This paper investigates the developmental trajectories for adolescent-to-mother violence by analysing administrative data using machine learning techniques. Although this is a novel approach to investigating the factors related to adolescent-to-mother violence, there are several limitations to this manuscript.

Introduction: The narrative for this introduction needs clarifying. This study has the potential to significantly add to the literature, but the introduction seems to focus on whether abuse against mothers/fathers should be explored, touching on feminist theory and evidence related to other forms of family violence. To me, this is a missed opportunity to explore the current evidence regarding the longitudinal relationship between witnessing, experiencing, and perpetrating abuse, as well as the relationship between mental health and parental abuse. Particularly as this is a nursing journal, this feels like not discussing the link to mental health is an oversight as this may be the most relevant aspect for nurses to consider. I would explore both literature that explored mental health and adolescent to parent violence as well as problem behaviours in first episode psychosis/other mental health issues (if you have space).
While I am not privy to the publication plan for the project as a whole, I am somewhat uncertain why this paper was framed as a paper on adolescent-to-mother violence when it could have easily discussed both parents, but separately. Indeed, in the last bit of the results fathers were discussed to help contextualise the similarities and differences with mothers. Given that you have access to data from both parents and that 37% of the sample are fathers, it seems strange that the focus remained on mothers.

Method: Proof read to ensure the correct use of past tense throughout.
It would be useful to describe the type of data collected in more detail. For instance. What type of police data was collected? Is it any contact with police? Just family violence incidents? If it is just family violence incidents, what counts as an incident in NSW? Family violence police differs significantly in Australia compared to North America and Europe.
I thought that you provided a reasonably good explanation of the health data, but I am wondering if you used counts, admission time, etc. More specifics would be much appreciated.
I am new to the idea of process mining, so more knowledgeable reviewers should be sought to comment on the statistics. However, it may be useful to describe the procedure in greater detail in the method. Are there statistics that indicate goodness of fit or reliability? How do we know how well these models are performing? For novel statistics, it is useful to provide the reader with guidelines of how to interpret the statistics (e.g., X is a small effect size, y is a moderate effect size, and z is a larger effect size).

Results: The sentence, “Proportionally, girls (64%, n = 155) engaged in family violence towards their mother at a higher rate than boys (62%, n = 331). “ concerns me, as it does not appear that an inferential test was calculated (i.e., a chi-square) but the authors stated that girls were more violently than boys. Please include the statistical tests, included effect sizes, when comparing boys/girls or mothers/fathers.
It seems strange that you direct the reader to a table comparing violence to mothers and fathers, but you do not mention it in the results.
Please include standard deviations every time a mean is stated.
Statistically, you would assume that exposure to P-IPV was father-to-mother violence, but did you check this? Or is it just any incident between parents.
For families where P-IPV occurred, what was the mean number of incidents?
Where did you get the parental drug use variable from? Was that from the police or the health dataset? If it is police, do you know how the police rated this item? (e.g., self report, appeared intoxicated at the incident). If it is health, same question as above. This should be addressed in the method, did you assume that if a variable was mentioned in either dataset that it was true? Did you look for consensus between datasets.
Do you have information regarding the child physical abuse variable? Do you know if it is physical abuse by mom or dad?
In your results, the focus sometimes shifts between children to mothers (e.g., 44% of mothers had adolescents involved in police recorded offences). Is that just 44% of the sample? If so, I would keep the focus on the adolescences who are the subjects of your research, not on the mothers, when discussing the results.
Reading the paragraph, “events preceding adolescent-to-mother violence’ made me realised that ‘violence’ needs to be operationalised in the method, as it seems that you are separating it from police report or verbal conflict.
Do you have statistics on what the diagnoses tended to be? This would be useful for nurses.
How many adolescents were excluded from the analyses with both parents?

Discussion: Given the limitation of the journal in regards to the number of references that can be included in this manuscript, I appreciate that you cannot cite many papers. However, I would have liked to see better integration of your research into the broader context of literature related to adolescent to parent abuse, emerging mental health issues, the relationship between mental health and offending.

References: Please fix the formatting of the references

Tables: Please adhere to APA formatting
Reading the tables, I have concerns that you did not adjust for multiple comparisons. Additionally, I would also add Odds Rations to help interpret the ChiSquare effect size
The figure is really interesting, but it would be helpful to the reader if you walked through how to interpret it given the lines appear to be different widths and following different directions. Also, what does “instant” mean in the context of the diagram?
Overall, I think that there is merit in this paper and I am intrigued by the results but the manuscript requires considerable work

Overall, the authors did a great job at responding to the concerns raised in the original manuscript. However, there are a few small suggestions that I would like to make.

Introduction
– use of semi-colons at the start of a list, when it should be a colon
-“Nine-one per cent of index cases” appears to be a typo
Method
- Would like to see the chi-square test in the following statement, "Overall, mothers were more likely to be victims of this family violence (63% n = 486), with mother only violence occurring at 3.3 times the rate of father only violence."
- "First offences against a mother predominantly involved an act of violence (57%, n = 276; violence: 45%, n = 218, serious violence: 12%, n = 58)." I would suggest coming up with something other than ‘violence’ for the less-severe violence category, as this sentence is rather confusing.
Discussion
– it would be useful for the first 1-2 paragraphs of the discussion to summarise the results of the study, rather than talk about the strength of the study.

Thank you for making the final changes. Well done on this paper.