A meta-analysis of children feeling caught between parents and parent-child relationship quality

Abstract

This meta-analysis reviews the findings from 26 studies (N = 8,061) examining the association between children feeling caught between parents (also referred to as parent-child triangulation) and parent-child relationship quality. The cumulative evidence indicated a small but meaningful inverse relationship between children feeling caught and relationship quality with parents (r = −.218, 95% CI: −.260, −.175). The overall effect was heterogenous. Moderator analyses revealed that the inverse association was stronger in studies using non-U.S. samples (k = 4, r = −.332, 95% CI: −.419, −.239) compared with U.S. samples (k = 22, r = −.197, 95% CI: −.237, −.155). Conversely, the association was not moderated by the divorce status of the parents, the Whiteness of the samples, gender of the children, measures of feeling caught and relationship quality, or by research design. Implications for family systems theory are discussed.

Keywords

feeling caught family systems parent-child triangulation parent-child relationship quality meta-analysis

One of the more challenging family dynamics for children to navigate are the loyalty divides that often emerge when parents become locked in destructive conflict patterns (e.g., Afifi & Schrodt, 2003; Amato & Afifi, 2006; Bowen, 1978; Buchanan et al., 1991; Minuchin, 1974). When children witness their parents’ chronic disputes (Schrodt & Afifi, 2007; Schrodt & LaFreniere, 2022; Schrodt & Ledbetter, 2007) or when parents create unhealthy alliances with their children by sharing inappropriate disclosures (Schrodt & Afifi, 2018a, 2018b), requesting information about the other parent, positioning their children as mediators of their disputes (Afifi, 2003), and undermining each other’s parenting efforts (Schrodt, 2016), children are more likely to feel caught between their parents. Feeling caught refers to a form of parent-child triangulation and boundary diffusion in which children are put in the middle of their parents’ relationship and forced to defend their loyalty to each of their parents (Afifi, 2003; Amato & Afifi, 2006). Being triangulated or drawn into parents’ disputes is consequential for children, as feeling caught is associated with children’s anxiety, depression, poorer self-esteem, and mental health symptoms (Amato & Afifi, 2006; Buchanan et al., 1991; Schrodt & Afifi, 2007).

Feeling caught is also consequential for children’s relationships with their parents, as adolescent and young adult children have reported less relationship satisfaction (Schrodt & O’Mara, 2019; Schrodt & Shimkowski, 2013; Thorson, 2021), less closeness (Afifi et al., 2008; Fosco et al., 2016; Fosco & Grych, 2010), more relational turbulence (Leustek & Theiss, 2020), greater insecurity in the parent-child relationship (O’Hara et al., 2024), and less healthy boundaries when they feel caught (Mayseless & Scharf, 2009; Urcan, 2011). Indeed, feeling caught is widely studied and maintains a key position in current theorizing about family functioning, parenting, and child adjustment, yet there is tremendous variability in the relationship outcomes that researchers have explored as correlates of children feeling caught. One way to synthesize this body of research, to generate an estimate of the overall effect of feeling caught on relationship outcomes, and to advance an understanding of how feeling caught undermines relationship quality with parents is to conduct a meta-analysis. For the purposes of this meta-analysis, parent-child relationship quality refers to the essential characteristics or attributes of parent-child relationships that distinguish them as excellent or superior. High quality parent-child relationships include those that are trustworthy, secure, satisfying, appropriately close, and have appropriate psychological boundaries that facilitate healthy childhood development. Consequently, this meta-analysis examines the association between children feeling caught and parent-child relationship quality in 26 studies that were published between 1985 and 2024.

This meta-analysis is the third report to come from a larger sample of studies analyzing children’s feelings of being caught between parents. In the first meta-analysis, I reported a positive and robust overall effect size estimate between interparental conflict and children feeling caught (Schrodt, 2025a). In the second meta-analysis, I shifted the focus away from interparental conflict and toward the parent-child relationship by summarizing the unique effects that antisocial and prosocial parent-child communication may have on feeling caught (Schrodt, 2025b). In what follows, I provide an overview of the primary theory researchers have used to investigate feeling caught and relationship quality. I then review potential moderators of the overall association before reporting a meta-analysis on this body of work.

Theoretical perspective

Most researchers investigating triangulation (in general) and feeling caught (specifically) owe intellectual debts to two related, but distinct family systems theories (FST) (i.e., Bowen, 1976, 1978; Minuchin, 1974). First, according to Bowen’s (1976, 1978) FST, the triangle is the fundamental building block of any system, and tension within a twosome often leads the most uncomfortable partner to recruit a third family member to help dispel the tension. Bowen (1978) suggested that emotional interdependence between family members is a key mechanism through which family members spread anxiety one to another. In support of his claim, researchers have documented the spillover effect (Cox et al., 2001) in which adolescents’ presence during interparental conflicts makes them convenient targets for their parents’ displaced hostility and aggression, ultimately undermining their emotional security (Davies & Martin, 2013; Fosco et al., 2014; Fosco & Grych, 2010). Bowen’s (1976, 1978) FST is emotionally and intergenerationally focused, centering on constructs like differentiation of self, emotional reactivity, and triangulation across multiple generations. Thus, from Bowen’s (1976, 1978) perspective, feeling caught emerges as a function of an emotional interdependence that exists between parents and children, one that in the wake of destructive interparental conflict, prompts feelings of stress and anxiety that may undermine children’s relationships with one or both of their parents.

Second, Minuchin’s (1974) FST focuses primarily on structural family therapy and emphasizes family organization and boundaries. According to Minuchin (1974), families operate through interaction patterns that establish when, how, and to whom to relate to – that is, through communication, family members establish and maintain boundaries that are critical for healthy family functioning and childhood development (see also, Kerig, 2005). Each family system is comprised of unique subsystems that influence, and are influenced by, other subsystems (i.e., the tenet of complex relationships), such that interactions occurring within one subsystem (e.g., between parents) affect interactions in others (e.g., between the child and each parent). Although family systems strive to maintain a sense of balance or equilibrium among the subsystems comprising them, they inevitably encounter moments of system disequilibrium (e.g., interparental and parent-child conflict) in which family members feel that other members are not fulfilling their obligations, at which point calls for family loyalty and guilt-inducing maneuvers begin to appear (Minuchin, 1974). As a result, some parents may turn on each other and seek to create unhealthy alliances with their (typically older) children that blur and diffuse healthy boundaries among interparental and parent-child subsystems. From Minuchin’s (1974) FST, it is parents’ attempts to marshal inappropriate social support from their children in response to disequilibrium in the interparental relationship that diffuses healthy boundaries within the family system and induces feelings of being caught.

Taken together, both FSTs suggest that once children feel caught between their parents, relationship quality with both parents may suffer, whether due to the stress and anxiety that comes from witnessing their parents’ fights and being an involuntary target of displaced hostilities and aggression (Bowen, 1978), or to the boundary diffusion and elevation of the child to an inappropriate role within the family system (e.g., peer, confidante, fellow parent, etc., Minuchin, 1974). In other words, feeling caught should be inversely associated with relationship quality in the parent-child relationship because children find stressful, anxiety-inducing, insecure, and enmeshed family relationships to be less satisfying, less trustworthy, more turbulent, and more uncomfortable. Given both theoretic logic and empirical evidence to suggest that the overall association between children feeling caught between parents and parent-child relationship quality should be negative, I advanced my first hypothesis:

H₁: Children feeling caught between their parents is negatively associated with children’s reports of relationship quality with parents.

Moderators of the children feeling caught – relationship quality association

In addition to generating an overall effect size estimate for feeling caught and parent-child relationship quality, I also explored potential moderators of the overall association. First, according to FST (Minuchin, 1974), family structure (or organization) may moderate the association due to the diffusion and expansion of subsystems and boundary patterns within the larger family system via divorce and remarriage. For instance, children from intact families are less likely to feel caught between parents than children from post-divorce families (Afifi & Schrodt, 2003; Schrodt & Ledbetter, 2007; Schrodt & Shimkowski, 2013), unless the intact families include high-conflict marriages, in which case children from intact families are more likely to feel caught than those whose parents are divorced (Amato & Afifi, 2006). Beyond mean-level differences in feeling caught between the family types, feeling caught may be more predictive of diminished relationship quality with parents from intact families because children are unable to escape their parents’ marital problems. Whereas post-divorce families often include remarried adults and other stepfamily members living in multiple households, which in turn may provide additional sources of social support and respite from being triangulated into their parents’ disputes, intact families offer no such forms of escape. Thus, to test family structure as a potential moderator of feeling caught and relationship quality, I advanced a second hypothesis:

H₂: The inverse association between children feeling caught and parent-child relationship quality is stronger for studies that sampled primarily (or exclusively) intact families than for studies that sampled exclusively post-divorce families.

Second, the association between feeling caught and relationship quality may also depend on the age of the children. As children age and develop greater information-processing and communication skills, they may become more sensitive to their parents’ disputes and drawn into their disputes as parents build unhealthy coalitions within the family. As Afifi (2003) discovered, (ex)spouses will sometimes turn to the oldest child during ongoing conflicts and disclose negative information about each other in an effort to marshal social support. Likewise, Buchanan et al. (1991) reported that older adolescents are more likely to feel caught between parents than younger adolescents. As adolescents grow into emerging adults and leave the home, the inverse association between feeling caught and parent-child relationship quality may grow in magnitude given continued pressure to remain involved in their parents’ disputes and relationship despite increased independence from the family. Therefore, I advanced a third hypothesis to test age group as a potential moderator:

H₃: The inverse association between children feeling caught and parent-child relationship quality is stronger for studies that sampled young adult children than for studies that sampled adolescent children.

Other sample characteristics may also moderate the association, such as gender and race, although evidence concerning the former is mixed and concerning the latter, non-existent. In terms of child gender, some studies have suggested that boys are more likely than girls to be caught in parents’ conflict (Johnston et al., 1989), whereas other studies have reported that girls experience more feelings of being caught than boys (Buchanan et al., 1991). Still others have reported no significant differences in feeling caught based on child gender (Afifi & Schrodt, 2003). Thus, in the present meta-analysis, I explored the gender composition (via the percentage of children in each sample who were female) and racial composition of the samples (via the percentage of participants who were White) as potential moderators of children feeling caught and parent-child relationship quality.

Finally, I explored different characteristics of research designs as potential moderators of the association. In addition to several measures of feeling caught,¹ for instance, I considered whether the association was stronger for cross-sectional as opposed to longitudinal research, for U.S. samples as opposed to non-U.S. samples, and for published as opposed to unpublished reports.

Method

Sample of studies

This is the third meta-analysis to come from a larger sample of studies analyzing children’s feelings of being caught between parents (Schrodt, 2025a, 2025b). The sample was created from an initial search for published work using the terms “feeling caught,” “loyalty conflict,” “child AND triangulation,” “parent AND triangulation,” and “family AND triangulation” in the following databases: Academic Search Complete, Communication Source, PsychArticles, PsycInfo, and the Social Sciences Citation Index. After searching and cross-referencing relevant citations in the obtained reports, I collected unpublished scholarship (i.e., grey literature) using the same keyword search in ProQuest Dissertations and Theses and Education Resources Information Center to obtain doctoral dissertations, master’s theses, and undergraduate theses. I also searched online convention programs from the International Communication Association, the National Communication Association (NCA), the International Association for Relationship Research, and regional communication associations. Finally, I called for unpublished manuscripts and data via NCA’s COMMNotes listserv. After screening more than 4,700 records published (or reported) between 1977 and August of 2024, I identified 171 manuscripts (150 published, 21 unpublished) that reported some form of data on feeling caught for possible inclusion in the analysis.

Selection criteria

Consistent with standard meta-analytical procedures (Borenstein et al., 2021), I considered a study a potential source of data in this analysis if (a) the predictor variable included assessments of children feeling caught between their biological parents, (b) the criterion variable included at least one affective indicator of children’s relationship quality with their parent(s) (e.g., closeness, satisfaction) or one psychological boundary indicator (e.g., parent-child insecurity) (see Online Supplemental Table 1), and (c) the study reported statistics necessary for effect size calculations.

Using these criteria, I carefully examined each of the identified studies to determine whether results had been reported in a manner that enabled meta-analytic comparison. I excluded studies if they: (a) were non-empirical (k = 9), (b) focused solely on the measurement of feeling caught or parent-child triangulation (k = 5), (c) reported findings in a manner that prevented the calculation of an effect size (k = 25), (d) reported thematic analyses using qualitative data (k = 14), (e) did not measure children feeling caught or a similar dimension of parent-child triangulation based on an inspection of measurement items (k = 16), or (f) examined feeling caught in a different relationship context (k = 3).² This review process produced the present and total sample of 26 primary empirical studies representing 8,061 individuals from 7,879 families published between 1985 and 2024.³

Table 1 lists the primary studies retained in the meta-analysis. These studies represented a combination of cross-sectional and longitudinal designs that assessed children feeling caught and parent-child relationship quality in both intact and post-divorce families from the United States, Canada, Bangladesh, Israel, and Germany. Across samples, the average age of the children was 18.9 years (range = 10.3–38.3 years), 61.7% of them were female, and 69.9% of all individual family members were White.

Table 1.

Descriptors and effects for studies included in the feeling caught and relationship quality meta-analysis (k = 26, N = 8,061).

Authors	Year	N	Family type	M Age of child	% Female	%White	Country	Research design	Unit of Analysis	Overall r
1. Afifi et al.	2008	224	Both	16.0	45.0	91.0	U.S.	C-S	Dyad	−.295
2. Afifi & Schrodt	2003	601	Both	18.7	64.6	93.7	U.S.	C-S	IND	−.290
3. Amato & Afifi	2006	632	Both	26.9	51.0	91.0	U.S.	LONG	IND	−.125
4. Bosco et al.	2003	150	Both	13.7	50.0	82.7	U.S.	C-S	Triad	−.120
5. Bray & Harvey – Study 2	1992	712	Intact	19.0	51.5	NR	U.S.	C-S	IND	−.283
6. Buchanan et al.	1991	336	DIV	14.0	49.0	73.0	U.S.	LONG	IND	−.185
7. Dahiyut	1997	84	NR	35.8	100	65.0	U.S.	C-S	IND	−.290
8. Fogel	2000	149	Both	17.5	49.0	89.0	U.S.	C-S	IND	−.201
9. Fortin et al.	2011	79	Intact	10.3	48.1	100.0	Canada	C-S	IND	−.340
10. Fosco & Grych	2010	326	Both	16.3	60.4	56.7	U.S.	LONG	IND	−.180
11. Fosco et al.	2016	236	Both	16.4	60.6	56.8	U.S.	LONG	IND	−.070
12. Grych et al.	2004	388	Intact	16.3	52.3	56.5			IND	−.120
13. Jacobvitz & Bush	1996	84	Intact	20.6	100.0	78.0	U.S.	C-S	IND	−.135
14. Khaleque et al.	2016	524	Intact	13.3	52.0	0.0	Bangladesh	C-S	IND	−.350
15. Koepke	1986	153	Intact	14.5	37.0	82.4	U.S.	C-S	Triad	−.320
16. Leustek & Theiss	2020	171	DIV	20.2	63.2	54.1	U.S.	C-S	IND	−.280
17. Mayseless & Scharf	2009	120	Intact	17.5	100.0	0.0	Israel	C-S	IND	−.228
18. Moye	1985	120	Intact	16.3	100.0	NR	U.S.	C-S	Triad	−.324
19. Nebus	1998	115	Both	19.0	57.4	55.7	U.S.	C-S	IND	−.284
20. O’Hara et al.	2024	1032	Both	12.5	49.0	82.0	U.S.	C-S	IND	−.183
21. Penner et al.	2024	565	Intact	15.1	67.1	45.0	U.S.	C-S	IND	−.125
22. Schrodt & O’Mara	2019	230	Both	19.4	66.5	79.6	U.S.	C-S	IND	−.320
23. Schrodt & Shimkowski	2013	493	Both	20.3	59.4	70.0	U.S.	C-S	IND	−.259
24. Thorson	2021	196	Both	29.4	72.4	82.7	U.S.	C-S	IND	.130
25. Urcan	2011	90	Intact	38.3	50.0	93.3	U.S.		IND	−.077
26. Walper et al.	2005	493	Both	14.1	48.7	100.0	Germany	LONG	IND	−.365

Note. NR = not reported. DIV = divorced. U.S. = United States. C-S = cross-sectional. LONG = longitudinal. IND = individual.

Coding of studies

Given that some articles reported results from more than one study or findings from multiple participant samples, I followed established meta-analytical practice and selected the individual study or unique sample grouping as the unit of analysis for this review rather than the published article (Borenstein et al., 2021; Lipsey & Wilson, 2001). All samples included within each meta-analysis had to be independent. If an individual report re-analyzed data from a previous study, the data were not counted twice unless they represented independent effects. Most studies used children’s reports of feeling caught and relationship quality in cross-sectional research designs, although one study collected data using dyads but reported the correlations using the individual as the unit of analysis. In two studies that collected data using triads, parent composite scores of triangulation were averaged with individual children reports. In such cases, I re-calculated sample sizes from studies that used triads as the unit of analysis so that N reflected the total number of family members in a given report. To account for this, I added unit of analysis (i.e., individual vs. triad) as a potential moderator of the overall effect. Likewise, some longitudinal studies reported effect sizes at one time point, whereas others reported effect sizes across some or all time points. Given that only one effect estimate per primary study was used in the analysis, I aggregated effects across time points but added research design as a potential moderator of the overall effect.

Data analysis procedures

I selected the correlation coefficient r as the effect size statistic because of its intuitive appeal to a broad audience. When possible, I extracted effect estimates directly from the correlation coefficients reported in primary studies. In cases of multiple measurement of the same variable, I computed average correlations following the principles of variance-centered meta-analysis (Lipsey & Wilson, 2001). Consistent with a random-effects meta-analysis (Borenstein et al., 2021), I did not correct for statistical artifacts (e.g., measurement error). The overall effect estimates reflecting the relationship between children feeling caught and parent-child relationship quality for each primary study appear in Table 1.

To test my hypotheses, I conducted a random-effects model meta-analysis using Comprehensive Meta-Analysis (CMA, v. 4.0) (Borenstein et al., 2022). Random-effects models assume that a sample of primary studies represent a random sample from a universe of potential studies on a topic (Borenstein et al., 2021). In this study, the results from random-effects models can be generalized to the universe of studies examining children feeling caught between their parents and indicators of relationship quality in parent-child relationships. One advantage of meta-analysis is that the presence of Type II error can be minimized by computing an average correlation, weighting each effect size by sample size. In the present meta-analysis, weights were assigned to each primary study using the inverse of the error variance for the study (Borenstein et al., 2021). Study correlations were transformed to the Fisher’s z scale to calculate average effects and confidence intervals. These estimates were then back-transformed into weighted mean zero-order correlations (r) and 95% confidence intervals. These confidence intervals display the possibility that the mean correlation in the defined universe of studies could fall anywhere between the lower limit and upper limit (Borenstein et al., 2021).

Assuming that effect sizes would vary across populations, I used several measures to estimate the extent of the variation and quantify heterogeneity. Specifically, for each effect size, I calculated: (a) the Q statistic, which tests the null hypothesis that all studies in the analysis share a common effect size; (b) tau-squared (T²), which represents the variance of true effect sizes; (c) tau (T), which represents the standard deviation of true effect sizes; and (d) the I² statistic, which is the proportion of observed variance attributable to differences in true effect sizes as opposed to sampling error (Borenstein et al., 2021). In addition, I obtained prediction intervals to provide the range of true effect sizes expected for 95% of all populations and representing highly probable values for true effects in future research (Borenstein et al., 2017).

To test my hypotheses and explore other sources of heterogeneity in effect sizes, I coded the following categorical moderators: family type (0 = intact, 1 = divorced, 2 = both), age group (0 = adolescent, ages 10–17, 1 = young adult, ages 18+), research design (0 = cross-sectional, 1 = longitudinal), unit of analysis (0 = individual, 1 = dyad, 2 = triad), measure of feeling caught,⁴ country (0 = U.S., 1 = Non-U.S.), and publication status (0 = published, 1 = unpublished). I also coded three continuous moderators for studies that provided the following information: average age of the children (in years) (k = 26), percentage of female participants (k = 26), and percentage of White participants (k = 24). I examined categorical moderators using mixed-effects models with random effects within subgroups and fixed effects across subgroups to test for significant differences (Q_B) between subgroup mean effect sizes (Hedges & Pigott, 2004). I examined continuous moderators with random-effects meta-regression models using the Z distribution with unrestricted maximum likelihood estimation. Finally, I conducted sensitivity analyses to estimate publication bias, which is the tendency for studies with nonsignificant findings to remain unpublished and excluded from the meta-analysis. Publication bias becomes problematic when it leads to an inflated estimate of the weighted mean effect across a sample of primary studies (Rothstein et al., 2005).

Results

H₁ predicted that children feeling caught between their biological parents would be inversely associated with parent-child relationship quality. A random-effects model meta-analysis showed that the weighted mean effect estimate for the relationship between children feeling caught and relationship quality across the entire sample was r = −.218, k = 26, N = 8,061. The 95% confidence interval (95% CI) for the mean effect ranged from −.260 to −.175. These results indicated that greater levels of children feeling caught between their parents were inversely associated with parent-child relationship quality. Thus, H₁ was supported.

Measures of heterogeneity were then inspected. The Q-statistic revealed that the effect estimates in the sample were heterogenous, Q(25) = 92.09, p < .001. The values for τ² and τ were .009 and .094, respectively. The value for I², which is a descriptive statistic representing the ratio of true to total variance in a sample of effect estimates (i.e., a signal-to-noise ratio), was 73%, indicating that some 73% of the variance in observed effects reflected variance in true effects rather than sampling error. Collectively, these results suggested the presence of moderators. Finally, a prediction interval, which is a measure of dispersion indicating the distribution or range of true effect estimates (Borenstein et al., 2021), was computed for the weighted mean effect estimate. Assuming that the true effects are normally distributed, the 95% PI ranged from −.397 to −.022. This indicated that the true effect size in 95% of all comparable populations falls in this interval. The forest plot for this meta-analysis, including each study’s correlation, plotted CI, and relative weight, is displayed in Figure 1.

Figure 1.

Forest plot for feeling caught and relationship quality.

Moderators of feeling caught and relationship quality

H₂ predicted that family structure would moderate the association between children feeling caught and parent-child relationship quality, such that the inverse association would be stronger for samples that included predominantly intact families than for samples that included only post-divorce families. Although the overall effect size for studies that sampled exclusively intact families (r = −.234, k = 10, 95% CI: −.306, −.159; p < .001) was slightly larger than the overall effect sizes for studies that included both family types (r = −.201, k = 13, 95% CI: −.262, −.139; p < .001) or sampled exclusively post-divorce families (r = −.228, k = 2, 95% CI: −.379, −.065; p < .001), the differences were not statistically significant, Q_B(2) = 0.457, p = .796. Thus, H₂ was not supported.

H₃ predicted that the inverse association between children feeling caught and parent-child relationship quality would be stronger for young adult children than for adolescent children. Contrary to this hypothesis, the omnibus Q-test for age group was not statistically significant, Q_B(1) = .030, p = .863, as the observed effect estimate for the relationship between children feeling caught and parent-child relationship quality was nearly identical for studies that sampled adolescent children (r = −.218, k = 14, 95% CI: −.276, −.158; p < .001) compared with studies that sampled young adult children (r = −.210, k = 11, 95% CI: −.277, −.140; p < .001). Likewise, a random-effects meta-regression using the average age of the children in each study as the predictor of the observed effect estimate for feeling caught and relationship quality was not statistically significant, k = 26, b = .007, SE = .004, t(24) = 1.99, p = .058, R² analog = .20. Hence, the results failed to support H₃.

I then searched for additional moderators, beginning with other demographic characteristics of the samples in each primary study. Studies conducted outside of the U.S. (r = −.332, k = 4, 95% CI: −.419, −.239; p < .001) produced a much larger observed effect estimate than studies conducted within the U.S. (r = −.197, k = 22, 95% CI: −.237, −.155; p < .001), Q_B(1) = 6.76, p = .009. However, two separate random-effects meta-regression models revealed that neither the gender composition of the samples (as indicated by percentage of female participants) included in the primary studies, k = 26, b = .0006, SE = .002, t(24) = 0.38, p = .704, nor the racial composition of the samples (as indicated by percentage of participants who identified as White) included in the primary studies were statistically significant predictors of the observed effect estimate, k = 24, b = .0003, SE = .001, t(22) = 0.34, p = .735.

In terms of other potential moderators related to research design and methods, studies using parent-child triads as the unit of analysis produced slightly larger observed effects (r = −.255, k = 3, 95% CI: −.385, −.114) than studies using individual children as the unit of analysis (r = −.211, k = 22, 95% CI: −.257, −.164), although the difference was not statistically significant, Q_B(1) = .349, p = .555.⁵ Likewise, cross-sectional studies produced slightly larger observed effects (r = −.225, k = 20, 95% CI: −.275, −.174; p < .001) than longitudinal studies (r = −.195, k = 6, 95% CI: −.281, −.107; p < .001), though again, the difference was not statistically significant, Q_B(1) = .339, p = .560. Neither the specific measure of feeling caught that researchers used, Q_B(3) = .440, p = .932, nor the publication status of the research, Q_B(1) = .618, p = .432, explained heterogeneity in the overall effect estimate for children feeling caught and parent-child relationship quality.

Taken as a whole, the nationality of family members explained some of the heterogeneity in the observed effect size estimate but left much of the heterogeneity unresolved. Per Levine and Weber’s (2020) recommendations for exploring other sources of heterogeneity (e.g., combined construct invalidity), I conducted one final test by recoding and comparing studies that used psychological boundary indicators of relationship quality (e.g., individuation/differentiation, parent-child insecurity) with studies that used affective indicators of relationship quality (e.g., satisfaction, closeness, positive/negative feelings). The former produced a larger observed effect size estimate (r = −.253, k = 14, 95% CI: −.309, −.194; p < .001) than the latter (r = −.181, k = 12, 95% CI: −.240, −.119; p < .001), although the difference was not statistically significant, Q_B(1) = 2.85, p = .091. Nevertheless, some of the unresolved heterogeneity may be due to combining different indicators of relationship quality, a higher-order construct that is admittedly less concrete than specific outcomes such as relational satisfaction, turbulence, or closeness.

Publication bias

I used three complementary approaches to evaluate the potential for publication bias in this project. First, as illustrated in Figure 2, I constructed a contour-enhanced funnel plot (Peters et al., 2008) using the overall effect estimate for each primary study on the horizontal axis and the corresponding standard error on the vertical axis. If the meta-analysis captured all of the relevant studies, the funnel plot would be symmetric, that is, the studies would be dispersed equally on either side of the overall effect, which is represented by the vertical dashed line. The shading represents different levels of statistical significance, and the areas without shading indicate non-significance. Publication bias is likely when the plot is asymmetrical and primary studies are missing from the unshaded areas of the plot. With the exception of one primary study that appeared in the unshaded area of the plot, the points of the plot in Figure 2 appear to be symmetrical.

Figure 2.

Contour-enhanced funnel plot depicting observed effect estimates and standard errors for the relationship between children feeling caught between parents and parent-child relationship quality (k = 26).

Second, I used Begg and Mazumdar’s (1994) rank correlation test to determine the correlation between effect size and the standard error of a primary study. Although a significant correlation suggests that bias exists, it does not directly address the implications of this bias, whereas a non-significant correlation may be due to low statistical power and cannot be taken as evidence that bias is absent. In this case, the correlation was not statistically significant, Kendall’s τ = .003, p_one-tailed = .491, providing no evidence to suggest that primary studies containing smaller samples reported larger effect estimates.

Finally, I obtained a regression model using the inverse of the standard error for primary studies as a predictor of the observed standardized effect estimates (Egger et al., 1997). According to Egger et al. (1997), under some circumstances, this approach constitutes a more powerful test than the rank correlation test, with publication bias determined by evaluating whether the intercept differs from zero. In this case, the intercept was not statistically significant, b₀ = .174, t(24) = .167, p_one-tailed = .435. Taken together, these two additional tests were consistent with the funnel plot and indicated that the impact of publication bias was probably trivial. If all relevant studies were included, the observed effect estimate would probably remain largely unchanged.

Discussion

By synthesizing an interdisciplinary and voluminous body of research that spans nearly 40 years, this meta-analysis provides an empirical summary and examination of the available data on the relationship between children feeling caught between parents and parent-child relationship quality. Overall, the results indicate that, on average, feeling caught has a small but meaningful inverse association with parent-child relationship quality. Subsequent analyses of potential moderators revealed only one statistically significant difference in the average effect size based on the nationality of the families sampled, although the average age of the children may also moderate the average effect. Conversely, the results provide no evidence to suggest that the overall effect varies as a function of divorce, children’s gender, or the Whiteness of the children sampled, nor as a function of which measure researchers used to assess children’s feelings of being caught or their relationship quality with parents. Consequently, these results extend the feeling caught literature by providing at least three implications worth noting.

First, although the overall effect size for feeling caught and relationship quality was small in magnitude, it is nevertheless meaningful given the tremendous breadth and scope of the samples and measures scholars have used to examine the association. The overall effect estimate of r = −.22, which is based on 26 primary studies that included more than 8,000 individuals from more than 7,800 families across five countries, is nearly identical with the mean effect estimate generated by other meta-analyses of communication phenomena. For instance, in their synthesis of 149 meta-analyses conducted across the communication discipline, Rains et al. (2018) reported a mean effect estimate for communication phenomena of r = .21. Given that meta-analysis provides greater statistical power and broader generalizability than any single primary study, feeling caught appears to be a fairly consistent predictor of diminished relationship quality with parents.

A second set of implications comes from two potential moderators of the overall effect size for feeling caught and relationship quality. Specifically, the findings indicate that feeling caught may undermine relationship quality with parents to a greater degree for children outside of the U.S. than for children living in the U.S.. Although this difference should be interpreted with caution given a limited number of studies that sampled families from other countries, it does raise interesting questions as to what might explain the difference. Perhaps differences in cultural values, such as individualism as opposed to collectivism, permeate family socialization processes in a manner that renders parent-child triangulation and feeling caught less harmful to relationship quality when family members value the individual more so than the family as a whole. In other words, few countries are more individualistic than the U.S., and perhaps high levels of individualism encourage children to make stronger distinctions between their parents’ relationship and their own individual relationships with each parent, which in turn might reduce the negative effect that feeling caught has on their reports of relationship quality. However, this explanation is limited by the fact that the non-U.S. samples in this meta-analysis came from both individualistic countries (e.g., Germany and Canada) and collectivistic countries (e.g., Bangladesh, and to a lesser degree, Israel).

An alternative explanation is that the difference in magnitude was created by differences in the types of measures scholars used to assess parent-child relationship quality in families outside of the U.S.. A closer inspection of the non-U.S. studies, and specifically the two largest and most heavily weighted non-U.S. studies (i.e., Bangladesh and Germany), reveals measures of relationship quality that focus on insecurity, rejection, and psychological separation in the parent-child relationship. Perhaps the significant difference in the magnitude of the overall effect size estimate between U.S. and non-U.S. samples represents something akin to moderated moderation, whereby the difference based on nationality depends on the type of outcome measure researchers used to assess relationship quality. In support of this explanation, the average effect size was larger when scholars assessed boundary-focused as opposed to affective measures of relationship quality, but the difference was not statistically significant. Nevertheless, the observed moderation based on nationality may be due to an interaction between nationality and type of relationship quality measure, although meta-analysis is not equipped to further explore moderated moderation in this sense.

In addition to nationality, the findings also point to age of the child as a potential moderator (though to a lesser extent). Although the group comparison between adolescent and young adult children revealed nearly identical effect size estimates, a meta-regression using the average age of the samples as the predictor variable approached statistical significance (p = .058) and explained 20% of the between-studies variance in the overall effect size estimate. One explanation for this potential moderator of feeling caught and relationship quality comes from Arnett’s (2004) theory of emerging adulthood. According to Arnett (2004), emerging adulthood is a developmental time period (typically ages 18–29) in which children transition from being completely dependent upon their parents to assuming full financial and emotional responsibility for their own well-being. Emerging adults are often dependent on their parents’ emotional support for many years. Thus, interparental conflict and parents’ attempts to involve their emerging adult children in their disputes may threaten both the amount and the types of emotional support they can provide their children during this developmental transition, ultimately hurting relationship quality with one or both parents. It may also be the case that parents’ guilt-inducing behaviors, calls for loyalty, and pressures to remain involved in their ongoing disputes undermine relationship quality because such experiences run contrary to the emerging adult child’s movement toward complete independence from the family.

The final set of implications come from the unresolved heterogeneity in the overall effect size estimate, and more precisely, from the inability of other potential moderators to explain the heterogeneity. Contrary to what I hypothesized, the divorce status of the parents did not moderate the overall effect size estimate of feeling caught on parent-child relationship quality. This is somewhat surprising given that divorce status moderates the direct and indirect effects of interparental conflict (Schrodt & LaFreniere, 2022), coparental communication (Schrodt & Shimkowski, 2013), and family communication patterns (Schrodt & Ledbetter, 2007) on children’s personal and relational well-being vis-à-vis feeling caught. Perhaps divorce and remarriage alter the communicative antecedents of feeling caught more so than the relationship outcomes of feeling caught. In other words, if feeling caught emanates from unhealthy coalitions within the family system and constitutes a form of boundary dissolution that problematizes how children interact with both parents, then perhaps such feelings undermine relationship quality comparably in both family types because the parent-child relationships persist even if the spousal relationship does not.

Likewise, this meta-analysis provided no evidence to suggest that the average effect size changes in magnitude as a function of child gender, the Whiteness of the samples participating in the research, or as a function of the measures used to assess feelings of being caught and relationship quality. Although the indicators of gender and race are limited and function more so as proxies than as precise measures, the fact that they did not moderate the overall effect size estimate strengthens, to some degree, the generalizability of the findings. On the other hand, the failure of most moderators to explain the heterogeneity in the overall effect size estimate does point to combined construct invalidity as one possible source or explanation for the unresolved heterogeneity (Levine & Weber, 2020). After probing this further, I did find that relationship outcome measures that were more psychological and boundary-focused in nature tended to produce larger effect sizes for feeling caught, on average, than affective measures such as satisfaction, positive feelings, and closeness. Although the difference was not statistically significant, from the perspective of FST (Bowen, 1976, 1978; Minuchin, 1974), boundary-focused outcomes (e.g., psychological separation/enmeshment, fusion, relationship insecurity) may be more closely associated with feeling caught (as a form of triangulation and boundary dissolution) than global indicators of satisfaction, positive affect, and closeness. At a minimum, relationship quality may function as a higher-order construct that is less concrete than some of the individual outcomes that were combined in this meta-analysis, and the less concrete the construct, the more susceptible the analysis will be to combined construct invalidity and to unexplained heterogeneity (Levine & Weber, 2020).

Theoretically, the results of this meta-analysis confirm the theoretic logic of FST (Bowen, 1978; Minuchin, 1974). They support the idea that system disequilibrium in the (ex)spousal subsystem that spills over into the parent-child subsystem via boundary diffusion, guilt-inducing behaviors, and coalition-building threatens the quality of children’s relationships with their parents. In essence, the overall effect size estimate obtained here confirms the emotional interdependence that exists between the interparental subsystem and the parent-child subsystem, as distress within the former inculcates a unique form of tension and stress in the latter that undermines children’s feelings of satisfaction, safety, emotional security, and trust with their parents.

Limitations and directions for future research

Although this meta-analysis benefitted from interdisciplinary diversity, a systematic process of identifying primary studies, and trivial amounts of publication bias, the implications and contributions should be interpreted with caution given the inherent limitations of the research design. Whereas most researchers collected data from individual children, one research team collected data from parent-child dyads but reported results using the individual as the unit of analysis, and two collected data from triads but reported their results using composite scores for parents’ and children’s reports. In order to summarize the effects from these studies using the correct number of data points, I followed suit and used the individual rather than the couple or the triad as the N for such studies. This procedure may have biased the results by averaging across different family member reports and by giving more weight to those studies that analyzed feeling caught using the individual rather than the dyad or triad as the unit of analysis. However, the weights assigned to each study in the random-effects meta-analysis differed only slightly. Although I was unable to conduct a formal test of moderation based on the study’s unit of analysis, the degree of bias introduced by this procedure appears to be trivial.

Second, the samples in this meta-analysis are highly homogenous with respect to its populations. The test of heterogeneity for family race was limited to percentages of White participants within samples, preventing more meaningful comparisons among different racial/ethnic groups. Despite including children from five countries on three different continents (Asia, Europe, North America), no country other than the U.S. was represented in more than one primary study, preventing more specific comparisons of country of origin as a moderator. Finally, 25 studies were excluded from the initial database for failing to report statistical information necessary to calculate an effect estimate. This emphasizes the importance of providing descriptive statistics and simple effect sizes in future research reports for the purposes of meta-analysis.

These limitations notwithstanding, this meta-analysis provides a conceptual overview and empirical summary of children feeling caught and parent-child relationship quality. It confirms the small but meaningful, inverse association that feeling caught has with relationship quality. It suggests that the association may vary in strength as a function of differences in the cultural values of family members and as the age of the child increases. Given unresolved heterogeneity in the overall effect size estimate, this meta-analysis also raises questions for future scholars to consider as they seek to identify other sources of contingency upon which the association might depend. To date, most investigations of the outcomes of children feeling caught have focused either on mental and emotional well-being or on relationship outcomes. Much less is known about how children communicate with parents in the aftermath of feeling caught, as well as how the ongoing management of feeling caught changes relationship quality over time. Thus, an important direction for future research would be to identify and compare the relative strength of different communication behaviors in helping children cope with the tension, confusion, and emotional insecurity that often comes from feeling caught. Likewise, researchers could consider how children appraise parent-child interactions once feelings of being caught have been triggered, and examine whether and how attributions of responsibility (e.g., for their parents’ relationship problems) and parents’ intentions tie their own feelings of being caught to relational uncertainty and future intentions to maintain relationships with their parents. Finally, efforts to develop and test a communication theory of feeling caught would help strengthen this body of work by providing a message-centered explanation for what is essentially a communicative and relational phenomenon. By considering the messages that follow feelings of being caught and tying them to changes in relationship quality, scholars can work toward a more complete theoretical explanation of this difficult communication dynamic in family systems.

Supplemental Material

Supplemental Material - A meta-analysis of children feeling caught between parents and parent-child relationship quality

Supplemental Material for A meta-analysis of children feeling caught between parents and parent-child relationship quality by Paul Schrodt in Journal of Social and Personal Relationships

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Ethical Considerations

Ethical approval was not required for this meta-analysis.

ORCID iD

Paul Schrodt

Open Research Statement

Anonymized: As part of IARR’s encouragement of open research practices, the author has provided the following information: (a) the research was not pre-registered, (b) the data used in the research are available upon request, and (c) the materials used in the research are available upon request.

Non-anonymized: As part of IARR’s encouragement of open research practices, the author has provided the following information: (a) the research was not pre-registered, (b) the data used in the research are available upon request to p.schrodt@tcu.edu, and (c) the materials used in the research are available upon request to p.schrodt@tcu.edu.

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Afifi

T. D.

(2003). ‘Feeling caught’ in stepfamilies: Managing boundary turbulence through appropriate communication privacy rules. Journal of Social and Personal Relationships, 20(6), 729–755. https://doi.org/10.1177/0265407503206002

Afifi

T. D.

Afifi

W. A.

Morse

C. R.

Hamrick

(2008). Adolescents’ avoidance tendencies and physiological reactions to discussions about their parents’ relationship: Implications for postdivorce and nondivorced families. Communication Monographs, 75(3), 290–317. https://doi.org/10.1080/03637750802342308

Afifi

T. D.

Schrodt

(2003). “Feeling caught” as a mediator of adolescents’ and young adults’ avoidance and satisfaction with their parents in divorced and non-divorced households. Communication Monographs, 70(2), 142–173. https://doi.org/10.1080/0363775032000133791

Amato

P. R.

Afifi

T. D.

(2006). Feeling caught between parents: Adult children’s relations with parents and subjective well-being. Journal of Marriage and Family, 68(1), 222–235. https://doi.org/10.1111/j.1741-3737.2006.00243.x

Arnett

J. J.

(2004). Emerging adulthood: The winding road from the late teens through the twenties. Oxford University Press.

Begg

C. B.

Muzamdar

(1994). Operating characteristics of a rank correlation test for publication bias. Biometrics, 50(4), 1088–1101. https://doi.org/10.2307/2533446

Borenstein

Hedges

L. V.

Higgins

J. P. T.

Rothstein

H. R.

(2021). Introduction to meta-analysis (2nd ed.). Wiley.

Borenstein

Hedges

L. V.

Higgins

J. P. T.

Rothstein

H. R.

(2022). Comprehensive meta-analysis (Version 4) [Computer software]. Biostat, Inc. https://www.meta-analysis.com/.

Borenstein

Higgins

J. P. T.

Hedges

L. V.

Rothstein

H. R.

(2017). Basics of meta-analysis: I² is not an absolute measure of heterogeneity. Research Synthesis Methods, 8(1), 5–18. https://doi.org/10.1002/jrsm.1230

10.

Bowen

(1976). Theory in the practice of psychotherapy. In Guerin

P. J.

(Ed.), Family therapy: Theory and practice (pp. 42–90). Gardner Press.

11.

Bowen

(1978). Family therapy in clinical practice. J. Aronson.

12.

Buchanan

C. M.

Maccoby

E. E.

Dornbusch

(1991). Caught between parents: Adolescents' experience in divorced homes. Child Development, 62(5), 1008–1029. https://doi.org/10.2307/1131149

13.

Cox

M. J.

Paley

Harter

(2001). Interparental conflict and parent-child relationships. In Grych

J. H.

Fincham

F. D.

(Eds.), Interparental conflict and child development: Theory, research and applications (pp. 249–272). Cambridge University Press. https://doi.org/10.1017/CBO9780511527838.011

14.

Davies

P. T.

Martin

M. J.

(2013). The reformulation of emotional security theory: The role of children’s social defense in developmental psychopathology. Development and Psychopathology, 25(4), 1435–1454. https://doi.org/10.1017/S0954579413000709

15.

Egger

Smith

G. D.

Schneider

Minder

(1997). Bias in meta-analysis detected by a simple, graphical test. British Medical Journal, 315(7109), 629–634. https://doi.org/10.1136/bmj.315.7109.629

16.

Fosco

G. M.

Grych

J. H.

(2010). Adolescent triangulation into parental conflicts: Longitudinal implications for appraisals and adolescent-parent relations. Journal of Marriage and Family, 72(1), 254–266. https://doi.org/10.1111/j.1741-3737.2010.00697.x

17.

Fosco

G. M.

Lippold

Feinberg

M. E.

(2014). Interparental boundary problems, parent-adolescent hostility, and adolescent-parent hostility: A family process model for adolescent aggression problems. Couple & Family Psychology, 3(3), 141–155. https://doi.org/10.1037/cfp0000025

18.

Fosco

G. M.

Xia

Lynn

M. G.

Grych

J. H.

(2016). Triangulation and parent-adolescent relationships: Implications for adolescent dating competence and abuse. Journal of Research on Adolescence: The Official Journal of the Society for Research on Adolescence, 26(3), 524–537. https://doi.org/10.1111/jora.12210

19.

Hedges

L. V.

Pigott

T. D.

(2004). The power of statistical tests for moderators in meta-analysis. Psychological Methods, 9(4), 426–445. https://doi.org/10.1037/1082-989X.9.4.426

20.

Johnston

J. R.

Kline

Tschann

J. M.

(1989). Ongoing postdivorce conflict: Effects on children of joint custody and frequent access. American Journal of Orthopsychiatry, 59(4), 576–592. https://doi.org/10.1111/j.1939-0025.1989.tb02748.x

21.

Kerig

P. K.

(2005). Revisiting the construct of boundary dissolution. Journal of Emotional Abuse, 5(2-3), 5–42. https://doi.org/10.1300/J135v05n02_02

22.

Leustek

Theiss

J. A.

(2020). Family communication patterns that predict perceptions of upheaval and psychological well-being for emerging adult children following late-life divorce. Journal of Family Studies, 26(2), 169–187. https://doi.org/10.1080/13229400.2017.1352531

23.

Levine

T. R.

Weber

(2020). Unresolved heterogeneity in meta-analysis: Combined construct invalidity, confounding, and other challenges to understanding mean effect sizes. Human Communication Research, 46(2–3), 343–354. https://doi.org/10.1093/hcr/hqz019

24.

Lipsey

M. W.

Wilson

D. B.

(2001). Practical meta-analysis (Applied Social Research Methods Series, Vol. 49). SAGE.

25.

Mayseless

Scharf

(2009). Too close for comfort: Inadequate boundaries with parents and individuation in late adolescent girls. American Journal of Orthopsychiatry, 79(2), 191–202. https://doi.org/10.1037/a001562

26.

Minuchin

(1974). Family and family therapy. Harvard Press.

27.

Moher

Liberati

Tetzlaff

Altman

D. G.

PRISMA Group . (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Medicine, 6(7), e1000097. https://doi.org/10.1371/journal.pmed.1000097

28.

O’Hara

K. L.

Cummings

E. M.

Davies

P. T.

(2024). Interparental conflict and adolescent emotional security across family structures. Family Process, 63(2), 265–283. https://doi.org/10.1111/famp.12872

29.

Peters

J. L.

Sutton

A. J.

Jones

D. R.

Abrams

K. R.

Rushton

(2008). Contour-enhanced meta-analysis funnel plots to help distinguish publication bias from other causes of asymmetry. Journal of Clinical Epidemiology, 61(10), 991–996. https://doi.org/10.1016/j.jclinepi.2007.11.010

30.

Rains

S. A.

Levine

T. R.

Weber

(2018). Sixty years of quantitative communication research summarized: Lessons from 149 meta-analyses. Annals of the International Communication Association, 42(2), 105–124. https://doi.org/10.1080/23808985.2018.1446350

31.

Rothstein

H. R.

Sutton

A. J.

Borenstein

(2005). Publication bias in meta-analysis. In Rothstein

H. R.

Sutton

A. J.

Borenstein

(Eds.), Publication bias in meta-analysis: Prevention, assessment and adjustments (pp. 1–7). Wiley. https://doi.org/10.1002/0470870168.ch1

32.

Schrodt

(2016). Coparental communication with nonresidential parents as a predictor of children’s feelings of being caught in stepfamilies. Communication Reports, 29(1), 63–74. https://doi.org/10.1080/08934215.2015.1020562

33.

Schrodt

(2025a). Interparental conflict and parent-child triangulation: A meta-analytical review of children feeling caught between parents. Human Communication Research. https://doi.org/10.1093/hcr/hqaf018

34.

Schrodt

(2025b). A meta-analytical review of parent-child communication and children feeling caught between parents. [Manuscript submitted for publication]. Department of Communication Studies, Texas Christian University.

35.

Schrodt

Afifi

T. D.

(2007). Communication processes that predict young adults’ feelings of being caught and their associations with mental health and family satisfaction. Communication Monographs, 74(2), 200–228. https://doi.org/10.1080/03637750701390085

36.

Schrodt

Afifi

T. D.

(2018a). Untying the ties that bind: Dispositional and relational patterns of negative relational disclosures and family members’ feelings of being caught. Journal of Family Issues, 39(7), 1962–1983. https://doi.org/10.1177/0192513X17739050

37.

Schrodt

Afifi

T. D.

(2018b). Negative disclosures and feeling caught mediate coparental communication and mental health. Personal Relationships, 25(4), 480–496. https://doi.org/10.1111/pere.12256

38.

Schrodt

LaFreniere

J. R.

(2022). Witnessing interparental conflict, feeling caught, and mental health: A conditional process analysis involving parental confirmation and divorce status. Journal of Social and Personal Relationships, 39(2), 303–324. https://doi.org/10.1177/02654075211036508

39.

Schrodt

Ledbetter

A. M.

(2007). Communication processes that mediate family communication patterns and mental well-being: A mean and covariance structures analysis of young adults from divorced and non-divorced families. Human Communication Research, 33(3), 330–356. https://doi.org/10.1111/j.1468-2958.2007.00302.x

40.

Schrodt

O’Mara

(2019). The development and validation of the emotion labor in families scale: Associations with emotion regulation, feeling caught, and relational satisfaction in parent-child relationships. Communication Quarterly, 67(4), 383–404. https://doi.org/10.1080/01463373.2019.1596143

41.

Schrodt

Shimkowski

J. R.

(2013). Feeling caught as a mediator of coparental communication and young adults’ mental health and relational satisfaction with parents. Journal of Social and Personal Relationships, 30(8), 977–999. https://doi.org/10.1177/0265407513479213

42.

Thorson

A. R.

(2021). Triangulation and parental infidelity: Faithful parents’ attempts to make their adult children feel caught, rumination, and satisfaction in the unfaithful parent-child relationship. Journal of Divorce & Remarriage, 62(5), 327–348. https://doi.org/10.1080/10502556.2020.1833293

43.

Urcan

J. D.

(2011). Relationship of family of origin qualities and forgiveness to marital satisfaction (Publication No. 3460962) [Doctoral dissertation. Hofstra University]. ProQuest Dissertations and Theses Global.

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.

0.00 MB

0.34 MB