Parenting goals and perceived shared and non-shared agency among kirogi mothers of a youth

Abstract This study examined the parenting goals and practices of Korean transnational mothers (referred to as ‘kirogi’) in order to understand how parents are preparing youth for success within a world of increasing globalization. Data are based on a cross-sectional convenience sample of Korean mothers (M age = 45.88) accompanying their youth (M age = 15.72) in the US and in NZ (N = 153). Results indicated that mothers rated youth’s educational/occupational self-development and humanistic life goals to be as important as educational/occupational achievement. Self-development and humanistic life goals were also found to be the most correlated to parenting strategies that accommodate youth, which, in turn, was the most predictive of outcomes. Discussion focuses on how globalization allows for parents with resources to find opportunities outside of their home country to achieve their parenting goals and to allow their youth to self-regulate their schoolwork successfully.


Parenting goals
Goals guide behavior and can be thought of as 'organizers' of action because it allows for plans for goal attainment to be developed. Within the context of parenting, parenting goals may be thought of as the outcomes parents aim for while interacting with their child. The research on parenting goals has been primarily conducted among young children and toddlers and as such, parenting goals have been defined to be what the parent hopes to accomplish in a particular child-rearing situation (Dix, 1992;Grusec, Rudy, & Martini, 1997). Parenting goals have been found to differ in focus (child-, parent-, or relationship-centered; Coplan, Hastings, Lagacé-Séguin, & Moulton, 2002) and in time orientation (short vs. long term; Kuczynski, 1984). Child-centered goals, for example, organize parents' interactions around the needs of children and their self-development whereas parent-centered goals organize interactions around parents' needs.
Parenting goals for adolescents may be qualitatively different in focus and time orientation compared to goals for young children. Older and more experienced parents have had more time to know their child and due to significant developmental shifts in the parent-child relationship, their parenting goals may be qualitatively different. for example, parents of youth may be more focused on fostering adolescent identity achievement, particularly due to choices youth will have in selecting institutions of postsecondary education, deciding upon a career path and selecting a mate. Research based on parents of youth (ages of [14][15][16][17] found that the highest importance ratings were for relationship-centered and selfdevelopment goals (Horvath & Lee, 2015). Thus, greater mutuality within the parent-child relationship can refocus parenting efforts towards recognizing the unique aspects of their child as they prepare to launch youth while at the same time attempting to maintain a positive relationship with them.
for the current study, we defined parental goals for youth as an expression of what parents want when they consider their youth's future. Thus, to ask parents what their goals are for their children is to ask parents to consider many aspects of their child, including the child's point of view, their abilities and their own perceptions of their child's abilities. We expected that parents would have a variety of life goals for their youth, particularly because kirogi parents tend to be future oriented. Previous research has found that youth's self-generated life goals reflect the major developmental tasks of adolescence and young adulthood, such as finishing their education, getting a job and starting a family (Chang, Chen, Greenberger, Dooley, & Heckhausen,2006;Nurmi, 1991). The current study examines the importance of these various life goals for youth among kirogi parents. It is likely that their unique family arrangement reflects importance of familistic goals as well as the importance of achievement in education and/or occupation.

Shared agency with youth for educational goals
Given that one must pursue goals in order to achieve them, strategies of goal pursuit or goal regulation can explain the link between goals and subjective well-being (Headey, 2008). The extant literature on parenting goals and parenting behaviors has indicated positive correlations between the two constructs across different studies over time. Hastings and Grusec (1998) found that parents who reported higher levels of concern over compliance or parents' own needs were found to be more punitive in their responses to vignettes of disagreements with a child. On the other hand, parents with a higher concern for socialization goals and their relationship quality with children were found to use more reasoning with the child. In addition, another line of research has found that parents of toddlers who self-generated long-term educational goals for their child devoted more of their conversations to academic topics than to pragmatic topics (Rowe & Casillas, 2011); and still another line of research has confirmed that parents of children with Type 1 Diabetes who felt that diabetes-specific parenting goals were important were more involved in helping their child manage diabetes (Robinson et al., 2011). Unfortunately, these and other studies of child-rearing goals and parenting behaviors have not made use of models of behavioral regulation resulting in very little understanding about which parenting goals take priority or how parenting processes can regulate parenting goals and therefore parenting outcomes (see Dix & Branca, 2003; for a review).Moreover, although there is evidence of socialization from parent to child, many ignore the fact that parenting behaviors may be elicited from the child or mutually reinforcing within their relationship (e.g. Bell, 1968;Briley, Harden, & Tucker-Drob, 2015;Zhang, Haddad, Torres, & Chen, 2011).
Goals are certainly not decided upon and pursued within a social vacuum although some goals may be more or less in need of assistance from others. Subsequently, there is growing interest in how goals may be co-regulated with social partners. In pediatric settings, parents' close monitoring and involvement in managing children's glucose levels has been described as a form of dyadic coping between parent and child (Berg, Meegan, & Deviney, 1998). Parental management of blood glucose may be necessary for young children due to their inability to adhere to treatment. However, research has indicated that during adolescence, parents who say that they collaborate with youth on their glucose management adhere better to treatments and doctor's orders (Wiebe et al., 2005). The advantage of parental collaboration with youth in health settings is that it promotes youth autonomy as well as preserves a high level of parental monitoring in service of developing youth's health-preserving behaviors.
The importance of social partners in the developmental regulation of life goals is underscored by research conducted by Young and colleagues, who found through a series of interviews of parent and youth dyads over time that youth's career development is best thought of as a project that is carried out jointly by family members with shared goals (Young et al., 2001). Since kirogi youth's educational goals can be considered a family project, parents may be more motivated to ensure that their youth's educational activities are to their satisfaction. Thus, we refer to a model of shared and non-shared agency that considers the goal engagement of the parent and the youth at the same time (Chang, 2013;Chang et al., 2010). This model was applied in order to understand parenting behaviors more accurately -as they occur within the parent-child relationship, which during adolescence is more mutual and bidirectional than in childhood.
The model of shared and non-shared agency consists of five strategies of educational goal pursuit, which vary depending upon whether each person is goal engaged or goal disengaged in youth's education. When both youth and parent are goal engaged, there is a sense of shared agency, which can differ uniquely in three particular ways. In the example of the mother-youth relationship, mothers can be collaborative, supportive or accommodating to youth (Chang et al., 2010). When parents feel as if they are collaborators with youth, they are more likely to discuss alternatives and negotiate terms with youth in order that they stay on track (Wiebe et al., 2005). Parents who collaborate with youth may feel more strongly about the 'right way' to do things and insist more strongly than parents who are willing to be supporters for youth autonomy. Parents who support youth autonomy are involved in schooling but more willing to allow youth to choose his/her own way of doing things as they encourage their child to succeed and stay the course (Steinberg, Lamborn, Dornbusch, & Darling, 1992). In contrast, when parents accommodate to youth, they are willing to let go of their own goals for youth but remain interested in youth's goal progress. In addition, two strategies of non-shared agency have also been identified. When the parent is not goal engaged, s/he is uninvolved in youth's educational pursuits and leaves it up to youth to accomplish his/her educational goals on their own. On the other hand, parents may be over involved relative to their youth because youth may be either unable to manage goals or unmotivated in general with educational pursuits. In this case, parents may find themselves to be directing their youth by setting study times or constantly reminding their youth to study.
The adaptiveness of each respective strategy is expected to differ across cultures and across the child's lifespan. In particular, cultural differences may be present in the acceptability of parental authority and the adaptiveness of parental directing given Confucian-influenced values and beliefs (Luo et al., 2013). However, it is hypothesized that maternal collaboration would be the most effective role for parents within the kirogi family context due to the investment of parents in children's education and the necessity of youth to have and regulate their own educational goals (Chang, 2013;Wigfield & Eccles, 2000).

The current study
This study proposes to help inform at least two important gaps in understanding parenting as a goaldirected activity. One is in exploring and describing what parents want for the development of their youth beyond specific situations and long-term socialization goals, and the other is in hypothesizing how they may best work towards achieving their goals with their children.
The first aim was to understand which life goals were important for kirogi mothers of a youth. Due to mothers' investment in transnational living arrangements and the competitive nature of the Korean educational system, it was expected that kirogi mothers would deem goals related to educational and occupational achievement to be very important goals for youth (Chew, 2009;finch & Kim, 2012;Lee & Koo, 2006). Reflective of their Confucian-influenced cultural background, kirogi mothers were also expected to rate familistic goals and traditional mate selection goals as important (Luo et al., 2013).
The second aim of this research was to investigate the relationship between kirogi mothers' life goals and each strategy of shared and non-shared agency with youth. It was hypothesized according to theory (Chang et al., 2010), that mothers who rated educational/occupational achievement and self-development life goals as more important were more likely to endorse statements indicating shared agency with youth (i.e. collaboration, support, accommodation). We also hypothesized two different directions of associations for each strategy of non-shared agency. first, it was expected that the more important educational and occupational goals were to mothers, relating both to achievement and youth self-development, the less likely mothers would report maternal noninvolvement. Second, the opposite would be expected with regard to maternal directing of youth's educational goals. It was hypothesized that mothers who rated educational and occupational goals relating both to achievement and youth self-development as more important would be more likely to report that they directed youth's educational activities. Since there is very little knowledge about how educational goal pursuits for youth correlate to the setting of other life goals, such as family obligations, expectations for future family formation, self-development and humanistic goals, there was no basis for further hypothesis development.
finally, the third aim of this study was to test for the unique significance of perceived maternal collaboration. Since parenting effectiveness depends on selecting goals and strategies that successfully elicit their child's motivation and skills to enable plans to be carried out, it was proposed that collaboration, in particular, would be positively and uniquely correlated to indicators of mothers' perception of youth's educational goal progress and satisfaction with their relationship with youth (e.g. Chang, 2013). It is argued that the kirogi family context would foster collaboration due to the stakes that both parents and youth have in youth's education.

Methods
At the time of survey, participants were Korean women who were married, and living in New Zealand or in the US for the purposes of their children's education while their husbands remained in Korea. The sample was further restricted to mothers who had an adolescent child between the ages of 12 and 18 years. Participants were recruited through snowballing and online postings on Korean community websites from the beginning of 2012 to the end of 2013. Postings described the purpose of the study and directed those interested to a link to an online survey to complete the questionnaire. The advertisement also explained that if respondents wished to have a paper copy of the questionnaire, they should contact the researcher and request one. A Korean version of the questionnaire was prepared by a bilingual and bicultural research assistant. The questionnaire package was accompanied by a survey questionnaire and a stamped self-addressed envelope for the return of their completed survey. If there was more than one adolescent in the specified age range (ages 11-19) in the family, mothers were asked to pick one and keep that person in mind for all of the survey questions relating to their youth. Mothers were also instructed to provide the name, age, grade of their target adolescent and the grade the child started schooling overseas. Mothers were compensated with a token amount of cash for their time. This study's procedures were approved by the ethics committee of the first author's institution.
The final sample of mothers (N = 153) were at midlife (M age = 45.88 years; SD = 4.62 years), married for approximately 20 years (M = 19.28; SD = 3.95), reported 1.78 children on average (SD = .73) and tended to be highly educated (e.g. 57.5% of mothers completed a 4-year degree & 11.8% of mothers completed a graduate degree). The average age of target youth was 15.72 years (SD = 1.90 years) and the median years of children's schooling in the US/NZ was 4 years. The proportion of each gender among target youth was almost equal: 51.7% of mothers reported on a son (n = 78) and 48.3% of mothers reported on a daughter (n = 73).
A t-test was conducted on demographic variables using country of current residence (US or NZ) as the grouping variable to examine differences between the two subsamples on six demographic variables and nine study variables. Results indicated that only four significant mean-level differences were found. Mothers residing in the US were slightly more likely to report higher educational attainment than mothers residing in NZ: Ms = 4.90 (SD = .78) & 4.44 (SD = .92), respectively; t (145) = 3.31, p = .001). Mothers residing in the US also had slightly more biological children on average (M = 2.00; SD = .81) than their counterparts in NZ (M = 1.56; SD = .55, t (151) = 3.93, p < .001). Additionally, mother-reported grades were much higher in the USA sample than in the NZ sample (Ms = 4.59 and 3.53, respectively; t (151) = 6.36, p < .001) and maternal directing was slightly higher in the USA sample than in the NZ sample (Ms = 2.78 and 2.61, respectively; t (147) = 2.05, p < .05). In all multivariate analyses, we added a control for country of residence to control for the country differences found even though the frequency of differences was minor.

Demographics
Mothers indicated their age, country of birth, ethnic background, marital status, number of biological and/or stepchildren, their target child's gender and the highest level of education they attained (1 = junior high, 2 = some high school, 3 = high school graduate, 4 = some college, 5 = BA or BS, 6 = MA or other professional degree).

Parenting goals
A 26-item inventory of life goals, which was developed for self-report among older youth (LGI: Chang, Chen, & Kim, 2015 ), was adapted for use among parents' life goals for their youth. Mothers rated life goals for their target youth in importance on a four-point scale (1 = not at all important; 4 = very important). Goals were diverse in life domain and also varied in focus for educational and occupational goals. for example, educational and occupational goals that were focused on achievement (6 items, α = .74; 'getting all A's' , 'earning a lot of money') were separated from goals focused on self-development in education and occupation (3 items, α = .72; 'studying an interesting subject').Beyond educational and occupational domains, other life domains were included such as relational goals that emphasized the strengthening of familistic ties (7 items, α = .86; 'being a good son/daughter'), humanistic goals that emphasized social welfare (3 items, α = .78; 'helping the unfortunate') and mate selection goals that were traditional in nature (7 items, α = .78; 'marrying someone who is financially stable'). See figure 2 for item wording and full inventory of goals in each category.

Shared and non-shared agency for educational goals with youth
Mothers were asked to think about their target youth and asked to indicate the extent to which they agree or disagree with statements measuring different aspects of shared and non-shared agency with him/her for educational goals (SNSA; Chang et al., 2010; 1 = Strongly disagree; 4 = Strongly agree). Since the original measure was established among older youth in college and instructed them to think about their parents, the measure was adapted for use among mothers and for high school youth for the purpose of this study.
Three subscales of shared agency consisted of maternal accommodation (3 items; e.g. 'I tend to follow my child's lead when it comes to his/her education'), support (5 items; e.g. 'I am very supportive of how my child manages his/her school activities') and collaboration with youth (5 items; e.g. 'When my child falls short of his/her educational goals, I give him/her the confidence to try harder'). In addition, maternal noninvolvement (6 items; e.g. 'I am not responsible for helping my child achieve his/her educational goals') and maternal directing (5 items; e.g. 'I need to know the details of what my child does during the day at school') of youth's educational goals to assess perceived non-shared agency with youth. All subscales were reliable (αs ranged from .67 maternal directing to .80 maternal noninvolvement).

Educational goal progress
Two indicators measured mothers' perceptions of educational goals progress. first, mothers were asked to indicate, to the best of their knowledge, the target child's average school grades on a six-point scale (1 = Mostly C's; 6 = All A's). Mothers also reported on the extent to which they felt satisfaction with their youth's schooling experiences on a four-point scale (1 = Not at all satisfied; 4 = Very satisfied; Chang et al., 2010). Items involved assessing satisfaction with their child's progress in English language learning, the amount of time youth invested in studying and with youth's school grades.

Parent-child relationship quality
Both negative and positive aspects of relationship quality were measured in order to tap into the quality of relationship mothers perceived to have with their youth.
Conflict with youth was measured by 11 items that tapped into typical topics of conflict between parents and youth on a four-point scale (Greenberger & Chen, 1996; 1 = Rarely, or none of the time; 4 = Most, or all of the time). Sample items are conflict over appearance, schoolwork, friends, dating and substance use (α = .89).
Satisfaction with their relationship with youth was measured by the extent to which mothers endorsed the statements 'In almost all aspects, the relationship between my child and I is close to ideal' and 'I am satisfied with my relationship with my child' on a six-point scale (1 = Strongly disagree; 6 = Strongly agree; α = .79).

Kirogi mothers' parenting goals
The first research question explored which goals would be the most important among kirogi mothers. On a four-point scale, a score above three indicated a goal that was at least 'Somewhat important' . As expected and as can be seen by figure 1, goals related to educational and occupational achievement were rated to be, on average, above 3 (M = 3.16; SD = .44). However, goals related to youth's self development in education and occupation were, on average, highest in importance (M = 3.61; SD = .48) and surprisingly, humanistic goals were also reported to be the among the important life goals for their youth (M = 3.33; SD = .59).
Life goals that were rated to be on average below the three mark were traditional mate selection goals and familistic goals (Ms = 2.86 & 2.79; SDs = .54 & .56, respectively), indicating that these were goals that were 'Somewhat not important' . We conducted one sample t-tests to confirm that means that were above 3 were significantly above 3 and suggestive of 'important' life goals whereas means that were lower than 3 were significantly below 3 and suggestive of 'unimportant' life goals. Results indicated that all t-tests were significant (all ps < .001). Thus, self-development in education/occupation, humanistic and achievement in education/occupation were important for their youth whereas traditional mate selection and familistic goals for their youth were not important to them. figure 2 displays the means of each life goal in order from the most important to the least important for further detailed analysis. The first and second most important life goal was indicative of youth's self-development in

Parenting goals and parenting strategies
The second research question was to examine the associations between parenting goals and strategies of shared and non-shared goal pursuit. Correlations were conducted to test whether mothers who rated goals related to educational/occupational achievement and self-development as more important would also be more likely to endorse strategies of shared agency with youth (i.e. accommodation, support, and collaboration) as well as maternal directing. It was also hypothesized that goals related to educational/occupational achievement and self-development would be negatively correlated with maternal noninvolvement. Results of correlational analyses are summarized in Table 1. As hypothesized, mothers who rated educational/occupational achievement goals as important were also significantly more likely to endorse statements relating to maternal directing of youth, as well as support for and collaboration with youth. Additionally, mothers who placed lower importance on achievement in education/occupation were also significantly less likely to endorse statements indicating maternal noninvolvement in youth's educational goals, as was expected. Contrary to hypotheses, importance ratings of educational/occupational achievement were not significantly correlated to maternal endorsements of statements accommodating to youth. Instead, maternal accommodation was correlated significantly with importance ratings related to youth's educational and occupational self-development. No other strategies of shared and non-shared agency were significantly correlated to educational/occupational self-development, which was contrary to hypotheses.
The importance of other life goals for youth was also correlated significantly with mothers' perceived agency with youth for educational goals. As can be seen on Table 1, the more important mothers rated humanistic goals, the more accommodating they were to youth in education. The more important traditional mate selection goals were to mothers, the more mothers were inclined to manage youth's educational goals. finally, the more that familistic goals were rated to be important to mothers, the more likely they were to report that they managed or collaborated with their youth on educational goals.

Correlates of shared and non-shared agency
Prior to reporting upon the correlates of shared and non-shared agency with outcomes, it is noteworthy to report upon the inter correlations between each strategy within the model. As expected, positive and significant correlations were found between strategies of shared agency (see Table 2). Correlation coefficients ranged from .24 between accommodation and collaboration to .55 support and collaboration. Also, and as expected, strategies of non-shared agency were negative and significant, though modest in size (r = −.17 between directing and noninvolvement). However, it is interesting to note that support and collaboration were both positively and significantly correlated to directing (.22 and .36, respectively); and noninvolvement was differentially associated with each strategy of shared agency. Noninvolvement was positively associated with accommodation, negatively associated with support and unrelated to collaboration (rs = .24, −.25 and −.06, respectively; see Table 2 for significance levels). In order to assess the potential effectiveness of shared agency and the ineffectiveness of non-shared agency, we relied on correlational and linear regression analyses. first, we correlated each strategy of shared and non-shared agency with two indicators of maternal perceptions of educational goal progress (i.e. youth's school grades and maternal satisfaction with youth's educational experiences) and with two indicators of mother-youth relationship quality (i.e. P-C conflict and P-C relationship satisfaction). As can be seen on Table 2, results indicated partial confirmation of hypotheses. The strongest support for hypotheses relating to the adaptiveness of shared agency among kirogi mothers was indicated by the positive correlations between strategies of shared agency with outcomes relating to maternal satisfaction (with either youth's educational experiences or maternal satisfaction with relationship quality with youth; rs ranged from .21 to .47). The strongest support for hypotheses relating to the maladaptiveness of non-shared agency with youth among kirogi mothers was indicated by one outcome, namely conflict with youth. Perceived noninvolvement in youth' educational goals was modestly correlated (r = .19, p < .05) and maternal directing was positively associated (r = .26, p < .01) with maternal report of conflict with youth. Both strategies of non-shared agency were unrelated to 2 out of the 4 outcomes (maternal report of youth's school grades and relationship satisfaction) and all three strategies of perceived shared agency were unrelated to maternal reports of conflict with youth.

Unique contributions of collaboration vs. directing
finally, we conducted 4 two-step hierarchical linear regression analyses to test for the unique effectiveness of collaboration and maladaptiveness of maternal directing for each outcome variable (third research question). We entered country of residence in Step 1 to control for mean level differences found between subsamples from NZ and the US; and the five strategies of shared and non-shared agency in Step 2. The results are tabled in Table 3. In order to see whether results would change due to significant associations between outcomes and demographic factors, we conducted 3 post hoc hierarchical linear regression analyses. In these regressions, we controlled for country of residence, as in the original set of analyses, and added the demographic factor that was significantly correlated with the corresponding outcome measure such that the demographic factor (child' gender or sibship size) was entered in Step 1 followed by the set of 5 strategies of shared and non-shared agency with youth in education entered in Step 2. for example, we entered child's gender in Step 1 and the five strategies of shared/non-shared Table 3. shared and non-shared agency for educational goals regressed onto indicators of adjustment in education and within parent-child relationship.

Educational goal progress
Mother-youth relationship quality Step 2 for analyses that used satisfaction with educational experiences and P-C conflict as the dependent variable. Since results did not change after controlling for each respective demographic factor, we report on the original set of analyses.
As can be seen on Table 3, strategies of shared and non-shared agency explain significant variance across outcomes of interest, ranging from 7% of the variance in maternal report of youth's school grades to 28% of the variance in maternal satisfaction in youth's educational experiences (see ΔR 2 from Step 1 to Step 2). Results also indicated that contrary to hypothesis, maternal accommodation was uniquely and positively associated (for 3 out of the 4 outcomes of interest) whereas maternal collaboration with youth was not a uniquely significant predictor of outcomes even though it reached trend levels for 2 outcomes. More consistent with hypotheses, we found that maternal directing was shown to be maladaptive. Mothers who reported higher levels of maternal directing were also more likely to report lower levels of satisfaction with their youth's educational progress as well as higher levels of conflict with youth.

Discussion
The current study made use of a unique family situation to test for the effectiveness of shared agency between youth and parents as well as for the ineffectiveness of non-shared agency. Our results partially confirmed that directing was maladaptive and identified maternal accomodation to youth, not collaboration, was the most effective strategy of perceived shared agency with youth. These results study make a few important contributions to the literature on parenting as a goal-regulated activity and academic socialization.
The first set of results reveals why maternal accommodation to youth was important to kirogi mothers. We found that kirogi parents' goals in developing their youth's self were as important as youth achievement for their child. An analysis of the importance of each goal revealed that 3 of the top 5 important life goals for youth were related to self-development, compared to just 1 achievement-related life goal (see figure 2). These findings were contrary to our predictions but consistent with a cross-cultural study that found that East Asians were more likely to endorse socialization goals that are usually assumed to be Western when compared to Westerners themselves (Park, Coello, & Lau, 2014). Although it is likely that mothers may have been influenced by their host culture's values, it is perhaps more likely that the focus on youth's self-development may be attributed to kirogi mothers' high educational attainment. for example, Western and non-Western parents with a high level of educational attainment tend to emphasize children's inner states more than children's social obligations and responsibilities in their parenting (Keller, 2012).
It is interesting to note that the life goals of lowest importance for mothers were familistic, particuarly for goals indicating greater focus on mothers than on youth. This suggests that kirogi mothers did not expect material benefits from their children in the future despite their parental investment. It also appears that values are changing from traditional Confucian traditions that expect adult children to care for parents' needs to modern values reflecting greater affluence and urbanization among highly educated Koreans. If true, this would be consistent with a study finding that expectations of a grown child have changed, namely from their utilitarian value to their psychological value (Kagitcibasi & Ataca, 2005). The lack of focus on the family may also reflect mothers' preoccupation with launching their youth into adulthood given the growing competition as a result of globalization.
It should be noted that there exists very little to no research on parents' humanistic life goals for youth. It was therefore surprising that humanistic life goals for youth were rated by mothers to be as important as achievement and self-development goals in education and occupation for youth and humanistic parenting goals were significantly intercorrelated with maternal accommodation. It should be recognized that this study used a short three-item measure to tap into humanistic goals, which can be expanded in future research for greater detail to include environmental themes and issues related to social justice. The unique perspective of family members who live in two countries in comparison to coresidential, single-nation families would be an important topic to explore in future research.
The second set of results provides evidence that the model of shared and non-shared agency performed according to theory despite the finding that collaboration did not appear to be uniquely significant among kirogi mothers. Since we differentiated between achievement and self-development goals in education/occupation, we found this distinction to have important implications for the type of goal pursuit mothers endorsed with youth.We found that mothers' achievement goals correlated significantly with mothers' perception of shared and non-shared agency and all in the expected directions but not signifcantly correlated to the importance of youth's self development in education/occupation. However, in both of the latter cases, maternal accomodation to youth was the exception.When mothers placed higher importance on youth's self-development, they were more willing to accommodate to how youth preferred to manage their education. Thus, shared agency strategies of support and collaboration were therefore more purposeful in achieving educational and occupational status whereas accommodation was more purposeful in developing youth's self. This differentiation will be important in informing future research since previous research on shared/non-shared agency assumed that parental agency was exclusively achievement-oriented.
The use of accommodation may be developmentally appropriate and effective given that youth are still minors and protected by their parents' full-time care. In other words, youth are able to make mistakes and learn from them with less consequence because mothers are there to guide and help them compared to mistakes in goal regulation after high school and away from full-time parental care. Parents high in accommodation may recognize that youth in middle and high school can be competent in regulating their educational goals, which remain predictable and controllable when compared to goal reguation after high school. Research examining the adaptiveness of maternal accommodation among youth in middle and high school in a more general population of parents may be more necessary given its high frequency during college (Chang et al., 2010). In this sample, maternal accommodation may indicate reasons for coming to the US/NZ for education. Parents may have been seeking social contexts for their youth to master their own educational goals perhaps because their youth was not adjusting well to the Korean educational context (i.e. thus requiring too much maternal directing).
Strategies of non-shared agency functioned as expected in this study and generally did not have unique negative consequences to outcomes (only to mothers feelings of youth's educational progress). However, maternal directing and noninvolvement were both uniquely related to higher reports of conflict with youth, confirming its maladaptiveness for families with a youth. Mothers who were managing youth's educational goals may have been prompted to do so due to the variety of parenting goals at stake. When referring back to correlations between parenting goals and maternal directing, our results indicated that mothers who managed youth's educational goals also felt that achievement, familistic and traditional mate selection goals were important. We believe that it will be important to conduct qualitative investigations to examine differences between mothers of youth in middle/high school who are accommodating versus managing youth's education in order to understand whether their parenting goals and practices changed during their youth's adolescence. It is possible that parenting an adolescent may be a period during which parents begin to let go of their own (selfish) goals.
We should also briefly note that kirogi mothers who were more likely to be uninvolved in youth's educational progress were also more likely to rate educational/occupational achievement goals lower in importance without any other goal domains that were significantly more important. This suggests a lack of investment in their parenting despite living transnationally for the sake of their child's education. Highly uninvolved kirogi mothers may be a vulnerable subpopulation in need of social services as a result. Although the evidence presented makes a case for the importance of parenting goals related to youth's self-development and the potential effectiveness of maternal accommodation in youth's education, it is important to acknolwedge the limitations of this study. Our results were based on the responses of a relatively small convenience sample of mothers, meriting some caution regarding the stability of our estimates (Lindsay, 2015). We also relied on mothers' report to measure motivational aspects of their relationship with youth, which could be positively biased due to the intergenerational stake (Birditt, Hartnett, fingerman, Zarit, & Antonucci, 2015). furthermore, kirogi mothers can be considered a special class of mothers who are distinct in their parenting goals and practices. Despite these limitations, we believe that this study contributes significantly to the study of parenting practices in the age of globalization and to the literature on academic socialization of youth as education becomes a gobal product and parents as their main consumers. Maternal accomodation to youth is an ideal way to share agency with youth and may have been the motivation for transnational living arrangements.

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes on contributors
Esther S. Chang is an Associate Professor of Psychology at Soka University of America. Her main research interest is in how family relationships impact developmental outcomes in the transition to adulthood across cultures.