Abstract
The Two Facet Model of Parental Psychological Control (PPC) stipulates that both intrusiveness and emotional manipulation are core facets of PPC, and intrusiveness varies by culture. However, established scales neglected to measure intrusiveness and operationalized intrusiveness with results mainly from white/European families. We tested the Two Facet Model of Parental Psychological Control with Korean adolescents (N = 354, 13–16 years old, 207 female-identified adolescents) via an online survey. We assessed what Korean adolescents believed to be intrusive parenting by asking about 25 areas in their lives. We measured emotional manipulation based on its definitions and tactics. As hypothesized, the intrusiveness and emotional manipulation scales were significantly positively correlated with established PPC scales, and intrusiveness functioned differently from emotional manipulation. These results buttress the utility of separately measuring intrusiveness and emotional manipulation. Hence, we provide empirical support for the Two Facet Model of Parental Psychological Control.
Parental Psychological Control
Parental Psychological Control (PPC) refers to a set of intrusive parenting behaviors that emotionally manipulate children to obey their parents (Choe & Read, 2019). Since Schaefer (1965) first measured PPC, and Barber (1996) revisited PPC, more than 750 published research papers reported PPC’s connections to a wide range of negative consequences for children’s development (Choe et al., 2023). Thus, studying PPC holds grave importance to understand healthy development in children. The Two Facet Model of Parental Psychological Control (Choe et al., 2023) theoretically specifies that PPC rests on both intrusiveness and emotional manipulation as its core facets, with intrusiveness varying by culture and developmental stage. Yet existing scales of PPC do not distinguish clearly between intrusiveness and emotional manipulation and were developed primarily based on data from white/European families. Thus, the meanings and manifestations of intrusiveness in non-white/European cultures remain understudied. In this study, we tested the Two Facet Model of Parental Psychological Control, with a new scale that separately measures intrusiveness and emotional manipulation, from the perspectives of Korean middle school adolescents.
The Two Facet Model of Parental Psychological Control and Cultural Relevance of PPC
PPC was conceptualized as parental behavior that intrudes upon the psychological and emotional development of the child (Barber, 1996). Moreover, previous research on PPC acknowledged its compelling force to change children’s thoughts, feelings, and behaviors (Barber et al., 2012; Soenens & Vansteenkiste, 2010). Such emotionally manipulative tactics include inducing guilt and shame, withdrawing love, invalidating feelings, and attacking personally (Olsen et al., 2002). Furthermore, the Two Facet Model of PPC (Choe et al., 2023) shows that PPC occupies the overlap of both intrusiveness and emotional manipulation (Choe et al., 2023, p. 4).
The Two Facet Model of PPC (Choe et al., 2023) provided a theoretical explanation for a discrepancy between the argument from some scholars that PPC might be associated with fewer adverse effects and even potential benefits in interdependent cultures compared with independent cultures (Fung & Lau, 2012; Rudy & Halgunseth, 2005), whereas other researchers found that PPC was associated with negative developmental child outcomes in interdependent cultures (e.g., Díez et al., 2019; Lau et al., 2022). The seminal papers that initiated the relativistic argument (Fung & Lau, 2012; Rudy & Halgunseth, 2005) operationalized PPC mainly as emotional manipulation (e.g., inducing guilt) and neglected intrusiveness (Choe et al., 2023). For example, the relativistic argument suggested that a parent emotionally manipulating a child could correct the child’s moral transgressions (e.g., when a child hits another child, a parent tells their child that it hurts their feelings that their child hit another child, then the child wants to restore the parent-child bond and stops hitting). However, a parent trying to stop a child’s hitting is not intrusive, as aggression must be addressed. Although the parent might have pressured the child to obey (the definition of emotional manipulation), the parent’s behavior must also be intrusive to qualify as PPC (Choe et al., 2023). Thus, measuring the intrusiveness facet of PPC is the key to study cultural relevance of PPC (Choe et al., 2023).
Separately Measuring Intrusiveness and Emotional Manipulation
Despite the two distinctive facets of PPC, many items of the two most widely used scales to measure PPC—the Psychological Control Scale–Youth Self-Report (Barber, 1996) and the Dependency-Oriented and Achievement-Oriented Psychological Control Scale (Soenens et al., 2010)—measure both intrusiveness and emotional manipulation simultaneously (Choe et al., 2023). For example, the item “my mother/father is less friendly with me if I do not see things her/his way” has two parts: parents being less friendly (emotional manipulation) and parents intervening when children do not see things as parents do (intrusiveness). Simultaneously measuring intrusiveness and emotional manipulation precludes examining separate impacts of intrusiveness and emotional manipulation. In contrast, measuring each facet of PPC via separate items allows researchers to investigate respective impacts of intrusiveness and emotional manipulation on children’s development, which will provide useful information to design preventive interventions for PPC targeted toward each facet of PPC.
Moreover, items of existing PPC scales appear to capture emotional manipulation rather than intrusiveness. Emotionally manipulative tactics may be more straightforward for people to report about than intrusiveness, and intrusiveness may be more implicit in existing items than emotional manipulation. For example, “walking away from me” or “stopping talking to me” describe discernable actions, whereas “not seeing things in her/his way” (intrusiveness) can entail more possible actions than “stopping talking to me” (emotional manipulation). Moreover, items like “will avoid looking at me when I have disappointed her/him” do not explicitly describe intrusiveness (Choe et al., 2023, p. 5). Therefore, testing items that separately measure intrusiveness and emotional manipulation with the established PPC scales can clarify whether the established PPC scales received responses about emotional manipulation rather than intrusiveness.
Measuring Intrusiveness
Intrusiveness entails parental intervention in areas in which caregivers do not have legitimate authority (Darling et al., 2007, 2008), such as areas that children believe should be up to them to decide. Areas in which caregivers have “legitimate” authority may differ by culture (Darling et al., 2005). In other words, a behavior that is considered to be intrusive in one culture might not be intrusive in another culture. For example, children from different cultural backgrounds may feel differently about caregivers tracking their locations on their phones. Some children (e.g., white/European culture) may think that it is intrusive, whereas other children (e.g., Korean culture) may feel affection that their caregivers care about their health and wellbeing. Ergo intrusiveness is culturally bound (Choe et al., 2023), and accounting for this cultural variability in intrusiveness is crucial to understand how PPC works in diverse cultures.
Because existing measures of PPC were developed primarily using data from white/European families in individualistic cultures, the manifestations of PPC in interdependent cultures have not been thoroughly studied using culturally appropriate measures (Choe et al., 2023). The Two Facet Model of Parental Psychological Control (Choe et al., 2023) stipulates that whether a parent’s behavior is intrusive is filtered through children’s interpretation, and children’s interpretation of PPC can be influenced by surrounding cultural norms and cultural values. For this reason, existing scales of PPC based on individualistic white families’ norms and values cannot simply be extended to culturally diverse families. In the present study, we took a culture-specific approach to measure intrusive parenting in Korean culture.
Given that items from established scales of PPC did not necessarily measure intrusiveness separately from emotional manipulation (Choe et al., 2023), we first consulted the adolescent children’s decision-making and parental legitimacy literatures. Parents’ and adolescent children’s decision-making was first measured in four areas (Dornbusch et al., 1985): choosing his or her clothes; how to spend his or her money; which friends to go out with; and how late he or she can stay out. Parents and adolescents were asked separately “who makes most of the decisions on the following?” with nine possible responses: youth alone, father, mother, both parents, father and youth, mother and youth, parents and youth, other person(s), and nobody. From the parental legitimacy literature, we found the Strategic Disclosure Questionnaire (Darling et al., 2007), with adolescent children answering six questions and parents three questions about each of 18 different issues based on sources of parent–adolescent conflict, such as “seeing problematic friends,” “relation with boy/girlfriend,” “smoking,” and “drinking.” It asked adolescent children to answer how frequently they obeyed their parents with four possible responses: never, sometimes, usually, and always. Later, the Strategic Disclosure Questionnaire (Darling et al., 2008) asked adolescent children to report about 20 issues using two questions; “Is it okay for your parents to set rules about this issue?” and “If you disagree with your parents, do you have to obey?” It provided adolescent children with a binary response option (i.e., yes or no). The issues were under several overlapping themes such as friendship, dating, and problematic behaviors. In sum, both the parents’ and adolescent children’s decision-making and parental legitimacy literatures have investigated who decides what and at which frequencies.
To advance the theory and measurement of intrusiveness, the literature can benefit from unpacking the layers of intrusiveness, beyond issues based on sources of conflict between parents and children. To achieve this aim, the meaning of intrusiveness can be explored regarding both agents in these dynamics: the child and the caregiver. When a child believes that an area of their lives should be up to them to decide, and yet their caregiver decides that area, it can be intrusiveness. When a child believes that it is not legitimate for the caregiver to decide an area of their lives, and yet their caregiver decides that area, it can be intrusiveness. Hence, three ratings derived from the definitions of intrusiveness can capture intrusiveness. First, identify which areas of their lives children consider parental intervention to be inappropriate in a specific culture, by measuring how much adolescents believe each area should be up to them to decide (“belief rating”) and how legitimate it is if their caregiver decides each area instead of them (“legitimacy rating”). Asking both questions can yield different answers, as one asks about the child’s decision-making capacity, whereas the other asks about caregivers’ decision legitimacy. These may not be mutually exclusive for children in certain cultures where they grow up hearing that they should obey their caregivers (e.g., Confucian culture), while they are also developing their own autonomy. Second, evaluate the frequency of intrusiveness by measuring how frequently caregivers decide the areas that children identified as intrusive (“frequency rating”). Therefore, examining the three theoretical ratings for intrusiveness can help us understand the meaning of intrusiveness in different cultures.
Methodologically, the literature lacks a scale with a systematic way of asking about exhaustive areas of children’s lives. Several aspects are missing in the literature. First, the earlier items (issues based on sources of conflict between parents and adolescents) focused on behaviors. However, PPC targets children’s feelings and thoughts as well as behaviors (Barber et al., 1994). Hence, feelings and thoughts should be included in the pool of items. Second, the earlier items were mostly developed using data from white European families. Areas that children in non-white European cultures, such as interdependent cultures, may deal with or be expected to handle should be introduced to the item pools as well, to measure intrusiveness in a culturally sensitive manner. Third, previous scales have used multiple categorical response options with varying numbers of options (two to nine options), which makes it difficult to assess degree or frequency across scales. Using the same answering options across scales to measure frequency or degree of intrusiveness will help ameliorate these measurement issues.
Measuring Emotional Manipulation
Emotional manipulation indicates the pressure that caregivers put on children and the anxiety that caregivers instill in children to coerce children into complying with caregivers’ wishes (Soenens & Vansteenkiste, 2010). Its tactics include inducing guilt (e.g., “tells me all the things s/he has done for me”), shaming (e.g., “tells me I am not a good family member”), withdrawing love (e.g., “avoids looking at me”), invalidating feelings (e.g., “tells me I shouldn’t feel that way”), and attacking personally (e.g., “brings up my past mistakes when criticizing me”) (Olsen et al., 2002). Emotional manipulation tactics entail subtle, covert, and indirect ways to psychologically control children, and these emotional manipulation tactics make children think less of themselves unless they obey their caregivers (Choe et al., 2020).
People in diverse cultures can emotionally manipulate children via different tactics (Choe et al., 2023). In other words, numerous ways exist to express one’s needs indirectly. However, the PPC research field lacks information on how people in diverse cultures differently use emotionally manipulative tactics (Choe et al., 2023). Inducing guilt has been studied in multiple cultures (e.g., Mandara & Pikes, 2008; McKee et al., 2014; Yu et al., 2015, 2019) more than other emotionally manipulative tactics of PPC (Choe et al., 2023). Hence, more research is needed to study how emotional manipulation manifests in different cultures.
Prior items of PPC scales listed different emotionally manipulative tactics. For example, “stops talking to me until I please her/him again” described withholding love, and “blames me for other family members’ problems” indicated attacking personally. However, caregivers in diverse cultures exhibit different behaviors to emotionally manipulate their children (Choe et al., 2023), different from what the earlier items described. One way to address this concern may be asking about each emotionally manipulative tactic and giving examples to help adolescents understand each tactic, to allow them to report a wider range of tactics than the prior PPC items. In addition, asking adolescents directly about the pressure and anxiety they feel to obey their caregivers—the definition of emotional manipulation—can enhance assessing emotional manipulation.
Cultural Frame: Korean Adolescents
To understand what people in a specific culture believe to be intrusive parenting, studying the individual culture is crucial, and the field can have information on multiple cultures by scholars studying each culture. We explain Korean culture to elucidate how both facets of PPC are relevant in this culture. Confucianism has been the dominant ideology in Korea for more than 500 years since the 14th century (JoSuhn dynasty) and greatly impacted Korean culture (Park & Cho, 1995). As Confucianism emphasizes hierarchical organizations and views family as a prototype social organization, people in Confucian cultures tend to expect parents to morally educate their children to ensure harmony and order within the family and within society by extension (Hyun, 2001). In other words, children learn to respect caregivers, which is often equivalent to obeying caregivers. Although the Confucian influence has waned (Zhang et al., 2005), a recent study still revealed the Confucian influence on familial relationships in Korea (Gurín & Brandt, 2024). In the Confucian culture where children are expected to obey adults, children may find it particularly difficult to withstand intrusive control.
Collectivism is deeply rooted in Korean family functioning (Kim et al., 2015). The collectivistic Korean society highly values togetherness and harmony among family members. When family comes first before individuals, and keeping harmony with family members greatly matters, children may feel pressured to not speak up to their caregivers when their caregivers cross their individual boundaries. Consistent with this, Korean adolescents who scored high on collectivism often complied passively with caregivers’ requests and less often defied caregivers’ requests than those who did not score high on collectivism (Soenens et al., 2018). Hence, Korean children may feel pressure to obey caregivers. Taken together, Korean children make an exceptional population in which to study intrusiveness and emotional manipulation.
Developmental Frame: Middle School Adolescents in Korea
In the Korean 6-3-3 education system, middle school ages are the 3 years in the middle, between 6 years of elementary school for (ages 8–13) and 3 years of high school (ages 17–19). School days (time in school) and individual classes in middle school are longer than in elementary school, and first grade students in middle school (typically age 14) already reported more academic stress and lower life satisfaction than fifth grade elementary school students (typically age 12) (Choi et al., 2019). But middle school students are still considered to be children with limited autonomy, compared with high school students who are considered close to adults. Ergo middle school adolescents in Korea may experience dramatically increasing desire for autonomy. The term “middle-school second-year syndrome,” coined to describe middle school adolescents in Korea acting out and experiencing drastic changes, depicts the struggles of middle school adolescents in Korean culture (The Korea Times, 2015). In addition, adolescents have the most conflict with their parents around ages 15 to 16 (Laursen et al., 1998), which falls into middle school in Korea. Thus, middle school adolescents (13–16 years old) make an ideal age group to study intrusiveness and emotional manipulation in Korean culture.
Current Study
This study tests the Two Facet Model of Parental Psychological Control (Choe et al., 2023) by separately measuring intrusiveness and emotional manipulation from a Korean middle school adolescent sample. We assessed internal consistency and confirmatory factor analyses’ (CFA) fit measures. We investigated if adolescents’ gender, age, and subjective perception of family’s wealth and primary caregiver’s education and age as manifest variables were associated with intrusiveness and emotional manipulation as latent variables. We also tested construct validity with established PPC scales, other parenting scales, and adolescents’ adjustment by examining latent variable correlations.
For construct validity, we first examined the validity of these new intrusiveness and emotional manipulation scales with the two mostly widely used PPC scales: Psychological Control Scale–Youth Self-Report (Barber, 1996) and Dependency-Oriented and Achievement-Oriented Psychological Control Scale (Soenens et al., 2010). We also investigated validity with other parenting constructs that are relevant to PPC and that previous studies of PPC reported in Asian samples (little or no data exist for Korean samples). Behavioral control indicates caregivers targeting children’s behaviors as opposed to children’s minds as PPC does (Barber et al., 1994) and was positively correlated with PPC in a Chinese adolescent sample (Wang et al., 2021). Acceptance describes the degree to which the adolescent receives care and attention from caregivers and was negatively correlated with PPC in a Chinese American middle childhood sample (Fung et al., 2017). Autonomy support illustrates the extent to which the adolescent feels listened to about their problems and supported to decide for themselves and was negatively correlated with PPC in Chinese adolescents (Wang et al., 2021). Self-esteem portrays how well the adolescent thinks of themself and was negatively correlated with PPC in Chinese adolescents (Chen et al., 2021). PPC was positively correlated with depression in Chinese adolescents (Qin et al., 2021) and anxiety in Chinese adolescents (Liu et al., 2022). Finally, PPC was positively correlated with adolescent drinking in Chinese adolescents (Wang et al., 2019).
Methods
Procedure
The institutional review board (IRB) of a university in South Korea approved the study and procedures. Homeroom teachers in a middle school in Seoul, South Korea distributed the forms for the caregivers’ permission and adolescent child’s assent to all of the students in three grades (N = 715). The students whose caregivers consented and themselves assented (“participants”) provided the signed forms to their homeroom teachers (N = 374, 52% participation). No financial compensation was offered for their participation, and the information on the students whose caregivers did not consent or themselves did not assent is not known. The students answered the online survey on Qualtrics during a class session in school (all classrooms at the same time), with their homeroom teachers present in the classrooms. Most students took about 16 minutes to finish the online survey. The school sent the signed caregivers’ permission and assent forms to the researcher at the university in Korea, who then verified that the number of permission and assent forms equaled the number of students who answered the online survey. We reported the data collection process to the IRB to ensure that we followed all the approved procedures and also reported the results to the IRB as they requested.
Participants
Data were drawn from the Korean Middle School Adolescents’ Decision Making, Adjustment, and Cultural Values project, conducted in February, 2023. Adolescents attending a middle school in Seoul, South Korea (N = 374) participated in the study. The answers from 20 students contained about 90% missing data. Removing these responses made an analytic sample of 354. Adolescents’ age varied from 13 to 16 years: Mage = 15.39 and SDage = .96. Age information from 12 adolescents was unknown. For gender, 207 students identified as female, whereas 141 students identified as male. Gender information for six students was unknown.
Adolescents identified their primary caregivers’ gender as mostly female (N = 315, 90%) with 299 mothers, 15 grandmothers, and one sister. Some adolescents identified their primary caregiver as male (N = 42, 10%): 31 fathers, two uncles, one brother, and five unknown. Primary caregivers’ age ranged from 29 to 81 years: Mage = 47.61 and SDage = 6.71. Adolescents reported their primary caregiver’s marital status in six categories: 292 married; 9 separated, but not divorced; 31 divorced; 0 cohabiting and not married; 1 in a relationship and not married; 3 single and never married. Marital status for 18 adolescents’ primary caregivers was unknown.
Socioeconomic Status
School Context
The information on the area where the middle school is located can provide insights on the adolescents’ background, as the Korean education system assigns students to public schools near their home addresses. Among the 25 districts in Seoul (“Gu”), the district where this middle school is located ranked 18th in mean income in 2023 (Seoul Credit Guarantee Foundation, 2023). In addition, this Gu has almost three times fewer private educational institutions per 10,000 students than other districts (Seoul Metropolitan Government and Seoul Institute, 2022). Hence, the adolescents came from a poor area and a disadvantaged environment.
Education
We examined primary caregivers’ education levels with 10 options ranging from “less than high school graduation” to “graduated from Doctoral program” (see Supplement). Because most Koreans finish high school education (91% of adults from ages 25–64 in 2022) (Organisation for Economic Co-operation and Development [OECD], 2024), we used high school as the lower limit. Compared with the statistics that 52% of Korean adults from ages 25 to 64 graduated from college (OECD, 2024), and 69% of Korean adults from ages 25 to 34 graduated from college (Ministry of Education in the Republic of Korea, 2023), the primary caregivers’ education level in this sample was less than the average.
Subjective Wealth
We asked about adolescents’ subjective perception of their family’s wealth, with seven options ranging from “struggle to afford basic resources (shelter, food, and clothing)” to “are able to afford luxurious things any time they want” (see Supplement). Information from 19 adolescents was unknown. While the median fell into the category “are able to afford things they want when there are not special occasions,” more than 17% reported that their families were concerned about basic resources (the first three categories). Given that South Korea ranks as the 13th largest economy in the world (WorldData.info, 2024), the participating adolescents came from a relatively low income area in South Korea.
Measures
We wrote the items for this new scale both in Korean and English and discussed them with a research team that is fluent in Korean and English to ensure the items have the same meaning in both languages. For the other existing scales, we found Korean translations that other researchers had used, compared the wording with the English versions, and changed wording slightly with a bilingual research team to represent the same meaning. Then we verified the meanings were the same by back-translation.
Slider
We provided a slider from 0 to 100 for each item to obtain continuous rather than ordinal indicators. In addition, using a slider with numbers can reduce ambiguity of responses like “frequently” or “infrequently,” given that some people may consider five times to be “frequent,” whereas others may consider “frequently” to mean a behavior occurs 10 times. To help adolescents use the slider, the instructions included guidance such as “if your primary caregiver decides the matter once out of 3 times that you consider the matter, it is 1/3 (0.33).” To not bias answers in either direction, we initially placed each slider at 50.
Two Facet Parental Psychological Control Scale
Intrusiveness
We used 25 areas that adolescents may believe are up to them to decide. We first wrote the areas reflecting the targets of PPC—thoughts and feelings as well as behaviors—based on prior conceptual work (Barber, 1996). We then reflected on Korean culture to generate intrusiveness areas such as relationships, time management, career choices, and religion choices (see Table 1 for rationale for each area; refer to the rationale for each area to come up with areas of conflict in another culture). Moreover, we considered collectivistic cultural perspectives and listed areas of contrast between what children in collectivistic cultures may believe are up to them to decide and what may be expected in the family- and group-oriented culture. Lastly, we adopted several intrusiveness areas from the parents’ and adolescent children’s decision-making literature (Dornbusch et al., 1985) and the Strategic Disclosure Questionnaire (Darling et al., 2007, 2008) and expanded the phrases for an exhaustive list for each area. We listed background information for generating each intrusiveness area to guide scholars in generating areas specific to the culture they would study (see Table 1). To ensure that we covered the areas that are relevant to Korean adolescents, we conducted a pilot study with 43 Korean adolescents attending different schools from the main study (a different sample from the main study, ages 14 and 17, see Supplement) in which we showed the 25 areas and asked them an open-ended question if there were any other areas that should be up to them to decide but their caregivers decided instead of them. Most students answered no, and no answer was listed twice. “Whether to end my life and when to end my life” came up as an answer, but given that it was only one student’s answer and due to the school teachers’ concern that asking about suicidal thoughts would encourage students to engage in suicidal behaviors, we did not add it as an area to test in the main study.
Intrusiveness Areas and Background for Generating Each Area.
Belief Rating
The instructions said “Compared to your primary caregiver, how much do you believe this should be up to you to decide?” Although this question for the belief rating may entail degree of beliefs, we decided to measure the frequency of the belief rating for three reasons. First, we need to use the same scale across all three ratings of intrusiveness, and the frequency of primary caregivers’ intervention denotes a literal frequency (number of events). Second, “degree of belief” can be an abstract concept that is difficult for adolescents to answer, whereas “frequency of belief”—how many times each area should be up to them to decide—asks about the number of concrete events that adolescents can recall or calculate patterns (e.g., three times out of seven times I decide this area). Third, quantifying a “degree of belief” is challenging, and one would measure it using a categorical scale with varying number of options such as ranging from disagree to agree. In contrast, “frequency of belief” can be measured with a quantitative number (times of decision-making). To help adolescents answer this rating, we indicated frequency on a slider: 100 = always my decision; 0 = not at all my decision; 50 = half of the time my decision & the other half of the time my primary caregiver’s decision. The Cronbach’s α coefficient for this sample was .95.
Frequency Rating
The instructions said “How frequently does your primary caregiver decide each matter instead of you? This is not based on a particular time duration (such as 24 hours). Please think about the frequency of your primary caregiver deciding each matter below. For example, if your primary caregiver decides the matter once out of 3 times that you consider the matter, it is 1/3 (0.33). Thus, place the slider at 33.” We indicated frequency on a slider: 100 = My primary caregiver always decides this matter instead of me when they are with me; 0 = My primary caregiver does not decide this matter instead of me at all; 50 = My primary caregiver decides this matter instead of me half of the time & I decide this matter the other half of the time. The Cronbach’s α coefficient for this sample was .99.
Legitimacy Rating
The instructions said “How legitimate is it if your primary caregiver decides this matter instead of you?” We measured the frequency of the legitimacy rating for the reasons aforementioned under the belief rating. We indicated frequency on a slider: 100 = always legitimate; 0 = not at all legitimate; 50 = legitimate half of the time & not legitimate the other half of the time. The Cronbach’s α for this sample was .99.
Emotional Manipulation
We created two items based on the definitions of emotional manipulation: pressuring children and instilling anxiety in children to make them obey. We also created five items based on the tactics of emotional manipulation that are relevant to Korean culture. Then we provided examples of each emotional manipulation tactic to help adolescents understand each tactic, and some of the examples came from the established PPC items. The instructions said “Children often feel that they need to change their behaviors or thoughts or feelings based on the caregiver’s opinions. Please indicate how frequently your primary caregiver engages in these behaviors toward you. This is not based on a particular time duration (such as 24 hours). Please think about the time you are with your primary caregiver (including video chats, phone calls, and texts) and indicate the frequency with which your primary caregiver engages in these behaviors. For example, if you are with your primary caregiver 7 times, and your primary caregiver presents the behavior described below 5 times, it is 5/7 (0.71). Thus, place the slider at 71.” We indicated frequency on a slider: 100 = My primary caregiver always engages in the behavior when my primary caregiver is with me; 0 = My primary caregiver does not at all engage in the behavior when my primary caregiver is with me; 50 = My primary caregiver engages in the behavior half of the time when my primary caregiver is with me and does not engage in the behavior the other half of the time when my primary caregiver is with me. The Cronbach’s α coefficient for this sample is .93.
Parental Psychological Control From Established Scales
For the PPC scales below, respondents indicated frequency on a slider in the same way as the Emotional Manipulation Subscale above.
Psychological Control Scale–Youth Self-Report
Students completed the eight items (Barber, 1996). The instructions said “My primary caregiver is a person who . . .”, and a sample item is “is less friendly with me if I do not see things her/his way.” The Cronbach’s α coefficient for this sample was .92.
Dependency-Oriented and Achievement-Oriented Psychological Control Scale
The original scale (Soenens et al., 2010) consists of 20 items with 10 items per subscale. To reduce the burden on students, we administered the five items in each subscale with the highest factor loadings (Soenens et al., 2010, p. 226, Table 1), based on the authors’ recommendation (personal communication with Bart Soenens on 12/20/2022). The instructions said, “My primary caregiver . . .”, and sample items include “blames me that I no longer want to do things that we used to enjoy” for dependency-oriented psychological control (DPC) and “shows that s/he loves me less if I perform badly” for achievement-oriented psychological control (APC). The Cronbach’s α coefficients for this sample were .92 for DPC and .91 for APC.
Parenting
Respondents indicated frequency on a slider, in the same way as the Emotional Manipulation Subscale, for the parenting scales below.
Behavioral Control
Students responded to the Behavioral Control Subscale from the Child Report of Parent Behavior Inventory (Barber et al., 1994; Schaefer, 1965) with four items. The instructions said “My primary caregiver . . .”, and a sample item is “gives me as much freedom as I want.” The Cronbach’s α coefficient for this sample was .76.
Acceptance
Students answered the Acceptance Subscale of the Child Report of Parent Behavior Inventory (Schaefer, 1965) with 10 items. The instructions said “My primary caregiver . . .”, and a sample item includes “gives me a lot of care and attention.” The Cronbach’s α coefficient for this sample was .95.
Autonomy Support
Students completed the Autonomy Support Subscale of the Perceptions of Parents Scales (Grolnick et al., 1997) with seven items. The original subscale has nine items, but to reduce the burden for students, we removed two items that might be less relevant in Korean culture than the other seven items and whose interpretations in Korean language could be ambiguous: “seems to know how I feel about things” and “isn’t very sensitive to many of my needs.” The instructions said “My primary caregiver . . .”, and a sample item includes “helps me to choose my own direction.” The Cronbach’s α coefficient was .73.
Adolescents’ Adjustment
We provided adolescents with this instruction for self-esteem, depression, and anxiety: “These questions ask about your feelings. Please indicate how true each item is to describe you.” We indicated frequency on a slider: 100 = all the time, I feel this way; 0 = not at all, I feel this way; 50 = half of the time, I feel this way, and the other half of the time, I do not feel this way.
Self-Esteem
Students responded to the Brief Rosenberg Self-Esteem Scale (Monteiro et al., 2022) with five items. A sample item includes “On the whole, I am satisfied with myself.” The Cronbach’s α coefficient for this sample was .72.
Depression and Anxiety
Students completed the Depression and Anxiety Subscales of the Depression Anxiety Stress Scale 21 (Lovibond & Lovibond, 1995) with seven items, respectively. Sample items are “I felt life was meaningless” for depression and “I felt scared without any good reason” for anxiety. The Cronbach’s α coefficient for this sample was .92 for depression and .87 for anxiety.
Drinking Frequency
Students answered one question: “During the last 12 months, how many times have you had alcohol or beverages that contain alcohol?” We indicated frequency on a slider: 0 = not at all; 99 = 99 times; 100 = more than 99 times.
Statistical Analyses
Mplus version 8.10 (Muthén & Muthén, 2023) performed CFA and computed latent variable correlations, using the maximum likelihood (ML) robust estimator given the nonnormality of some variables. We performed the Little’s missing completely at random (MCAR) test.
Confirmatory Factor Analyses
Intrusiveness
Given that the 25 areas we tested for intrusiveness theoretically indicate one factor, we tested a one-factor CFA model for each rating for intrusiveness: belief, frequency, and legitimacy. We investigated scree plots to confirm one factor for each rating. Starting from the 25 intrusiveness areas and the belief rating, we removed the intrusiveness area with the lowest factor loading from the intrusiveness belief rating until the comparative fit index (CFI) became greater than .90, based on the recommendation that CFI between .90 to .99 indicates an acceptable fit (Little, 2013). We also checked if the root mean square error of approximation (RMSEA) and the standardized root mean square residual (SRMR) were less than .08. We determined which areas to include based on the factor loadings and theoretical perspectives. After selecting the intrusiveness areas to include for the belief rating, we created parallel factors with the same areas for the frequency and legitimacy ratings.
Emotional Manipulation
Given the emotional manipulation items theoretically suggest one construct, we also tested a one-factor CFA model.
Computation
Mplus version 8.10 (Muthén & Muthén, 2023) computed these CFA models. We used a robust estimator for nonnormality—ML parameter estimates with standard errors and a chi-square test statistic that are robust to nonnormality and nonindependence of observations (MLR)—after consulting with the founder of Mplus regarding the appropriate type of robust estimator in Mplus for this data set (personal communication with Linda Muthén on 1/18/2024). We also performed CFA to create latent variables for all of the other scales (implemented as a one-factor model for each scale), but we did not remove any items from the other scales. We used the latent variables for further analyses described below to test validity of this new scale.
Correlations
We report correlation coefficients and their significance among latent variables. We tested construct validity: if intrusiveness and emotional manipulation were significantly correlated with established PPC scales and other parenting and adolescent adjustment scales that have been associated with the established PPC scales.
Results
Descriptive Statistics
Table 2 shows descriptive statistics of the intrusiveness areas for each of the three ratings, Table 3 shows descriptive statistics for the emotional manipulation items, and Table 4 shows the mean scores for all of the scales and the demographics. The belief rating was negatively skewed as Figure 1 shows; most areas were up to the Korean adolescents in the sample to decide. The frequency rating displayed a bimodal distribution as Figure 1 shows; some students did not think their primary caregivers intervened in these areas, whereas other students thought that their primary caregivers did intervene. The most common response was 0: My primary caregiver did not decide this matter at all instead of me. The legitimacy rating also shows a bimodal distribution as Figure 1 shows. The most common response was 100: It was always legitimate for primary caregiver to decide this matter instead of me. Emotional manipulation was positively skewed, showing the same pattern as the established PPC scales (see Figure 2).
Descriptive Statistics for Intrusiveness Areas by its Ratings.
Note. The belief rating = how frequently this area is up to you to decide; the frequency rating = how frequently your primary caregiver decides each area instead of you; the legitimacy rating = how frequently this area is legitimate for your primary caregiver to decide instead of you.
Descriptive Statistics for Emotional Manipulation Items and Standardized Factor Loadings From Confirmatory Factor Analyses.
Descriptive Statistics for Mean Scores of all of the Scales and Demographic Variables.
Histograms of intrusiveness ratings.
Histograms of emotional manipulation and the established Parental Psychological Control scales.
Missing Data Analyses
Little’s MCAR test indicated that the data were MCAR, χ2(1,043, N = 354) = 1020.65, p > .68. Thus, we used the full information maximum likelihood (FIML) method (Enders, 2010; Little, 2013) to deal with missing data.
Confirmatory Factor Analyses
Intrusiveness
The scree plots we investigated (see Supplement) show one factor for each rating. Table 5 shows fit measures, and Table 6 shows the standardized factor loadings from CFA for the three intrusiveness ratings. Although the 20-area CFA model also fit well (CFI > .90, RMSEA < .06, SRMR < .05), we decided to exclude “whether I take care of younger siblings (e.g., help them with their homework) & how I take care of them” from the factor. Conceptually, this area describes family responsibilities similar to the other excluded items such as “whether I go to family gatherings (e.g., birthday parties, family reunions, funerals, weddings) & which family gatherings I attend” and “whether I do house chores & which chores I do.” These are areas that children in collectivistic cultures may be expected to do, and children might have internalized caregivers’ legitimacy over those areas. Out of 25 areas included in the initial intrusiveness belief rating, fit measures of 19 areas suggested a good fit. Using the same 19 intrusiveness areas, the fit measures of the frequency rating and the legitimacy rating also suggested good fits. Factor loadings showed that the same 19 included areas identified using the belief rating also had the highest factor loadings for the frequency and legitimacy ratings.
Fit Measures of the Scales.
Standardized Factor Loadings for the Intrusiveness Areas by its Ratings, From Confirmatory Factor Analyses.
Note. The number after each rating denotes the number of areas.
Emotional Manipulation
The one-factor CFA model for emotional manipulation with seven items showed a good fit. Factor loadings for the emotional manipulation items were over .60, except for one item: withdrawing love, .56. Theoretically, withdrawing love is a tactic for emotional manipulation. Hence, we kept the item. Table 5 shows the standardized factor loadings for the items.
Construct Validity
We reported latent correlation coefficients and their significance in Table 7.
Correlations Among Latent Variables and Demographics.
Note. p: a.001. b.01. c.05, drinking frequency, age, education, and wealth are items (manifest variables). I_Blf = intrusiveness: the belief rating; I_Frq = intrusiveness: the frequency rating; I_Lgt = intrusiveness: the legitimacy rating; EM = emotional manipulation; PCS = Psychological Control Scale; DPC = dependency-oriented psychological control; APC = achievement-oriented psychological control; BC = behavioral control; A = acceptance; AS = autonomy support; SE = self-esteem; Dprs = depression; Anx = anxiety; Drk = drinking frequency; age = adolescents’ age; ed = primary caregiver’s education; wlt = adolescents’ subjective perception of family’s wealth.
Intrusiveness
Frequency Rating
The frequency rating was significantly positively correlated with the established PPC scales—Barber scale, DPC, and APC—despite the small effect sizes (r = .20–.25) and also with anxiety, drinking frequency in the past 12 months, the legitimacy rating, and emotional manipulation. The frequency rating was significantly negatively correlated with primary caregivers’ education. It was not significantly correlated with behavioral control, acceptance, autonomy support, self-esteem, depression, adolescents’ age, or adolescents’ subjective perception of family’s wealth.
Legitimacy Rating
The legitimacy rating was significantly positively correlated with the established PPC scales—Barber scale, DPC, and APC—and drinking frequency and the frequency rating and negatively correlated with primary caregiver’s education. It was not significantly correlated with behavioral control, acceptance, autonomy support, self-esteem, depression, anxiety, the belief rating, emotional manipulation, adolescents’ age, or adolescents’ subjective perception of family’s wealth.
Belief Rating
The belief rating was significantly positively correlated with behavioral control, acceptance, autonomy support, or adolescents’ subjective perception of family’s wealth and negatively correlated with the established PPC scales—Barber scale, DPC, and APC—and emotional manipulation, depression, and anxiety. It was not significantly correlated with self-esteem, drinking frequency, the frequency rating, the legitimacy rating, adolescents’ age, and primary caregivers’ education.
Emotional Manipulation
Emotional manipulation was significantly positively correlated with the established PPC scales, with large effect sizes (r = .72–.83). It showed the same direction as the established PPC scales; it was significantly negatively correlated with behavioral control, acceptance, autonomy support, and self-esteem and positively correlated with depression, anxiety, and drinking frequency. It and the established PPC scales were not significantly correlated with adolescents’ age, primary caregivers’ education, or adolescents’ subjective perception of family’s wealth.
Discussion
This study tested the Two Facet Model of Parental Psychological Control using a new scale that separately measures intrusiveness and emotional manipulation with Korean middle school adolescents. It addresses important gaps in the literature on the measurement of PPC and the lack of culturally sensitive operationalizations of intrusiveness and emotional manipulation. Our study also provides an example of how to test this scale in a specific culture, enabling scholars to test this scale in each culture and establish culturally sensitive conceptualizations and operationalizations of intrusiveness and emotional manipulation in diverse cultures. This scale shows strong internal consistency, a good fit in one-factor CFA, and good construct validity with other related scales.
Separately Measuring Intrusiveness and Emotional Manipulation
Measuring intrusiveness and emotional manipulation separately revealed that both intrusiveness and emotional manipulation were significantly positively correlated with established PPC scales, yet intrusiveness functioned differently from emotional manipulation. Emotional manipulation showed similar patterns as the established PPC scales, but intrusiveness did not. These results show that intrusiveness and emotional manipulation may affect adolescents in different ways and both can harm adolescents’ development. The reality may be more nuanced than a simple statement that intrusiveness or emotional manipulation impairs children’s development, and separately measuring intrusiveness and emotional manipulation offers ways to test if and how intrusiveness or emotional manipulation operationalized by what people in a given culture believe influences children’s development. This suggests the need to study intrusiveness separately from emotional manipulation. These results empirically support the Two Facet Model of Parental Psychological Control (Choe et al., 2023).
Intrusiveness
Although Confucianism and collectivism greatly influenced Korean culture, Korean middle school adolescents nonetheless reported that most of the time, these areas should be up to them to decide. These results may reflect Korean middle school adolescents’ desires for autonomy, despite the cultural pressure from Confucianism to obey adults (e.g., caregivers) and the pressure from collectivism to place the family before the individual. The frequency rating directly represents parental intrusiveness, given that we included the 19 areas that adolescents reported should be up to them to decide. Interestingly, the distribution for the frequency rating was bimodal. Unlike the stereotype of intrusive Asian caregivers (Chua, 2011), nearly a quarter of the Korean adolescents reported that their primary caregivers did not decide those areas at all. Our results suggest wide variability in intrusiveness within Korean culture and this developmental stage.
The legitimacy rating results show that adolescents are reasonable in acknowledging caregivers’ legitimacy in decision-making. One might expect that adolescents would not acknowledge caregivers’ legitimacy in most areas because they want to decide on their own, and that adolescents would perceive caregivers’ interventions in these areas as intrusive. However, looking at the areas excluded from the intrusiveness factor tells a different story. The responses for “when I wake up & when I go to bed,” “what foods I eat & when I eat,” and “how much time I spend on social media” suggest that adolescents acknowledge caregivers’ legitimacy in deciding health-related areas such as time for sleep, food consumption, and media usage.
The other three areas excluded from making a factor—“whether I go to family gatherings (e.g., birthdays, funerals, weddings) & which family gatherings I attend,” “whether I take care of younger siblings (e.g., help them with their homework) & how I take care of them,” and “whether I do house chores & which chores I do”—may reflect collectivism and filial piety in Korean culture. Because family tends to come before the individual in Korean culture, Korean adolescents may believe that attending family gatherings might be legitimate for the primary caregiver to decide instead of them. Taking care of younger siblings and doing household chores can be a way of Korean adolescents’ paying back their caregivers who are raising them (filial piety). Consistent with this, repayment was considered a facet of filial piety among Korean adolescents and emerging adults (Sung, 1995). Perhaps adolescents in more individualistic cultures believe that attending family gatherings should be up to them to decide, raising children is parents’ responsibility (not older siblings’), and that they could negotiate if or which household chores they are responsible for completing. However, this is speculation at present. Hence measuring the extent to which adolescents in individualistic cultures believe it is legitimate for their primary caregivers to decide these areas can reveal cultural similarities or differences in intrusiveness in future research.
It appears intuitive that adolescents reported depression and anxiety symptoms when their primary caregivers decided the 19 areas that they believed were up to them to decide. However, it is noteworthy that adolescents reported less depression and anxiety when they simply believed the 19 areas were up to them to decide, given that the belief rating does not entail any actions from the caregivers or adolescents themselves. These results may emphasize the importance of beliefs about autonomy in developing adolescents, even in collectivistic and Confucian based cultures like Korean. Interestingly, adolescents reported more frequent drinking in the past 12 months when they acknowledged more legitimacy for their primary caregivers’ decision-making in the 19 areas. This may suggest that Korean adolescents may struggle when they feel pressure from their culture (based on Confucianism and collectivism) to acknowledge caregivers’ legitimacy in the 19 areas they believed should be up to them to decide. This may cause stress to adolescents due to losing autonomy or due to conflicts arising with their caregivers over autonomy, and some adolescents may drink to cope with the stress (Cooper et al., 1992).
The three ratings of intrusiveness reflect different aspects of intrusiveness. The answers from the belief rating indicate which areas in a specific culture that children believe should be up to them to decide. Thus, the belief rating guides which areas to include to make a factor and include in latent variables for all three ratings for analyses. The belief rating differs from the legitimacy rating, whose subject is the primary caregiver, not the child themselves. Interestingly, the belief rating was not significantly correlated with the legitimacy rating in this Korean middle school adolescent sample. It supports our rationale to use different ratings with different subjects, children versus caregivers, and more research applying these ratings in different cultures will provide rich information on what adolescents believe about their decision-making capacity versus caregiver’s legitimacy in each culture. The legitimacy rating conveys adolescents’ beliefs about how frequently each area is legitimate for their caregivers to decide. Given that legitimacy taps into the definition of intrusiveness, the legitimacy rating can also determine which areas to include for computing caregivers’ intrusiveness with the frequency rating. The legitimacy rating may especially help if answers for the belief rating in another culture are not straightforward, unlike our results from Korean middle school adolescents in which the same areas had low factor loadings across all three ratings. The legitimacy rating may also reveal cultural differences between collectivistic and individualistic cultures in future research.
Emotional Manipulation
When Korean middle school adolescents felt pressure or anxiety to obey their primary caregivers and also felt that their caregivers instilled guilt, shamed, invalidated their feelings, attacked personally, or withheld love from them, they did not think well of themselves and reported depression and anxiety symptoms and frequent drinking in the past 12 months. It is striking that regardless of which areas primary caregivers decided instead of the adolescents (intrusiveness), the way that these Korean middle school adolescents felt controlled (emotional manipulation) was associated with poor adjustment. Moreover, these results are consistent with the argument that emotional manipulation may impair children’s development in any culture (Choe et al., 2023). Emotional manipulation tactics might constitute emotional abuse by sending the message to the child that they are worthless, unwanted, and of value only for meeting others’ needs (Choe et al., 2023). Studying emotional manipulation in diverse cultures will help illuminate culture-specific and universal effects of emotional manipulation on children’s development and adjustment.
The emotional manipulation subscale shows similar associations as the established PPC scales do with other parenting and adjustment scales. The distribution of emotional manipulation is also similar to the distributions of the established PPC scales: positively skewed. The results support our argument that the established PPC scales may capture emotional manipulation rather than intrusiveness per se. This suggests that some adverse outcomes attributed to PPC might result from emotional manipulation rather than intrusiveness. These results call for attention to interpret previous results and a nuanced approach to understanding the role of emotional manipulation versus intrusiveness in parent-child relationships and children’s development.
Limitations and Future Directions
This study has some weaknesses that future studies can address. The study lacks caregivers’ answers due to the difficulties of engaging caregivers in Korea, and caregivers’ perspectives on the belief, frequency, and legitimacy ratings would enrich the PPC literature. Lack of secondary caregivers’ answers is another weakness. Many of the secondary caregivers could have been fathers, and the potentially unique role of fathers’ psychological control has been reported (Choe et al., 2022). Future studies with secondary caregivers’ answers can investigate whether fathers’ intrusiveness and emotional manipulation work differently from mothers’ intrusiveness and emotional manipulation. A lack of information on the caregivers or adolescents who declined to participate in the study (due to the complexity of international data collection) make it difficult to discern any sampling bias. We were not able to test all three ratings of intrusiveness and emotional manipulation as four factors, as it requires 820 subjects (25 intrusiveness areas * 3 ratings + 7 emotional manipulation items = 82 items, 10 subjects per item is a rule of thumb for data analyses). Larger samples in future research will afford such analyses. Finally, cross-sectional data make it difficult to conclude temporal ordering between parenting and child outcomes. Longitudinal data can inform how intrusiveness and emotional manipulation change over time and how they predict distal child outcomes.
Conclusion
We tested the Two Facet Model of Parental Psychological Control by separately measuring intrusiveness and emotional manipulation with a Korean middle school adolescent sample. We provide a way to measure what children (and caregivers) in a specific culture believe to be intrusive, hence allowing us to test if or how each facet of PPC, operationalized by what people in a given culture believe, predicts children’s development and adjustment. Our results illuminate the utility of measuring intrusiveness with the three intrusiveness ratings: beliefs, legitimacy, and frequency. The belief rating shows which areas the child believes should be up to them decide, the frequency rating indicates caregivers’ intrusiveness in those areas, and the legitimacy rating addresses possible cultural, familial, and individual differences in what children in collectivistic and individualistic cultures believe about caregivers’ legitimacy. Our results also show that intrusiveness functions differently from emotional manipulation, while suggesting that the established PPC scales might have captured emotional manipulation rather than intrusiveness. This is an important distinction because intrusiveness may predict child outcomes differently from emotional manipulation, but measuring intrusiveness and emotional manipulation simultaneously makes it not possible to test such respective impacts of intrusiveness and emotional manipulation. Our findings suggest that intrusiveness and emotional manipulation are indeed two facets of PPC, empirically buttressing the Two Facet Model of Parental Psychological Control (Choe et al., 2023).
Supplemental Material
sj-docx-1-asm-10.1177_10731911261440140 – Supplemental material for The Two Facet Model of Parental Psychological Control: Separately Measuring Intrusiveness and Emotional Manipulation Based on What Korean Middle School Adolescents in Seoul Believe to Be Intrusive Parenting
Supplemental material, sj-docx-1-asm-10.1177_10731911261440140 for The Two Facet Model of Parental Psychological Control: Separately Measuring Intrusiveness and Emotional Manipulation Based on What Korean Middle School Adolescents in Seoul Believe to Be Intrusive Parenting by So Young Choe, Sarah J. Schoppe-Sullivan, Qing Zhou and Kyoung-Uk Lee in Assessment
Footnotes
Acknowledgements
We thank Gustavo Carlo for his contribution to the design of the study and discussion of the results. We also thank Judith Smetana for her aid in conceptualizing intrusiveness ratings and her consultation on international data collection. Moreover, we thank Joel W. Grube for his contribution to the data analyses and writing suggestions. Finally, we thank the adolescent participants in South Korea, their caregivers who provided the permission to participate, the teachers who distributed the forms and who were present in the classrooms during the study, and the vice principal and principal of the middle school who allowed us to collect data from their middle school in South Korea.
Authors’ Note
Most of the work of the first author was conducted at UC Berkeley, although she has moved to the University of Washington.
Ethical Considerations
The institutional review board (IRB) of a university in South Korea approved the study and procedures.
Consent to Participate
Homeroom teachers in a middle school in Seoul, South Korea distributed the forms for the caregivers’ permission and adolescent child’s assent to all of the students in three grades. The students whose caregivers consented and themselves assented (“participants”) provided the signed forms to their homeroom teachers and participated in the study.
Author Contributions
So Young Choe conceived the ideas, designed the study, wrote the IRB application, analyzed the data, and wrote the manuscript. Sarah J. Schoppe-Sullivan contributed to the design of the study, discussion of results, and writing. Qing Zhou contributed to the Asian parenting literature and writing. Kyoung-Uk Lee worked on the IRB process in South Korea, ensured safety of the adolescent participants in South Korea, and contributed to writing.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Analyzing the data and writing the manuscript by Dr. Choe were supported by Grant T32AA014125 from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) of the National Institutes of Health (NIH). Writing the manuscript by Dr. Choe was also supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health (NICHD) under Award Number K99HD115797. The content of this manuscript is solely the responsibility of the authors and does not necessarily reflect the views of NIAAA or NICHD or NIH.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The participants of this study did not give written consent for their data to be shared publicly, so due to the sensitive nature of the research, supporting data are not available.
Preregistration
This study was not preregistered.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
