Hovering Is Not Helping: Relationships among Helicopter Parenting, Attachment, Academic Outcomes, and Mental Health in College Students

Round 1

Reviewer 1 Report

Thank you for the opportunity to review this manuscript, which examines associations between college students' perceptions of their mother's and father's helicopter parenting in a sample of 136 college students.  The paper is clearly written and reviews much of the relevant literature in this space.  Opportunities for improvement include:

1. Measurement. Throughout, the authors underscore that helicopter parenting/ overparenting is defined as developmentally inappropriate amounts of parental involvement, and indeed make a case that college would be a time when we would expect to see normative decreases in involvement and increased student autonomy. However, I found myself wanting more information about this measure (Schiffrin et al., 2019) of helicopter parenting, and how it operationalizes developmentally inappropriate involvement. I went to the citation, and it appears to include items from several other scales that include an element of student attributions about parental behaviors (i.e., their perception that the involvement is unwanted, intrusive, or excessive).  This is in contrast to some other measures that try and capture the frequency of objective behaviors (e.g. about whether or how much parents engage in behaviors like solving problems FOR the child).  The perceived excessive involvement approach is not without its merits, which could be highlighted here. However, there may drawbacks to this approach (i.e., perceptions of parenting are more likely to be influenced by negative attribution biases that often accompany internalizing problems).  Also measurement related- what were the timescales on all the measures? (i.e., past month/year? lifetime?). What are the implications of this timescale for your hypothesized direction of association (i.e., helicopter --> other variables)? For instance, might attachment to parent be primarily rooted in childhood and predate current overparenting behaviors? 

2. Related to the above point on developmental fit. I was surprised to see that there were not any covariates included in the regression models. In particular, I would have expected age to covary both with level of involvement and some of the outcomes which have a developmental course.  Gender, SES, and race/ethnicity would also be covariates to consider as potentially confounding results.  

3. Is there any literature to inform our understanding of potential mother/father differences on the front end? 

4. Methods: My understanding is that centering only reduces multicolinearity with an interaction term, not individual predictors, and thus this appears not relevant here. Not wrong to center, just not going to matter for multicolinearity.  

5. Results: I found the differences between the bivariate results and the regression analyses confusing. Why collapse measures into overarching constructs (e.g., internalizing) for one but not the other? Further, lack of bivariate associations might not necessarily preclude multivariate significant results (especially if other variables are included as covariates).  I would think that the bivariate section could be greatly reduced, to put more emphasis on primary multivariate analyses (with appropriate controls as mentioned above).  

6. The first sentence to the Discussion section seems to be setting up a literature review rather than this empirical article (i.e., "further explored the developing literature").  

7. In the discussion, the authors cite Padilla-Walker et al.'s 2021 LPA paper as support for the importance of dimensionality in helicopter parenting. I am afraid that this somewhat misrepresents the nature of that study's findings, which actually concluded that helicopter parenting was not the distinguishing feature of any of their extracted profiles, and rather that differences with outcomes seemed to be driven by parental control (which they conceptualized as distinct).  

Author Response

Thank you for taking the time to review the article – We found your comments insightful and helpful in improving the quality of the manuscript. We especially appreciate the comments regarding clearing up confusion about bivariate and regression analyses. We believe that we’ve addressed your concerns in this revision.

1. I found myself wanting more information about this measure (Schiffrin et al., 2019) of helicopter parenting, and how it operationalizes developmentally inappropriate involvement.

We added additional information about the Schiffrin et al. (2019) measure (see pp. 4-5) to provide more examples of how they defined developmentally inappropriate involvement.

1. The perceived excessive involvement approach is not without its merits, which could be highlighted here. However, there may drawbacks to this approach (i.e., perceptions of parenting are more likely to be influenced by negative attribution biases that often accompany internalizing problems).

We have now addressed this approach as a possible limitation in our discussion (see p. 10).

1. Also measurement related - what were the timescales on all the measures? (i.e., past month/year? lifetime?).

We have now added specifications regarding timescales into the respective measures paragraphs (pp. 5-6). The Motivated Strategies for Learning Questionnaire (MSLQ; p. 5) does not specify a timescale, instead asking respondents to indicate the extent to which each statement is true for them.

1. What are the implications of this timescale for your hypothesized direction of association (i.e., helicopter --> other variables)? For instance, might attachment to parent be primarily rooted in childhood and predate current overparenting behaviors?

This is a good note! We have now added this as a potential limitation within our discussion section (see p. 9).

1. Related to the above point on developmental fit. I was surprised to see that there were not any covariates included in the regression models. In particular, I would have expected age to covary both with level of involvement and some of the outcomes which have a developmental course.  Gender, SES, and race/ethnicity would also be covariates to consider as potentially confounding results.  

Thank you for this note. As requested, we have added several covariate control variables into the regression models. Namely, we added age and SES into step 1 of the regression model, controlling for these variables before adding helicopter parenting to the model.

1. Is there any literature to inform our understanding of potential mother/father differences on the front end?

We have added extra information from the literature regarding differences in mother and father helicopter parenting (see p. 9).

1. Methods: My understanding is that centering only reduces multicolinearity with an interaction term, not individual predictors, and thus this appears not relevant here. Not wrong to center, just not going to matter for multicolinearity.  

I believe you are correct about this; specific reference to multicollinearity has been removed from the manuscript (p. 6).

1. Results: I found the differences between the bivariate results and the regression analyses confusing. Why collapse measures into overarching constructs (e.g., internalizing) for one but not the other? Further, lack of bivariate associations might not necessarily preclude multivariate significant results (especially if other variables are included as covariates).  I would think that the bivariate section could be greatly reduced, to put more emphasis on primary multivariate analyses (with appropriate controls as mentioned above).  

We can see how this approach might be problematic and have opted to use the same variables in our reported bivariate analyses that are used in the regression analyses. We also reduced the emphasis on the bivariate analyses and added tables summarizing the regression analyses for further clarity (see pp. 6-8). We have thoroughly reworked how we discuss our analyses in response to your feedback.  

1. The first sentence to the Discussion section seems to be setting up a literature review rather than this empirical article (i.e., "further explored the developing literature"). 

This is a good catch! We have edited the language in this section to better suit the direction of the discussion.

1. In the discussion, the authors cite Padilla-Walker et al.'s 2021 LPA paper as support for the importance of dimensionality in helicopter parenting. I am afraid that this somewhat misrepresents the nature of that study's findings, which actually concluded that helicopter parenting was not the distinguishing feature of any of their extracted profiles, and rather that differences with outcomes seemed to be driven by parental control (which they conceptualized as distinct).

Thank you for bringing this to our attention. We have edited the discussion to more accurately reflect the conclusions from the referenced paper to avoid any confusion or misrepresentation. Thank you again for your help!

Reviewer 2 Report

The following paper, "Hovering is Not Helping: Relationships Among Overparenting, Attachment, Academic Outcomes, and Mental Health in College Students" extends our previous knowledge on overparenting by further examining the associations between overparenting, attachment, mental health, and academic-related variables. I felt the study provided sufficient empirical and theoretical support and the analyses were appropriate for the sample. Below, are my comments in hopes to improve the article's impact.

I think it is helpful to clarify that overparenting is synonymous with helicopter parenting, but I believe it may be clearer to stick to one term throughout the paper after this statement.

The two paragraphs going from pp. 2-3 between lines 81-106 could be revised for clarity and flow. For example, the last sentence of the first paragraph, "Finally, overparented students also reported lower levels of autonomy, competence, and feelings of relatedness [23]." could be incorporated when the authors are discussing research with similar associations in the beginning of the paragraph. I felt there were similar issues with flow in the paragraph after as well. 

As I was reading the participants section, I noticed that the authors stated the sample ranged "between the ages of 18 and 39" (Line 142, page 3). Since emerging adults are typically defined as 18-29 years old, can you provide more information and a rationale for the extended age range?

I believe it may be helpful to include a table that presents the results of the regression analyses.

I did not see any discussion of the gender-specific associations that were presented in the results.

I enjoyed reviewing the manuscript. Good luck with the paper!

Author Response

Thank you for taking the time to review our article! We have revised the document in response to your suggestions.

I think it is helpful to clarify that overparenting is synonymous with helicopter parenting, but I believe it may be clearer to stick to one term throughout the paper after this statement.

This is a very good point. We have decided to specify that the two terms are interchangeable in the beginning of the paper and have edited the rest of the paper only use the term helicopter parenting. We think that this will help with the flow of reading.

The two paragraphs going from pp. 2-3 between lines 81-106 could be revised for clarity and flow. For example, the last sentence of the first paragraph, "Finally, overparented students also reported lower levels of autonomy, competence, and feelings of relatedness [23]." could be incorporated when the authors are discussing research with similar associations in the beginning of the paragraph. I felt there were similar issues with flow in the paragraph after as well. 

We have revised the specific passage noted (see pp. 2-3) and have also numerous small edits to the introduction to increase flow and readability.

As I was reading the participants section, I noticed that the authors stated the sample ranged "between the ages of 18 and 39" (Line 142, page 3). Since emerging adults are typically defined as 18-29 years old, can you provide more information and a rationale for the extended age range?

The study was open to all students in the sampled classes, including those older than typically considered emerging adults. However, given our focus in the current study, we agree that it makes sense to omit these students from the analyses. Thus, we have now rerun analyses removing the those participants over the age of 29 from the sample.

I believe it may be helpful to include a table that presents the results of the regression analyses.

We agree that this could be a helpful way to depict the results, so we have added tables to p. 8.

I did not see any discussion of the gender-specific associations that were presented in the results.

We have added extra information from the literature regarding differences in mother and father helicopter parenting (see p. 9). Thank you for the suggestion!

Reviewer 3 Report

While this study did include a somewhat adequate literature review, there remains room for enhancing its overall quality. Notably, the author omitted any explicit mention of the study's underlying rationale or how it distinguishes itself from prior research in the field. The variables investigated in this study have already been the subject of extensive prior examination, and it would be beneficial for the author to elucidate the unique ways in which this study contributes to the existing literature. Moreover, the data analysis could benefit from significant improvement, with a strong recommendation to employ statistical equation modeling, meticulously illustrating individual paths and their combined effects

The English quality of this study requires minor revisions

Author Response

Thank you for taking the time to review our article! We have revised the document in response to your suggestions.

The author omitted any explicit mention of the study's underlying rationale or how it distinguishes itself from prior research in the field. The variables investigated in this study have already been the subject of extensive prior examination, and it would be beneficial for the author to elucidate the unique ways in which this study contributes to the existing literature.

We have added further discussion of the rationale to the introduction (see p. 3).

Moreover, the data analysis could benefit from significant improvement, with a strong recommendation to employ statistical equation modeling, meticulously illustrating individual paths and their combined effects.

In response to this request as well as another reviewer, we have made efforts in this latest revision to improve statistical control in the included regression analyses. Chiefly, we have done this by adding covariates to the aforementioned regression analyses. Although we understand that SEM can be helpful for testing more complex relationships among variables, we believe that regression is able to test our hypotheses without significant changes to our models. We hope that the revisions in this version of the manuscript will help.

Round 2

Reviewer 1 Report

Thank you for the opportunity to review this revised manuscript. The revision is stronger than the past draft, and largely responsive to suggestions. Areas for additional improvement remain:

1. Page 2: the "A previous investigation suggested that helicopter parenting is a related but distinct construct with an emphasis on overprotectiveness and overinvolvement that is not 44 apparent in other styles of parenting."- needs citation

2. Measures: All measures could still benefit from some more detail. For instance, how were the scores used in analyses computed? The ranges in the correlation table (e.g., 70) make me assume summary scoring, but this should be noted explicitly? Although not necessary, mean scores might be easier to interpret. Also, please include covariates (age and SES) in the measure section. How was SES measured? 

3. A remaining item from prior review: I would like to see, in the intro, whether there is any past literature to suggest that mothers' vs. fathers' helicopter parenting matters more/less for emerging adult developmental outcomes.  Did the authors have hypotheses about whether mothers or father would matter more? If not, then this should also be mentioned.  

4. Summary of Analyses: on Page 6, in the second paragraph (detailing hierarchical linear regression analyses): it is not clear that maternal and paternal helicopter parenting are sometimes in separate models (i.e., for maternal and paternal attachment) and sometimes included together (e.g., for peer attachment and other outcomes).  

5. It is a bit unconventional to have all the primary results nested under the heading "Summary of Analyses" which feels like more a part of the method than results. 

6. Results: Although I appreciate the inclusion of covariates in this revision, I don't know that the results need to feature these (largely non-significant) associations so prominently in text. You could probably just point the reader to the tables and say "covariates of age and SES were not associated with outcomes, as seen in Tables 2-4." and cut out a lot of extra text. 

7. Results: In the paragraph on Paternal Insecure Attachment, it states: "The full model with all independent variables entered was significant (R2 = .78, F (3, 118) = 3.343, p = .022)." .78 seems like a really high R2 value, especially given the size of the coefficients reported from that model. (it is also much larger than all of the other R2 values reported for the other outcomes). Might this be a typo?

8. For models where mother and father helicopter parenting were included together, it should be noted that the coefficients for each are OVER AND ABOVE the other. 

9. Discussion:

9. a This is much improved, but a few spots for potential improvement: The added paragraph on mother and father differences doesn't reference any past studies (other than Van Ingen) that have uncovered differential associations.  Is this because there aren't any? If so, this should be stated explicitly. 

9.b There remains quite a lot of causal language for a study in cross-sectional data. (e.g. "effect") For instance, "our study suggests  that such parental behavior may profoundly impact how emerging adults view their relationships with their parents"- I don't know that the size of the effects here warrants the word "profoundly", and further we do not know from these cross-sectional findings that helicopter parenting is shaping perceptions of attachment, and not vice versa (see point 9c below). Although the limitations section now mentions this point, the language leading up to this paragraph on page 9 tends to invoke a causal relationship such that helicopter parenting --> outcomes (rather than the reverse). I would recommend a close read for directional language (e.g., "predictors" "effect" rather than the more neutral and accurate "association"). 

9.c A related point: the final paragraph on Page 9 discusses "Similarly, it is possible that the issues with attachment observed in relationship to higher levels of helicopter parenting may actually have predated said parenting practices; in other words, children’s anxiety and avoidance of the parent-child relationship may have led to intrusive parenting practices as children grew older." this is one possibility, but it seems to suggest that parent-child attachment is a primarily child driven process, rather than being HEAVILY influenced by the parents' behaviors. I think this manuscript (and the field) would be helped by a more nuanced consideration of what this finding might mean. Could it not be that parenting behaviors in childhood that helped shape anxious or avoidant attachment (e.g., intrusiveness, inconsistency, harsh parenting) might be persisting in some ways or changing shape in emerging adulthood in ways that might look like helicopter parenting? 

9.d I found the final paragraph of the limitations (page 10) very confusing.  The first sentence asserts that a limitation of the present research was that it takes a unidimensional approach to understanding helicopter parenting (implying that a multidimensional approach, which captures sub-facets of helicopter parenting might be superior?), but then the rest of the paragraph talks about different profile analyses that do not seem to have suggested that a multidimensional approach is warranted.  Rather, those studies appear to have examined how helicopter parenting co-occurs alongside other parenting behaviors (which is another important point, but not the same as unidimensional measurement).  

Author Response

Thank you for your thought comments on this version of the manuscript! We have done our best to address the issues that you discussed and feel that addressing these has led to a stronger manuscript. We appreciate you taking the time to work with us – the point-by-point format of your comments also made revision quite approachable.

1. Page 2: the "A previous investigation suggested that helicopter parenting is a related but distinct construct with an emphasis on overprotectiveness and overinvolvement that is not 44 apparent in other styles of parenting."- needs citation

Unfortunately, I was unable to locate the article I had in mind when this sentence was added to the manuscript, so the statement has been omitted from this draft.

2. Measures: All measures could still benefit from some more detail. For instance, how were the scores used in analyses computed? The ranges in the correlation table (e.g., 70) make me assume summary scoring, but this should be noted explicitly? Although not necessary, mean scores might be easier to interpret. Also, please include covariates (age and SES) in the measure section. How was SES measured? 

We added this information to the subsections for each measure (see pp. 4-5). As you anticipated, scores were indeed summed. For this revision, we opted to keep the score summed rather than averaged. Further Information about the covariates was added to the end of the section (see p. 6).

3. A remaining item from prior review: I would like to see, in the intro, whether there is any past literature to suggest that mothers' vs. fathers' helicopter parenting matters more/less for emerging adult developmental outcomes.  Did the authors have hypotheses about whether mothers or father would matter more? If not, then this should also be mentioned.  

We apologize for the oversight in our response to prior review. We have now considerably expanded our coverage of gender differences between mothers’ and fathers’ helicopter parenting, adding an additional paragraph to the introduction (see p. 2). We have also updated our hypotheses section to indicate that although we were interested in seeing whether there were differences between helicopter parenting in mothers and fathers, we did not make any specific hypotheses about this prior to analyses.

4. Summary of Analyses: on Page 6, in the second paragraph (detailing hierarchical linear regression analyses): it is not clear that maternal and paternal helicopter parenting are sometimes in separate models (i.e., for maternal and paternal attachment) and sometimes included together (e.g., for peer attachment and other outcomes).  

We have added clarification to p. 6.

5. It is a bit unconventional to have all the primary results nested under the heading "Summary of Analyses" which feels like more a part of the method than results. 

Agreed – I think this was an artifact of fitting the manuscript to the template. We have now added revised headings to Results (see p. 6).

6. Results: Although I appreciate the inclusion of covariates in this revision, I don't know that the results need to feature these (largely non-significant) associations so prominently in text. You could probably just point the reader to the tables and say "covariates of age and SES were not associated with outcomes, as seen in Tables 2-4." and cut out a lot of extra text. 

We have removed much of the text about the covariates from the results section (see pp. 6-7), directing readings to the included tables and streamlining the text.

7. Results: In the paragraph on Paternal Insecure Attachment, it states: "The full model with all independent variables entered was significant (R2 = .78, F (3, 118) = 3.343, p = .022)." .78 seems like a really high R2 value, especially given the size of the coefficients reported from that model. (it is also much larger than all of the other R2 values reported for the other outcomes). Might this be a typo?

Good catch, thank you! We dropped a 0 here. The number should be .078 instead.

8. For models where mother and father helicopter parenting were included together, it should be noted that the coefficients for each are OVER AND ABOVE the other. 

We have revised the text in section 3.2 (see p. 6) to more clearly indicate this.

9. Discussion:
10. a This is much improved, but a few spots for potential improvement: The added paragraph on mother and father differences doesn't reference any past studies (other than Van Ingen) that have uncovered differential associations.  Is this because there aren't any? If so, this should be stated explicitly. 

We further added to this section, including a few more references to studies for comparison (see p. 10).

9.b There remains quite a lot of causal language for a study in cross-sectional data. (e.g. "effect") For instance, "our study suggests  that such parental behavior may profoundly impact how emerging adults view their relationships with their parents"- I don't know that the size of the effects here warrants the word "profoundly", and further we do not know from these cross-sectional findings that helicopter parenting is shaping perceptions of attachment, and not vice versa (see point 9c below). Although the limitations section now mentions this point, the language leading up to this paragraph on page 9 tends to invoke a causal relationship such that helicopter parenting --> outcomes (rather than the reverse). I would recommend a close read for directional language (e.g., "predictors" "effect" rather than the more neutral and accurate "association"). 

We understand this criticism and have attempted to reduce the use of causal language throughout the text in many spots (for examples, see the referenced passage that formerly suggested “profound impact” on p. 9, and also changes to the last paragraph on p. 9).

9.c A related point: the final paragraph on Page 9 discusses "Similarly, it is possible that the issues with attachment observed in relationship to higher levels of helicopter parenting may actually have predated said parenting practices; in other words, children’s anxiety and avoidance of the parent-child relationship may have led to intrusive parenting practices as children grew older." this is one possibility, but it seems to suggest that parent-child attachment is a primarily child driven process, rather than being HEAVILY influenced by the parents' behaviors. I think this manuscript (and the field) would be helped by a more nuanced consideration of what this finding might mean. Could it not be that parenting behaviors in childhood that helped shape anxious or avoidant attachment (e.g., intrusiveness, inconsistency, harsh parenting) might be persisting in some ways or changing shape in emerging adulthood in ways that might look like helicopter parenting? 

I think that we are in agreement about this – We also think that it is most likely that the issues with attachment we found in this study stem from parental behavior. In this passage we were trying to acknowledge that there’s some possibility that the relationship functions differently, but we’ve added some updated text that hopefully makes our position clearer.

9.d I found the final paragraph of the limitations (page 10) very confusing.  The first sentence asserts that a limitation of the present research was that it takes a unidimensional approach to understanding helicopter parenting (implying that a multidimensional approach, which captures sub-facets of helicopter parenting might be superior?), but then the rest of the paragraph talks about different profile analyses that do not seem to have suggested that a multidimensional approach is warranted.  Rather, those studies appear to have examined how helicopter parenting co-occurs alongside other parenting behaviors (which is another important point, but not the same as unidimensional measurement).  

We can see that our use of the term “unidimensional” was confusing. What we really meant to say is that our study was somewhat limited by not investigating dimensions such as warmth and perceived control as separate variables. We hope that the revisions to this paragraph (see p. 10) will improve this section.

Reviewer 3 Report

The authors addressed my comments adequately

Author Response

Thank you for your help with the manuscript!

Round 3

Reviewer 1 Report

I appreciate this round of revisions and have no further comments to add. Interesting work!