Exploring the Decisional Drivers of Deviance: A Qualitative Study of Institutionalized Adolescents in Malaysia

Round 1

Reviewer 1 Report

The article was very clear in reading without rising questions about its development and rational thinking. The sample was small but the development covered it. Qualitative research is not based on heavy data but in deep understanding of the facts. I saw a very good abstract that attracted me to read the paper because the authors gave all the information a scholar may need. The structure was very well developed and the paper deserves to be published as it is.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

First of all, I would like to congratulate the authors for their hard work. The work seems to me relevant and necessary and addresses a field that needs to be updated. In addition, it includes information that can be very valuable for the scientific community in general, and especially for child and adolescent’s psychology. However, there are some minor improvements to be made, which I describe below.

• I think the introduction needs to be reorganised, there are multiple ideas that appear on multiple occasions in a disjointed manner. 
  It is necessary to include objectives and hypotheses that can really be answered with the work. 
• Authors should address in the introduction all variables to be analysed later.
• For me it is important to know what kind of professionals conducted the interviews and whether or not they were trained to do so. 
• In the participants section you need to expand the information about the participants: how many were there in the end, what was their gender, what were their ages (range, mean, standard deviation...)? I suggest moving the information from the results section to the participants section.
• I consider the lack of inclusion and exclusion criteria to be a major limitation of the study... As the authors are aware, these aspects could seriously bias the study. The type of offence, age, recurrence of the offence, presence or not of serious mental health problems, emotional stability, time since the offence... This aspect should be expanded.
• The extraction process is not clear to me, it is normal that at least two blind judges would have been involved. If so, the authors should calculate the inter-judge agreement rate.
• If the authors have followed criteria of rigour and credibility, this should be pointed out. For example (I quote an article by Póstigo and collaborators from 2019, http://dx.doi.org/10.6018/analesps.35.2.301201) In terms of the rigour and credibility criteria proposed for qualitative studies, triangulation of researchers (two researchers), methods (sociometric and qualitative), coding (two analysts), and theoretical triangulation (review and discussion of previous studies) were used, which increases the reliability and validity of this study (Denzin & Lincoln, 2008; Twining, Heller, Nussbaum & Tsai, 2017).
• It would be appropriate to assess the rate of agreement between respondents using the Kappa index and Cohen's.
• Cohen's Delta index, and to show whether there was adequate coding reliability or not (the following references can be consulted in this regard: Cohen, J. (1960). A Coefficient of Agreement for Nominal Scales. Educational and Psychological Measurement, 20, 37-46. Femia, P., Martín, A., & Álvarez, M. (2012). The Delta Model to evaluate the degree of agreement between two observers [El Modelo Delta para evaluar el grado de acuerdo entre dos observadores]. In García, Bouza, & Covarrubias (Eds.), Modelación de fenómenos del Medio Ambiente, Salud y Desarrollo Humano: estudios medio ambientales. (pp. 125-144). Universities of Guerrero (Mex) and Havana (Cuba)).
• Were the evaluators blinded during the process? This would be a strength of the study.
• How did the authors control for confounding variables? TThere may be notable differences between some of the socio-demographic factors that the authors assess (and others that the literature has pointed out). I think the direction of the work could be improved by statistical comparisons of the results according to these variables. For example: do the themes differ according to the group of origin? Contingency table or chi-square analyses could be carried out.
• Authors should address in the introduction all variables that are to be analysed later on.
• The limitations of the study need to be significantly expanded.
• The article should end with a brief conclusion, including the findings and implications of the study.
• Citations and references must be up to date.
• The entire manuscript should be in the same typeface (check figures).

  Again, I thank the authors for addressing this issue.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Thanks to the authors for their changes and clarifications. I think it would be necessary to address all the variables that are subsequently analysed in relation to the hypotheses and results. For example, if gender is expected to be an important or analysed variable, information from previous literature should be put in the introduction. 
I suggest calculating the indices I mentioned (at least Kappa), they are fundamental to give more rigour to the article. 
I suggest again:

How did the authors control for confounding variables? There may be notable differences between some of the socio-demographic factors that the authors assess (and others that the literature has noted). I think the direction of the work could be improved by statistical comparisons of the results according to these variables. For example: do themes differ by a group of origin? Contingency tables or chi-square analyses could be performed.

Thank you

Author Response

Please see attachment.

Author Response File: Author Response.docx