Review Reports - “Everything Sounds Great on Paper”: Drivers of Housing Instability as Youth Transition out of Foster Care

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors are bringing into light (really reminding the field given what is known about factors contributing to successful transitioning out of foster care) an important component necessary for transitioning successfully out of foster care. Overall, the authors have presented the rationale, methods, results, and discussion in a clear manner. The use of mixed methods is a plus and the inclusion of social network component is an important addition to the study of transitioning successfully out of foster care. The following suggestions are meant to further enhance the strength of the study.

Please specify the theoretical framework or conceptual framework (if applicable) guiding the study. In addition to the research basis for exploring the influence of social network, additional theoretical basis tied to research would be important to note.
Methodology: The qualitative analytic method and rationale for its selection that guided coding, development of themes, etc should be named and described.
The reader can infer that some strategies of rigor were applied (i.e., consensus building around codes). If any other strategies of rigor were attempted, they should be explicitly stated.
Discussion of Limitations should address the limitations of the analytic methods used, if any. The discussion of "housing capable" should be amplified in the discussion as well as not only the number/size of social network but the quality of those network connections.

Author Response

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions highlighted in the re-submitted files. We appreciate the detailed feedback and believe that the requested revisions have strengthened the manuscript. We have copied each comment followed by our response which is indented and in italics including references to specific page numbers where changes can be found. We are hopeful that with these revisions, the manuscript is suitable for publication in this exciting special issue.

Response to Reviewer 1 Comments

Please specify the theoretical framework or conceptual framework (if applicable) guiding the study. In addition to the research basis for exploring the influence of social network, additional theoretical basis tied to research would be important to note.

Thank you for this feedback and for pointing out that the theoretical framework was not clearly described. We have added a paragraph on p. 3, lines 110-118 that references the socioecological framework and the social network theoretical perspectives that inform this study.

Methodology: The qualitative analytic method and rationale for its selection that guided coding, development of themes, etc should be named and described.

Thank you for pointing out that this was not clearly stated. We have added additional description of our study design and rationale on p. 4, lines 147-158. And, we added the following at the beginning of the data analysis section on p.5, line 215 to more clearly describe our analytic method: “We used a team-based consensus coding process that drew on grounded theory and employed case summaries and data matrix displays to facilitate identification of patterns and connections across cases [18].”

The reader can infer that some strategies of rigor were applied (i.e., consensus building around codes). If any other strategies of rigor were attempted, they should be explicitly stated.

We appreciate this feedback and have added the following on p. 6, lines 241-243 to more explicitly describe our strategies to enhance rigor: “We specifically engaged in the team-based consensus process and kept an audit trail with team notes and individual reflections in order to enhance the rigor of our analysis.”

Discussion of Limitations should address the limitations of the analytic methods used, if any.

We have added a limitation of our analytic methods, specifically the matrix approach which involves data reduction. This is highlighted on p.14, lines 603-605.

The discussion of "housing capable" should be amplified in the discussion as well as not only the number/size of social network but the quality of those network connections.

This is an important point and we have added several sentences on p.14, lines 585-593 to explicitly highlight the importance of the quality and not just the quantity of network connections, particularly the importance of housing capable adults.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript presents findings from a mixed methods study of transition-aged youth (TAY; aged 18-25) in Texas regarding factors precipitating housing instability when youth transition out of foster care.

Strengths:

Overall well-written and organized
important topic : ongoing problem of TAY and high rates of housing instability
Makes the case for low- barrier housing options for transition age youth, probationary periods to support youth towards independence – Supervised Independent Living programs

Areas to strengthen:

Introduction, pg 3: More info needed here — in what way were the prior studies limited and how does your study address these limitations? Is there no prior research on the housing context in prior studies that could be generalized to foster youth in general? If not, be explicit. Drivers of housing instability? Factors contributing to remaining housed? Role of extended care? Role of social network —not clear what the value added is for this study
Materials and Methods, pg 3: More info needed on methodology; QUAL + Quant? Were data collected simultaneously? How were qual and quant data integrated?
Materials and Methods, pg 4: Rationale for not using CW admin data re: placement status, # of placements/homes, # of years in care?
Materials and Methods, pg 4: Given the emphasis on limited social network measurement in prior studies, it would be important to articulate here how operationalization of social network is better than prior studies
Materials and Methods, pg. 5: Why n = 25? Was this the number that reached saturation? Or the total # of participants volunteering?
Materials and Methods, pg 5: When coding, how many cases had 2 coders? What happened to those cases with no overlapping coders? What was interrater reliability assessed? How many coders int total?
Results: make sure to put participant responses in quotations – some confusion regarding which set of texts are exemplars and which is the narrative.

Author Response

Response to Reviewer 2 Comments

Introduction, pg 3: More info needed here — in what way were the prior studies limited and how does your study address these limitations? Is there no prior research on the housing context in prior studies that could be generalized to foster youth in general? If not, be explicit. Drivers of housing instability? Factors contributing to remaining housed? Role of extended care? Role of social network —not clear what the value added is for this study

We appreciate this feedback and have worked to more explicitly state the limitations of prior research and how our study specifically contributes. We have added language about the limitations of prior social network studies on p 2, lines 74-77. And added additional details about how our study adds value in the paragraph describing the current study on p.3-4, lines 119-144.

Materials and Methods, pg 3: More info needed on methodology; QUAL + Quant? Were data collected simultaneously? How were qual and quant data integrated?.

Thanks for pointing out that this was not clearly described. We have added additional detail about our methodology which is highlighted on p.4, lines 147-158. We specifically provide detail on the QUAL + Quant design where data were collected simultaneously and how the two data sources worked together. In the analysis section, there was already some information about how the qual and quant data are integrated specifically, see p. 5, lines 220-224.

Materials and Methods, pg 4: Rationale for not using CW admin data re: placement status, # of placements/homes, # of years in care?

The use of administrative data was not feasible or necessary in our study where our focus was to understand the experiences and perspectives of young adults. Gaining access to administrative data would have required additional permissions from the child welfare administration and we believed that participant self-report was actually valuable in understanding how the young adult saw their experiences. The use of self-report measures is an acknowledged limitation and we added additional language to acknowledge this specifically in relation to child welfare experiences on p. 14, lines 603-604.

Materials and Methods, pg 4: Given the emphasis on limited social network measurement in prior studies, it would be important to articulate here how operationalization of social network is better than prior studies

Thank you for this feedback. We have added the following statement to the measures section on p. 5, lines 199-203: “This approach allowed for more extensive analysis of social networks than in prior studies, allowing for visual depictions of the networks to understand how connected network members were to each other as well as identify key members of the network that were central and bridged different groups within it.”

Materials and Methods, pg. 5: Why n = 25? Was this the number that reached saturation? Or the total # of participants volunteering?

We appreciate this important question. The sample size target was set based on our analysis approach which was based on the grounded theory inductive process and the diversity of the experiences that we wanted to represent. We added a statement to clarify our approach to sample size on p. 4, lines 155-158.

Materials and Methods, pg 5: When coding, how many cases had 2 coders? What happened to those cases with no overlapping coders? What was interrater reliability assessed? How many coders int total?

Thank you for this specific feedback. We have provided more detail about the number or coders and the number of cases at each step of our process in the analysis description on p. 5, lines 215-243. We did not assess interrater reliability in the case abstraction process, however, the team felt confident following our analysis of the first cases that we had a strong process with clear consensus.

Results: make sure to put participant responses in quotations – some confusion regarding which set of texts are exemplars and which is the narrative.

We agree that this was confusing. We are not sure on the standard conventions for presenting longer quotes but have gone ahead and formatted all quotes to have quotations and be presented as part of the paragraph in which they appear rather than having some that are in offset paragraphs without quotes.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have responded to the comments and feedback with specificity and clarity which has enhanced the strength of the manuscript. The study will contribute to the body of existing research in a meaningful manner.

Reviewer 2 Report

Comments and Suggestions for Authors

None