey 1994). Indeed, research suggests that parents from a Conservative Protestant religious background are more likely to use physical discipline than other religions (Ellison et al. 1996;Grasmick et al. 1992). Gershoff et al. (1999) find that Conservative Protestant parents spanked their children significantly more than other religions and were less likely to report negative outcomes of spanking, such as an increase in their child's aggression. Hence, the literature consistently shows conservative religions back the use of harsh, physical discipline of children.
Additionally, the race, age, and gender of the parent influences parental discipline practices. Regalado et al. (2004) reveal that African American parents were twice as likely to use corporal punishment as White parents. Additionally, Barkin et al. (2007) find that African American parents were marginally more likely to spank their children and less likely to use time-outs to correct their children's behavior than White parents. As for age, research has found adolescent mothers are at an increased risk of committing physical abuse, sexual abuse, and neglect (Brown et al. 1998). Finally, regarding gender, females have been found to be more likely to commit physical abuse, corporal punishment, or neglect given that mothers typically have more interactions with their children than fathers (Gelles 1997;Jackson et al. 1999). Given these differences, these demographic predictors are important to consider in the current study.

Expectations for the Current Research
Based on the literature summarized above, the cycle of violence is a key perspective to explain how harsh parenting and abuse may occur across multiple generations. The literature also suggests that parenting attitudes, beliefs, and discipline practices are transmitted to younger generations. An additional factor that influences this transmission is the traits or demeanor exhibited by the parents during discipline. Hence, this study explores the intergenerational transmission of violence and parenting practices by investigating different types of discipline experienced during childhood, the demeanor of parents while administering discipline, and whether this influences how one perceives they discipline their own children years later. To address these issues, the current study poses the following research questions: (1) How does experiencing different discipline techniques impact one's perception of their parent's influence on their own discipline practices? (2) How does a parent's demeanor during discipline impact one's perception of their parent's influence on their own disciplining practices?

Sample
For this exploratory study, a random sample of residents from three counties (Florence, Chesterfield, and Marlboro) in Northeast South Carolina was selected from voter registration lists acquired from the South Carolina State Election Commission. The voter registration lists included the resident's name, address, and select demographics. The total number of residents randomly selected for this study was 850 for Florence County, 450 for Chesterfield County, and 450 for Marlboro County. Given Florence County is larger in population and contains more registered voters than Chesterfield and Marlboro Counties, a larger sample was selected for Florence. The samples were randomly selected by (1) importing the registration lists into SPSS and (2) using the random sample of cases function (found within the select cases function in SPSS) to select respondents from each county.

Data Collection
The authors designed a 101-item questionnaire booklet that was sent via first class mail to each respondent in the Florence, Chesterfield and Marlboro County samples. A stamped, addressed return envelope and two copies of the consent form were included with the questionnaire booklet. Respondents were instructed to sign and return one copy of the consent form with the completed questionnaire and retain the other copy for their records. Respondents were given the chance to be entered into a random drawing to win one of eight $25.00 Wal-Mart gift cards. Respondents were instructed to check a box on the consent form to indicate if they wanted to be entered into the drawing. A $1.00 bill was paper-clipped to the questionnaire booklet sent to all respondents in the Chesterfield and Marlboro samples as further incentive to participate in the study. Due to budget limitations, however, only a subset (approximately 24%) of the Florence sample received a $1.00 token. Approximately two weeks after the initial mailing, a reminder postcard was sent to all non-responders. In total, 213 questionnaires were completed and returned, yielding an overall response rate of 12.2%. Data were collected from January 2017 to September 2017.
Prior to engaging in data collection, the study was approved by the Francis Marion University Institutional Review Board. Those who volunteered to participate were not subject to risks beyond what is encountered in daily life. Confidentiality and anonymity were maintained through all phases of the data collection process. All identifying information was removed from the questionnaire and all results are reported in the aggregate.

Sample Selection
To determine the influence of disciplinary techniques on the transmission of discipline practices, respondents who indicated that they were parents were selected for inclusion in the data analysis. There were 179 parents within the larger sample. Upon removing cases with missing data, 130 respondents remained for analysis.

Dependent Variable
The dependent variable in this study is whether respondents perceive that their discipline techniques were influenced by the discipline they received as a child. Two questions were used in the construction of the dependent variable: "My parents'/guardians' method of discipline influences how I discipline(d) my child/children" and "I discipline(d) my child/children similar to how I was disciplined by my parents/guardian." These questions were measured on a six-point Likert scale (1 = Strongly Disagree to 6 = Strongly Agree). Respondents who did not answer both of these questions were removed from the analysis.

Independent Variables
The primary variables of interest measure the methods of discipline respondents experienced as a child. Respondents were asked how often their parent/guardian: (1) gave them a time-out; (2) took away privileges; (3) used humiliation to correct their behavior; (4) yelled at them; (5) cussed/swore at them; (6) threatened them; (7) hit them with a fist; (8) pushed, grabbed, or shoved them; (9) slapped them with an open hand; and (10) spanked them. These discipline techniques were measured on a five-point Likert scale (1 = Never to 5 = Always). In order to ensure an adequate sample size, mean imputation was used to estimate the answers of those who did not provide responses for three or fewer of the above discipline techniques. If more than three questions were left unanswered by a respondent, mean imputation was not conducted and those individuals were removed from the analysis.
Parent traits or demeanor served as another independent variable. Respondents were asked to rate their parent's/guardian's demeanor when administering discipline. Specifically, they were asked if their parent/guardian was: (1) calm; (2) loving; (3) patient; (4) consistent; (5) angry; (6) disrespectful; and (7) indecisive. These characteristics were measured on a six-point Likert scale (1 = Strongly Disagree to 6 = Strongly Agree). Characteristics that are perceived to be negative traits (angry, disrespectful, and indecisive) were reverse coded so that 1 = Strongly Agree and 6 = Strongly Disagree. For consistency, mean imputation was used to estimate the answers of those who did not provide a response two or fewer of the above characteristics. Those failing to respond to more than two questions were removed from the analysis.

Control Variables
Several demographic characteristics were controlled, including race, gender, and age of the respondent. Race was dummy coded to 1 = white and 0 = nonwhite, while gender was dummy coded to 1 = female and 0 = male. Age was measured as the respondent's age on their last birthday. The role of religion was also controlled through two distinct measures. Respondents were asked to report their religion (1 = Baptist and 0 = All others) as well as how often they attend religious services a month. To take into account the importance of family, respondents were asked to report how close they are to their family on a five-point Likert scale (1 = Not at all close to 5 = Extremely close). Finally, the primary disciplinarian of the respondent was controlled. Mother was used as the reference category, while father and another disciplinarian were included in the model.

Descriptive Statistics and Data Reduction
Table 1 provides the descriptive statistics for the dependent, explanatory, and control variables. On average, respondents somewhat agree that their discipline practices are influenced by how they were disciplined and that they discipline similarly to how their parents disciplined them. Additionally, respondents reported that their parents yelled at them, took away their privileges, or spanked them more frequently than they used other forms of discipline. On average, these techniques were used rarely to sometimes, unlike all other discipline techniques, which were reported to occur never or rarely. Generally, respondents somewhat agreed to agreed that their parent's demeanor during discipline was calm, loving, patient, and consistent. Conversely, they somewhat disagreed to disagreed that their parent's demeanor was angry, and disagreed to strongly disagreed that it was disrespectful or indecisive.
Analyzing the control variables reveals that more than two-thirds of respondents were female, nearly three-fourths were white, and on average, they were 57 years old. Less than half of the participants were Baptist and respondents reported attending religious services more than five times a month, on average. Additionally, respondents reported being somewhat to extremely close to their family. Finally, the mother was reported as being the disciplinarian by 60% of respondents.
Several measures utilized in the current study are theoretically and statistically associated with each other (correlation matrix available upon request). To ensure similar concepts are analyzed together and multicollinearity is reduced, obliquely rotated principal components factor analyses were conducted separately for the dependent variable and the explanatory variables. By conventional standards, variables with a factor loading score of at least 0.50 are thought to load together. Additionally, factors that produce an eigenvalue of at least 1 can be kept for inclusion in the regression analyses (Land et al. 1990). Table 2 displays the factor loading scores and eigenvalues of the three distinct factors produced from performing principal components factor analyses. The first factor includes the two measures of the dependent variable, which load equally within the factor. The second factor measures what would be considered abusive disciplining practices (both physically and emotionally), including being hit with a fist, slapped, pushed, yelled at, cussed at, threatened, and humiliated. Reliability analyses support the use of a factor to measure this variable, revealing a Cronbach's alpha of 0.886. Furthermore, the removal of any of these discipline techniques would diminish, rather than improve, the reliability of the factor. The remaining discipline techniques (parent spanked, gave a time-out, and took away privileges) did not load into a factor, and thus, are analyzed individually. Prior research suggests that spanking should load with measures of physical and emotional abuse (Afifi et al. 2017) as it is a harsh discipline technique, and like abuse, is on a continuum of violence against children (Dussich and Maekoya 2007). However, the sample from the current study is drawn from a southern state and southerners tend to be more accepting of corporal punishment (Straus and Mathur 1996). Thus, respondents from this study may not consider spanking to be violent and abusive. The final factor measures the parent's demeanor during discipline. All measures of parental demeanor loaded together, including those that were reverse coded for ease of interpretation. Reliability analyses also support the use of this factor as is, with a Cronbach's alpha of 0.913. Removal of any measure would not improve the factor's reliability.

Analytical Method
To determine what may influence a respondent's perception that their own discipline techniques are influenced by or similar to those of their parents, ordinary least squares (OLS) regression analyses were employed. The first model includes only control variables to establish a baseline adjusted R 2 value. Adjusted R 2 is reported instead of R 2 because unlike R 2 , it takes into account the number of variables in the model and only increases if any added variables are contributing to the explanatory power of the model (Frost 2013). Variables measuring each independent variable were then analyzed with the control variables to determine how much, if any, additional variance these measures were capturing. This method also allowed for the detection of any potential mediating effects. Once complete, the final model including all control and explanatory variables was run. Table 3 presents the OLS regression results predicting respondents' perceptions of parental influence on their own discipline practices. Model 1, the baseline model, shows that the control variables explain 12.6% of the variance in the model with only frequency of attending religious services and perceived closeness to the family attaining significance. Those who attend religious services more frequently and those who report being closer to their family are more likely to perceive that their own discipline practices are similar to or influenced by those of their parents. In Model 2, the measures of discipline experienced by the respondent were included. Experiencing abusive discipline and being spanked were both negative and significantly related to the respondents' perceptions of their own discipline techniques. Those who experienced either disciplinary technique more frequently were less likely to report that their discipline practices are similar to or were influenced by their parents. Additionally, frequency of attending religious services and perceived closeness to family maintained a positive, significant relationship. Religious affiliation is negative and marginally associated with reporting that one's discipline practices are similar to or were influenced by their parents. Specifically, those who identify as Baptist are less likely to report similarities to or influence by their parents when disciplining their own children. Finally, age is negatively and significantly related to perceiving one's disciplinary techniques as being similar to or influenced by one's parents. As age increases, respondents are less likely to report parental similarities or influence in discipline techniques. This model explains 18% of the variance, which is an improvement in variance explained over the baseline model.

Results
Model 3 in Table 3 removes the measures of discipline experienced and replaces them with the factor representing parental demeanor while administering discipline. This model reveals a positive, significant relationship between parental demeanor and perceptions of parental influence on discipline practices. Those who viewed their parents' demeanor positively at the time of the discipline are significantly more likely to report that their own practices are similar to or influenced by those of their parents. Family closeness is the only control variable that maintains significance in this model. The adjusted R 2 shows that approximately 24% of the variance in the dependent variable is explained by this model.
The final model, Model 4, is the full model and includes all control and explanatory variables. In this model, parent demeanor is the only explanatory variable that maintains significance, though the level of significance and the size of the effect are smaller than in Model 3. With regard to the control variables, family closeness, race, and religious affiliation are significantly related to perceptions of parental influence on discipline practices, though the latter two are only marginally significant. As seen in all prior models, as family closeness increases, perceiving discipline techniques to be similar to or influenced by one's parents increases as well. Whites are significantly more likely to perceive that their own discipline practices are similar to or influenced by those of their parents. Additionally, those who identify as Baptist are less likely to perceive that their own discipline techniques are similar to or influenced by their parents, as was seen in Model 2. The full model is shown to explain slightly more than 23% of the variance.
Model 4 also indicates that parent demeanor may have a mediating effect on the relationship between two of the four forms of discipline experienced by respondents and their perceptions of parental influence on their own discipline practices. In this model, abusive discipline and being spanked both lose significance and their coefficients are absolutely smaller in size than they were in Model 2. As part of a post-hoc analysis, mediation was tested using a bootstrapping method (Shrout and Bolger 2002). Using the PROCESS macro in SPSS (Hayes 2019), 5000 bootstrap samples with replacement were performed. This test revealed that parent demeanor mediates the relationship between abusive discipline techniques and perceived parental influence on discipline practices at a 99% level of confidence. This variable also mediates the effect between being spanked and perceived parental influence on discipline practices at a 99% level of confidence.

Discussion
The intergenerational transmission of violence or abuse, in conjunction with the intergenerational transmission of parenting attitudes and behaviors, provides an explanation for parental choices in administering discipline (Craft and Serovich 2005;Pears and Capaldi 2001). Understanding the factors that influence parents' choice of discipline may shed light on the intricacies of child abuse and harsh parenting, including the impact one's experience with discipline during childhood has on such choices in adulthood. In order to further explore this issue, the current study utilized a sample of respondents from three counties in northeast South Carolina. Specifically, the types of discipline this sample received as children, the importance of parental demeanor during discipline, and their perception of how they discipline(d) their own children was examined.
Results revealed an important finding involving parental demeanor, which was statistically significant in the final model. Specifically, respondents who perceived their parents to be calm, loving, patient, and consistent while administering discipline are more likely to perceive their own parenting practices to be similar to their parents. These findings are consistent with the previous literature on the importance of parental traits and demeanor on the parent-child relationship. Socolar et al. (2007) suggest that the type of discipline is only one factor affecting child outcomes and that research should also focus on how parents administer discipline. The findings from their longitudinal study on the parental discipline of young children reveal that both corporal punishment and a negative parental demeanor increase over time, indicating the notion that physical discipline may occur in conjunction with parents who may be stern, angry, rejecting, or use verbal aggression. Indeed, spanking and a negative parental demeanor are correlated (Socolar and Stein 1995) with research showing that anger affects how much physical discipline is given to a child (Peterson et al. 1994). Further, when children experience discipline with a parental demeanor that is warm and supportive, they tend to have a similar demeanor when parenting their own children later (Chen and Kaplan 2001). Thus, it is reasonable to expect that respondents from the current study who perceive their parents as having a positive demeanor while administering discipline would most likely have a good relationship with their parents and as a result would perceive that they discipline their own children in the same manner.
Spanking and abusive discipline techniques were statistically significant in Model 2, but unlike parental demeanor, the relationship between both techniques and respondents perceiving their own parenting practices to be influenced by or similar to their parents was negative. In contrast to the intergenerational transmission of violence and discipline practices, respondents who reported experiencing spanking and abusive discipline practices during childhood report utilizing discipline practices that are dissimilar to their parents. Respondents in this sample perceive themselves as not following in the footsteps of their abusive parents or parents who used corporal punishment, thus indicating a break in the cycle of violence and abuse across generations for these individuals. Although research generally supports the intergenerational transmission of violence (Bartlett et al. 2017;Pears and Capaldi 2001), other literature yields findings that show either mixed support or no support for this phenomenon. For instance, studies that examine the level of risk for the intergenerational transmission of violence show either weak to moderate support or no empirical evidence (Leve et al. 2015;Thornberry et al. 2012). Hence, the findings from the current study are consistent with the latter literature on this phenomenon. Moreover, spanking and abusive discipline was no longer significant in the final model, suggesting that for this particular sample the type of discipline is not as important as parental demeanor when it comes to whether one decides to parent similarly to previous generations.
While not an initial research question, the analysis revealed that parental demeanor has a mediating effect on the relationship between two measures of disciplinary techniques and the perception of parental influence on the respondent's own discipline practices during adulthood. Abusive discipline and being spanked were significant predictors of parental influence on one's perception of their parenting practices; however, these discipline techniques lost significance when they were added in the final model with parental demeanor. This suggests it is not the disciplinary techniques experienced as a child that impacts how one perceives they discipline their children; rather it is the parental demeanor during discipline that is key. In other words, parenting practices may be handed down to the next generation, but for these respondents the demeanor of the parent while administering discipline affects that transmission. Thus, a focus on demeanor may result in positive parenting outcomes in the intergenerational transmission of parental beliefs and practices and may result in reducing the cycle of violence/abuse across generations.
This study is important in the context of childhood and adult mental health outcomes. Harsh discipline and spanking are linked to aggression among children (Weiss et al. 1992), more depression and externalizing behavior among adolescents (Bender et al. 2007), and an increase in suicide attempts and drug and alcohol use/abuse in adulthood (Afifi et al. 2017). However, this exploratory study reveals the importance of having a calm, loving, patient, and consistent demeanor while giving discipline. Indeed, McLoyd and Smith (2004) conducted a study of maternal emotional support among Caucasian, African American, and Latino children. They find that emotional support moderates the relationship between spanking and behavioral problems, suggesting the vitality of parental demeanor and discipline. Though not generalizable, the findings of this study align with prior research, further informing mental health counselors and other family advocates on how to better instruct and work with parents on their demeanor in this context.
Although this study sheds light on the important relationship between the intergenerational transmission of violence/disciplinary practices and parental demeanor, there are several limitations that must be addressed. The results of this study are based on a small sample of residents from one region in South Carolina. The study also yielded a poor response rate despite mailing multiple reminders and utilizing tokens as an incentive to complete the questionnaire. As a result, the current study may suffer from non-response bias (Hager et al. 2003). Given the small sample size and poor response rate, these findings are not generalizable to the population from which this sample was drawn and must be treated as an exploratory analysis. Future research should explore these research questions using larger, representative samples that are generalizable to the U.S. population. Such studies will increase knowledge as to whether the patterns found in this study are applicable beyond a small, non-generalizable sample. Additionally, the dependent variable is based on respondents' perceptions of discipline practices. Due to survey limitations, data on actual discipline techniques used by respondents were not available. Furthermore, the data are retrospective as the items on the questionnaire ask respondents to reflect on the discipline they received as a child and their perceived parents' demeanor during discipline. Memory and recall of events during childhood may be faulty which can reduce both validity and reliability; however, there is research that shows that retrospective data, or data that involves recalling abusive discipline that occurred during childhood, is valid and reliable (Hardt et al. 2010;Widom and Morris 1997).
Moreover, this study did not focus on the influence of socioeconomic status (SES) or the occupation of the respondent or the respondent's parents. Research consistently suggests that parental beliefs regarding discipline vary depending on SES and parental occupation (Gunnoe and Mariner 1997;Hoff et al. 2002). Furthermore, some research reveals that a low SES increases the risk of child abuse and neglect (Brown et al. 1998;Whipple and Webster-Stratton 1991). Future analyses need to incorporate the SES and occupation of both respondents and their parents. This will enable researchers to not only understand how these aspects influence discipline practices, but to also examine if intergenerational social mobility (i.e., a child has a higher SES or occupational prestige than his or her parents) plays a role in the intergenerational transmission of discipline practices. Finally, the sensitive nature of this study may yield responses influenced by social desirability, which is a bias that results from respondents who may feel threatened by a survey question or are trying to avoid embarrassing answers (Fisher 1993). Although the survey was anonymous, respondents may not be forthcoming on answering questions about harsh physical discipline or abuse.
The current study contributes to the literature that recognizes the influence of parental demeanor on the intergenerational transmission of violence and parenting practices. However, additional research is needed to fully understand discipline practices across generations and establish causality. Longitudinal data from larger, representative samples would fill this void. Researchers could gain insight into why some individuals perpetuate harmful behaviors and further understand how parental demeanor mediates the transmission of disciplinary practices in families. Such knowledge is vital for reducing the incidence of child abuse in future generations.