Childhood Physical Victimization and Relationship Dysfunction in Justice-Involved Women: A Path Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Strengths:

1. Important and Underexplored Topic: Your study addresses a significant gap in criminology by examining the pathways linking childhood victimization to relationship dysfunction in justice-involved women.
2. Theoretical and Practical Contributions: The emphasis on gender-specific pathways and relational challenges enriches the discourse on trauma-informed interventions for justice-involved women.
3. Methodological Rigor: The use of path analysis is well-suited to explore the complex relationships among victimization, self-esteem, self-efficacy, and relational outcomes.
4. Clear Writing: Your manuscript is generally well-written, with clear explanations of the study’s aims, methods, and findings.

Suggestions for Improvement:

1. Introduction:
• Streamline the discussion of women’s pathways to justice involvement to avoid repetition (lines 40–58).
• Expand on the theoretical importance of self-esteem and self-efficacy as mediators. Why were these constructs chosen, and how do they fit into existing frameworks?
2. Literature Review:
• Incorporate more recent studies (post-2018) on trauma-informed approaches and gender-specific justice pathways (lines 90–130).
• Clarify the gap in the literature more explicitly by emphasizing how your focus on mediation is novel.
3. Methods:
• Justify the choice of Maui County, Hawaii, as the study setting, and discuss the potential limitations for generalizability (lines 135–140).
• Explain why certain items were excluded from the Relationship Dysfunction scale and how recall bias in self-reporting was mitigated (lines 150–180).
4. Results:
• Provide more interpretation of why self-esteem and self-efficacy did not mediate the relationships (lines 220–260). Consider exploring potential methodological or theoretical reasons for this result.
• Acknowledge the small sample size and its implications for statistical power and generalizability (lines 270–280).
5. Discussion:
• Deepen the analysis of the direct relationship between childhood victimization and relationship dysfunction. For example, discuss systemic factors that might contribute to this strong direct effect (lines 300–340).
• Expand on the practical applications of your findings, particularly how they might inform trauma-informed interventions for justice-involved women.
• Address the limitations of the study more explicitly, including the reliance on cross-sectional data and the absence of longitudinal designs.
6. Conclusion:
• Strengthen the conclusion by emphasizing the broader policy and practice implications of your findings (lines 380–400).
• Include specific suggestions for future research, such as exploring additional mediators, longitudinal designs, or cross-gender comparisons.
7. References:
• Verify that all in-text citations are included in the reference list and vice versa.
• Update the references with more recent studies where applicable.

Overall Impression:

Your study is a valuable contribution to the field, shedding light on the enduring impact of childhood victimization on relational outcomes in justice-involved women. With minor revisions to improve clarity, update references, and expand interpretations, this article has the potential to make a significant impact on both research and practice.

Thank you for your thoughtful work on this important topic!

Author Response

Comment 1: Introduction:

Streamline the discussion of women’s pathways to justice involvement to avoid repetition (lines 40–58).

Response 1: We agree. The literature review (p. 1-4) was edited significantly for clarity. The pathways discussion has been edited down (p.1-2) and more emphasis has been placed on the specific mechanisms being studied.

Comment 2: Expand on the theoretical importance of self-esteem and self-efficacy as mediators. Why were these constructs chosen, and how do they fit into existing frameworks?

Response 2: We agree. The literature review (p. 1-4) was edited significantly for clarity. Additional research was added regarding the role of self-esteem and self-efficacy in the child abuse and relationship quality literature (p. 2-3, 82-104 and p. 3, 122-138).

Comment3: Literature Review:

Incorporate more recent studies (post-2018) on trauma-informed approaches and gender-specific justice pathways (lines 90–130).

Response 3: More recent literature on trauma and pathways has been included to support the seminal works previously cited. This was updated throughout the paper.

Comment 4: Clarify the gap in the literature more explicitly by emphasizing how your focus on mediation is novel.

Response 4: The current study section (p. 4, 169-191) was updated to better articulate the gap in literature and question to be addressed.

Comment 5: Methods:

Justify the choice of Maui County, Hawaii, as the study setting, and discuss the potential limitations for generalizability (lines 135–140).

Response 5: Additional context and justification regarding the study sample was included on p. 4-5 (201- 204) and p. 7 (231-234).

Comment 6: Explain why certain items were excluded from the Relationship Dysfunction scale and how recall bias in self-reporting was mitigated (lines 150–180).

Response 6: Agreed. This was not sufficiently clear. Information on how the items were chosen from the broader scale used in the pilot was added (p.8, 268-282). Specifically, a correlation analysis and exploratory factor analysis were conducted. A table was added with the factor loadings and a complete list of the items. Prevalence of the 13 types of physical abuse were added to Table 1 (p.7) showing demographics of the sample.

Comment 7: Results:

Provide more interpretation of why self-esteem and self-efficacy did not mediate the relationships (lines 220–260). Consider exploring potential methodological or theoretical reasons for this result.

Response7: The discussion was edited to reflect more context for the findings on p.11 (358-377).

Comment 8: Acknowledge the small sample size and its implications for statistical power and generalizability (lines 270–280)

Response 8: Respectfully, the limitations of the sample size are addressed in the Discussion on p. 12.

Comment 9: Discussion:

Deepen the analysis of the direct relationship between childhood victimization and relationship dysfunction. For example, discuss systemic factors that might contribute to this strong direct effect (lines 300–340).

Response9: Additional context for this relationship was added to the front end of the paper in section 1.2 (p. 2-3, 82-104)

Comment 10: Expand on the practical applications of your findings, particularly how they might inform trauma-informed interventions for justice-involved women.

Response 10: The practical applications of the findings were added to the discussion on p.12 (395-401).

Comment 11: Address the limitations of the study more explicitly, including the reliance on cross-sectional data and the absence of longitudinal designs.

Response: Respectfully, the limitations of the study design are addressed in the Discussion on p. 12.

Comment 12: Conclusion:

Strengthen the conclusion by emphasizing the broader policy and practice implications of your findings (lines 380–400).

Response 12: Broader implications were added to the discussion on p.12 (401-406).

Comment 13: Include specific suggestions for future research, such as exploring additional mediators, longitudinal designs, or cross-gender comparisons.

Response 13: Respectfully, these recommendations are addressed with the limitations in the Discussion on p. 12.

Comment 14: References:

Verify that all in-text citations are included in the reference list and vice versa.

Response 14: These were reviewed and checked.

Comment 15: Update the references with more recent studies where applicable.

Response 15: Updated references were added throughout to support seminal texts.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

 

This study examined the relationship between childhood abuse and adult relationship dysfunction among a sample of justice-involved women, and explored the potential mediating roles of self-esteem and self-efficacy in this relationship. I think the study has the potential to make valuable contributions to the bodies of literature on childhood victimization and women’s justice-involvement. However, there are some drawbacks of the current version of the manuscript that need to be addressed before the paper is suitable for publication. Below, I detail my concerns, which are largely with the front end and methods sections.

 

Front End (intro/literature review):

 

The front end of the manuscript seems a bit too broad and not tailored enough to the aim of the paper, making it feel somewhat disconnected from the study. Much of the front end focuses on the impact of relationships on women’s criminal offending. The section on the pathways perspective (pp. 1-2) focuses on the relations between victimization and women’s offending, and the section on the relational model (pp. 2-3) focuses on the impact of relationship dysfunction on women’s offending. Together the front end seems to suggest that the dependent variable of the current study is offending, when it is in fact relationship dysfunction among a sample of justice-involved women. These sections do not contribute to setting up core relationship examined in the study – the impact of childhood victimization on later relationship dysfunction. I suggest the authors modify the paper to more strongly connect the concepts discussed in the front end to the relationships that are ultimately examined in the study. Specifically, the authors might make more of an effort to draw on the pathways and relational model perspectives to frame their paper around how childhood victimization can shape relationship dysfunction. I think the authors can do this while still acknowledging that the two perspectives are typically used to explain women’s offending.

 

The authors do a good job, however, in reviewing literature in a way that connects the mediating variables (self-esteem and self-efficacy) to the relationship between child victimization and relationship dysfunction (p. 3).

 

I also think the current study section could be stronger. More of an effort should be made to explicitly draw together the takeaways from the reviewed literature to ground the relationships the study examines. The current study reads, “The relationship between childhood victimization and adult relationship dysfunction is less clear” (p. 3), which is implied to be the overall gap that the study seeks to address. Since this appears to be the central premise on which the study is based, this point needs to be brought up much earlier. The lack of knowledge on this relationship should be explicitly stated earlier in the literature review. In addition, more elaboration is needed on the significance of this gap, beyond the fact that there has been little research on it. The authors could following this statement with a few sentences explicitly clarifying why this gap is problematic and what can be gained in, substantively, by gaining knowledge of this relationship and the path through which childhood victimization and adult relationship dysfunction among justice-involved women.

 

Relatedly, the use of a sample of justice-involved women needs to be more strongly situated in the set-up of the research. It is unclear how the justice-involved nature of this sample fits in within the broader study. Does drawing on a justice-involved sample provide some specific value to investigating the connection between childhood victimization and adult relationship dysfunction, perhaps due to the documented associations between the study’s constructs (from the pathways perspective and relational model) and women’s offending? Another idea would be to frame the study around the need to support justice-involved women’s wellbeing as they transition out of their sentences. The authors allude to the need for therapeutic interventions for this population a bit already: “Treating issues of self-esteem and self-efficacy are central goals of therapeutic interventions for survivors and can promote improved interpersonal relationships in the future” (p. 3). Perhaps the case could be made that relationship dysfunction is common among justice-involved women, which may be due to their disproportionate exposure to childhood victimization. Then the authors could 1) bring up the need to understand this relationship as a whole, and 2) make the cause that understanding the mechanisms through which childhood victimization influences relationship dysfunction (i.e., potential roles of self-esteem and self-efficacy) can provide potential points for intervening.

 

Methods:

 

In the methods section, the authors describe the basic characteristics of their sample, including demographics (age, race, education, marital status, motherhood) as well as prevalence of felony convictism and age of first offense. I think the prevalence of childhood abuse should also be clarified, since it is the study’s main independent variable. The details provided on childhood physical abuse among the sample are not entirely informative (i.e., measure ranging from 0-26 with a mean of 6.39). I would suggest the authors provide a general prevalence of physical abuse in the sample, and perhaps prevalence of the 13 types of physical abuse items for clarity.

Relatedly, the description of the abuse measure reads: “The current study pulled 13 items related to physical abuse experienced in childhood such as ‘spanked/slapped you’, beat you’, and ‘burned/scalded you’.” I was surprised that 13 types of abuse were mentioned, yet only 3 were provided as examples. The entire study is hinged on childhood abuse and its path to adult relationship function, so there is a need for much more information on the type of abuse examined and its prevalence. I realize that listing all 13 types of abuse in the measures description is quite wordy, so I would recommend the authors list the 13 types in Table 2 under the descriptive statistics of the overall measure. They could then provide a prevalence % of each of the 13 types.

Did the study include control variables in the models? I was surprised that none were described in the measures section, since sample demographics and criminal histories were summarized in the sample description and in Table 1. If control variables were included, this should be explicitly stated and their measurement should be described. If no control variables were included, then this needs to be stated and the decision not to include them needs to be justified.

 

More detail should be provided on the analysis. All that is mentioned is that a path model was estimated for the relationship between child abuse, self-esteem and self-efficacy, and relationship dysfunction in adulthood. The authors should indicate the specifics of the path model, for instance the method of determining significance of indirect/mediating effects (there are a wide range of approaches for testing for mediation and a lot of debate as to which is most appropriate depending on data characteristics). The authors indicate that bootstrapping was used, but do not clarify how many bootstrapped draws were used. This is a key analytic detail that should not be omitted. In addition, were self-esteem and self-efficacy modeled together in paths, or were each examined separately?

 

I was surprised that the authors did not provide any of the typical model fit statistics for path models: Comparative Fit Index (CFI), Tucker Lewis Index (TLI), Root Mean Square Root of Approximation (RMSRA). These should be provided, as they are essential in determining whether the path model actually fit the data.

 

One minor point:

 

There is some awkward phrasing in the abstract that should be modified:  “The current study uses a path analysis to investigate the question of if for justice-involved women, is childhood physical abuse related to relationship dysfunction for women in adulthood and is this relationship mediated by self-esteem and/or self-efficacy?”

 

 

 

 

Author Response

Comment 1: Front End (intro/literature review):

The front end of the manuscript seems a bit too broad and not tailored enough to the aim of the paper, making it feel somewhat disconnected from the study. Much of the front end focuses on the impact of relationships on women’s criminal offending. The section on the pathways perspective (pp. 1-2) focuses on the relations between victimization and women’s offending, and the section on the relational model (pp. 2-3) focuses on the impact of relationship dysfunction on women’s offending. Together the front end seems to suggest that the dependent variable of the current study is offending, when it is in fact relationship dysfunction among a sample of justice-involved women. These sections do not contribute to setting up core relationship examined in the study – the impact of childhood victimization on later relationship dysfunction. I suggest the authors modify the paper to more strongly connect the concepts discussed in the front end to the relationships that are ultimately examined in the study. Specifically, the authors might make more of an effort to draw on the pathways and relational model perspectives to frame their paper around how childhood victimization can shape relationship dysfunction. I think the authors can do this while still acknowledging that the two perspectives are typically used to explain women’s offending.

Response 1: We agree with this feedback. The literature review (p. 1-4) was significantly edited to tell a more cohesive story regarding the questions being asked and how they relate to the population being studied.

Comment 2: I also think the current study section could be stronger. More of an effort should be made to explicitly draw together the takeaways from the reviewed literature to ground the relationships the study examines. The current study reads, “The relationship between childhood victimization and adult relationship dysfunction is less clear” (p. 3), which is implied to be the overall gap that the study seeks to address. Since this appears to be the central premise on which the study is based, this point needs to be brought up much earlier. The lack of knowledge on this relationship should be explicitly stated earlier in the literature review. In addition, more elaboration is needed on the significance of this gap, beyond the fact that there has been little research on it. The authors could following this statement with a few sentences explicitly clarifying why this gap is problematic and what can be gained in, substantively, by gaining knowledge of this relationship and the path through which childhood victimization and adult relationship dysfunction among justice-involved women.

Response 2: We agree. The current study section (p. 4-5) has been edited to more clearly articulate the gap in the literature and its relevance to the population. This should also be better supported by the literature review.

Comment 3: Relatedly, the use of a sample of justice-involved women needs to be more strongly situated in the set-up of the research. It is unclear how the justice-involved nature of this sample fits in within the broader study. Does drawing on a justice-involved sample provide some specific value to investigating the connection between childhood victimization and adult relationship dysfunction, perhaps due to the documented associations between the study’s constructs (from the pathways perspective and relational model) and women’s offending? Another idea would be to frame the study around the need to support justice-involved women’s wellbeing as they transition out of their sentences. The authors allude to the need for therapeutic interventions for this population a bit already: “Treating issues of self-esteem and self-efficacy are central goals of therapeutic interventions for survivors and can promote improved interpersonal relationships in the future” (p. 3). Perhaps the case could be made that relationship dysfunction is common among justice-involved women, which may be due to their disproportionate exposure to childhood victimization. Then the authors could 1) bring up the need to understand this relationship as a whole, and 2) make the cause that understanding the mechanisms through which childhood victimization influences relationship dysfunction (i.e., potential roles of self-esteem and self-efficacy) can provide potential points for intervening.

Response 3: We agree. The current study section (p. 4-5) has been edited to more clearly articulate the gap in the literature and its relevance to the population. This should also be better supported by the literature review.

Comment 4: Methods:

In the methods section, the authors describe the basic characteristics of their sample, including demographics (age, race, education, marital status, motherhood) as well as prevalence of felony convictism and age of first offense. I think the prevalence of childhood abuse should also be clarified, since it is the study’s main independent variable. The details provided on childhood physical abuse among the sample are not entirely informative (i.e., measure ranging from 0-26 with a mean of 6.39). I would suggest the authors provide a general prevalence of physical abuse in the sample, and perhaps prevalence of the 13 types of physical abuse items for clarity.

Response 4: General prevalence of abuse was added (281-282). Prevalence of the 13 types of physical abuse were added to Table 1 (p.7) showing demographics of the sample.

Comment 5: Relatedly, the description of the abuse measure reads: “The current study pulled 13 items related to physical abuse experienced in childhood such as ‘spanked/slapped you’, beat you’, and ‘burned/scalded you’.” I was surprised that 13 types of abuse were mentioned, yet only 3 were provided as examples. The entire study is hinged on childhood abuse and its path to adult relationship function, so there is a need for much more information on the type of abuse examined and its prevalence. I realize that listing all 13 types of abuse in the measures description is quite wordy, so I would recommend the authors list the 13 types in Table 2 under the descriptive statistics of the overall measure. They could then provide a prevalence % of each of the 13 types.

Response 5: Agreed. This was not sufficiently clear. Information on how the items were chosen from the broader scale used in the pilot was added (p.8, 268-282). Specifically, a correlation analysis and exploratory factor analysis were conducted. A table was added with the factor loadings and a complete list of the items. Prevalence of the 13 types of physical abuse were added to Table 1 (p.7) showing demographics of the sample.

Comment 6: Did the study include control variables in the models? I was surprised that none were described in the measures section, since sample demographics and criminal histories were summarized in the sample description and in Table 1. If control variables were included, this should be explicitly stated and their measurement should be described. If no control variables were included, then this needs to be stated and the decision not to include them needs to be justified.

Response 6: The current study did not include control variables due to the specificity of the sample. This has now been noted and given context in the text (p.10, 332-337).

Comment 7: More detail should be provided on the analysis. All that is mentioned is that a path model was estimated for the relationship between child abuse, self-esteem and self-efficacy, and relationship dysfunction in adulthood. The authors should indicate the specifics of the path model, for instance the method of determining significance of indirect/mediating effects (there are a wide range of approaches for testing for mediation and a lot of debate as to which is most appropriate depending on data characteristics). The authors indicate that bootstrapping was used, but do not clarify how many bootstrapped draws were used. This is a key analytic detail that should not be omitted. In addition, were self-esteem and self-efficacy modeled together in paths, or were each examined separately?

Response 7: Additional detail and references were added to better articulate the analysis (p. 10, 326-343)

Comment 8: I was surprised that the authors did not provide any of the typical model fit statistics for path models: Comparative Fit Index (CFI), Tucker Lewis Index (TLI), Root Mean Square Root of Approximation (RMSRA). These should be provided, as they are essential in determining whether the path model actually fit the data.

Response 8: Thank you for spotting this oversight. Chi-squared, RMSEA and SRMR were added to Table 4 (p. 11). CFI and TLI were not included as they are a better representation of fit in comparison to another model. See (Schreiber et al., 2006).

Comment 9: One minor point:

There is some awkward phrasing in the abstract that should be modified: “The current study uses a path analysis to investigate the question of if for justice-involved women, is childhood physical abuse related to relationship dysfunction for women in adulthood and is this relationship mediated by self-esteem and/or self-efficacy?”

Response 9: This language was edited in the abstract and current study section (p. 5) to say “ Using a sample of justice-involved women, the current study uses a path analysis to investigate the impact of childhood physical abuse on adult relationship dysfunction and if this relationship is mediated by self-esteem and/or self-efficacy.”

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors were thorough in addressing all of my concerns and the revised manuscript has been strengthened. I recommend it be accepted for publication.