2/2/2023: Daniel O'Brien

[GabeNicholson]

Comment below with a well-developed question or comment about the reading for this week's workshop.

If you would really like to ask your question in person, please place two exclamation points before your question to signal that you really want to ask it.

Please post your question by Tuesday of the coming week, at 11:59 PM. We will also ask you all to upvote questions that you think were particularly good. There may be prizes for top question-askers.

[sdbaier]

Professor O'Brien, thank you for sharing your research on how violent crime persists across multiple geographical scales.

You mention the key role addresses play, “[...] accounting for half of the main effects and presenting in over 80% of the positive interaction effects. Streets were then moderately more impactful than neighborhoods, featuring more main effects and having twice as many interactions with addresses. This indicates that persistence and aggravation operate at each level, but most often through addresses and the interactions they have with their broader context. […] The predominant importance of addresses for both persistence and aggravation made them the heart of the proverbial “action”. […] If the underlying conflict is isolated to a few individuals, the subsequent violent events might be expected to be confined to their private spaces. The results largely support these localized interpretations.” (O’Brien and Ciomek, 2022)

I am curious if this finding holds for cases where the perpetrator is detached from the lowest scale of analysis. Put differently, does crime persist to cluster at an address level in cases where the perpetrator is removed (i.e., jailed or imprisoned) following the reported offense? I am assuming that much of the clustering of crime at a particular residential address boils down to one or a group of focal actors repeatedly engaging in crime (examples of private disorder and residential conflicts, or abandoned houses used to store contraband in the illegal drug trade: i.e., same location, same actors), more so than the case of different actors by chance all get criminally active in the same location (example of repeated robbery of the same house; i.e., same location, different actors). This would imply that for the fraction of violent crimes where perpetrators are arrested, crime should (at least temporarily) subside at an address level. Perhaps you have considered this while designing the study, and it might not even be feasible given the currently used archival data of Boston, but I would be curious to hear your thoughts.

[pranathiiyer]

Thanks so much for sharing your work with us, Professor. I'm not very familiar with literature around aggravation and criminology but I was trying to think about your taxonomy more broadly. For instance, you define reinforced aggravation to occur when one form of disorder or crime predicts another in the following year. What would we define to be disorder in this case? For example, it is possible that the pandemic led to unemployment which led to crimes for consecutive years in a neighbourhood. Could we redefine this and other such instances to be captured by reinforced aggravation or other metrics in your taxonomy? Thanks!

[JoeHelbing]

Professor O'Brien,

Am I correct in understanding that the analysis of address level data contradicted the expectations of Broken Window Theory? I may have misunderstood the analysis in this respect but would appreciate if you might expand on the connection to this as I remember when this was a core and somewhat controversial principle in New York policing.

Thank you

[yuzhouw313]

Hello Professor O'Brien, Thank you for sharing your research on the topic of the persistence and aggravation of crime and disorder! I am not familiar with the field of criminology, but I did assume that neighborhood as a geographic scale contains some level of homogeneity as you mentioned that should be debunked. I was wondering that, while crimes may persist at any of the scales you mentioned, how would you discompose the geographic unit to gain a more detailed perspective? Thank you again and look forward to meeting you on Thursday.

[Yuxin-Ji]

Hi Professor O’Brien, Thanks for sharing your research with us. It is interesting to learn about quantitative methods in investigating urban crimes. I am not familiar with this field, but I can see that it is a very meaningful topic and its academic and social value. I noticed that the data being used here is the archive of reported cases, but there might be many other cases unreported. I am curious if there are any study in the bias in reported vs. unreported crime cases, and how such bias could further affect crime patterns?

[Sirius2713]

Hi Professor O'Brien,

Thanks for sharing your research with us. It's very intriguing to read your work on criminology and how addresses, streets and neighborhoods impact the persistence and aggravation of violent crimes. May I ask if you have come across any evidence in your long-term data that suggests a connection between gentrification and violent crimes? For instance, in areas that were previously considered "violent" but have undergone gentrification in recent years, how does gentrification affect the occurrences of violent crimes in those neighborhoods?

[lbitsiko]

Dear Prof. O'Brien,

Many thanks for presenting your work in our workshop. Even though my knowledge of criminology is extremely limited, I found your analysis really interesting. I would appreciate your input on some thoughts I had while reading the paper. I understand that you are trying to predict crime in terms of geographic concepts. I might be mistaken but I feel there is a certain causal relationship implied between geographic concepts and crime. Isn't the geographic element a mediating factor between crime and other factors (such as socioeconomic status and segregation)? Furthermore, I wonder whether the predictions actually reflect an increase in crime awareness due to targeted policing in areas predicted to have high crime rates. Looking forward to your talk on Thursday.

Kind regards, Loizos

[iefis]

Hi Professor O’Brien, Thank you for sharing your work with us. One of the main findings of the research is about how violent crime in a neighborhood manifests itself through the most vulnerable addresses. Building on the cross-scale interactions observed, I'm curious about if there is any heterogeneity of such interactions among neighborhoods (and streets & addresses) with different social economic status (SES). In other words, how would SES interplays with the types of manifested vulnerability in different neighborhoods? Thanks.

[shaangao]

Hi Professor O’Brien, thank you so much for sharing your work with us! I'm new to criminology and would like to know more about the potential mechanisms through which addresses, streets, and neighborhoods interact to perpetuate crimes, e.g., the role of geography, history, SES, etc.

[XTang685]

Hi Professor O’Brien, thank you for sharing your work with us! Your study examines the interdependent effects of addresses, streets, and neighborhoods in driving the persistence and aggravation of crime over time. Using a database of records generated by 311 and 911 systems in Boston, MA, the study tests the independent effects of persistence and aggravation at each geographical scale and the cross-scale interactions between these effects. My question is what are the current study's contributions to the study of crime in communities?

[JunoWuu]

Hi Professor O’Brien, thank you for sharing your great work with us! I think the findings can be really important for policymakers and governments. I wonder what applications or what directions the results may imply and how can we actually use them to maybe target the persistence of crime.

[AlexPrizzy]

Thanks for sharing this work in out workshop, Dan! My question is how do these findings generalize across urban designs? Is there an effect of the urban design of Boston, where the street layout is quite inconsistent and confusing, on the outcomes found here? How would the interplay of violence at the address, street, and neighborhood levels vary for a city like Chicago which has a perfect grid system?

Intuitively I'd expect the degree of diffusion of crime across levels to vary between cities like Boston and Chicago. Boston having many winding roads may limit communication between neighborhoods. While Chicago's grid system may allow for greater spread of crime across neighborhoods by directly linking them via streets. Considering there are major streets which run straight from one end of the city to the other, crossing through many neighborhoods varying in socioeconomic and crime levels.

[secorey]

Hi Dr. O'Brien, Thank you for your presentation. With the results of this research in mind, do you think that a large part of this pattern is in how cities are structured? If so, does this reveal any ways to improve city structure? I would imagine that planned cities could find this information very valuable.

[taizeyu]

Dear Dr. O'Brien,

It is really an excellent article. Thank you for your presentation. I am curious that how this research can help communities grow in the future.

[javad-e]

Thank you for presenting your research at our workshop, Prof. O'Brien! The results are very interesting and have profound policy implications. I was wondering to what extent you think the results are generalizable to other cities. In particular, do you think factors such as more significant income segregation at the neighborhood level (e.g. in Chicago), lower population density (e.g. in Houston), and non-uniform distribution of urban facilities (e.g. in Los Angeles) compared to Boston can result in different outcomes? In other words, do you think the results might depend on the spatial distribution of wealth, population, and amenities in the city?

[zhiyun0707]

Hi Professor O'Brien!, thank you for sharing your work with us! In this paper, you primarily focused on the Boston area violent crime, I wonder that how will the geographical factors impact the result? (In other cities, e.g. Chicago)

[borlasekn]

Prof. O'Brien, thank you for sharing your work with us! As someone who studies comparative human development, family sociology, and education, your work is especially pertinent. In particular, I think the persistence of crime across time is especially relevant. I've seen studies that talk about failures to address mental health concerns that come from violent or traumatic upbringings and the contribution to the cycles of violence that so many communities seem to be stuck in. As you stated in your own research, results from these studies "have been mixed at best potentially unrelated to the elimination of disorder". I had a question about your approaches as a researcher when studying these topics, given that your work has direct impact to community efforts, including policing: How do you prepare yourself to be not only a bias informed researcher but a trauma-informed researcher, especially when dealing with such important community topics?

[tangn121]

Hello Professor O'Brien, thanks for presenting your research to us! Your paper analyzed data from Boston between 2011 to 2016. In light of the pandemic, do you anticipate any impact on your results or are there additional factors that need to be considered?

[yhchou0904]

Dear Prof. O’Brien, Thank you for sharing your idea with us! The relationship between crimes and locations is complicated and involves the context of cultural and historical background. With the analysis across different geographic scales and interactions, the research provides an overview of how communities are related to the occurrence of crimes. What I am wondering is that, since persistence was common at all scales, showing where the crimes happen is important, is it possible to break this persistence? If yes, does it requires certain characteristics of a community?

[Ry-Wu]

Hi Professor O’Brien, Thank you for sharing this amazing research! I find the multilayered complexity really interesting and the research provides great insights into the interplayed factors and the influence on environmental justice. Looking forward to your talk tomorrow!

[nijingwen]

Hi Professor O'Brien Thank you for sharing you idea. it is great to make relationship between crime and locations. looking forward to hearing from you tomorrow

[BaotongZh]

Dear Professor O'Brien,

I wanted to extend my gratitude for presenting such an insightful paper. May I inquire as to how the study suggests that practical steps can be taken in addressing issues in communities and places?

[zoeyjiao1104]

Hi Professor O'Brien,

Thank you so much for sharing your work. I was wondering about the unreported crime cases. Do you think these cases would potentially lead to bias in your study? In addition, what would be the applications or practical values of your studies in real life. Thanks!

[lguo7]

Hi Professor O'Brien, Thank you for sharing this meaningful and fascinating research. I learned about spatial geography in my previous course spatial data science. I think that the concept of geography in your study to predict crime is exciting. I would like to know if there is a crime that is continuous, say moving from one neighborhood A to another more distant neighborhood B, how do you handle this aspect of the data?

[fiofiofiona]

Hi Professor O'Brien,

Thank you for presenting your work. I am curious how the demographics (e.g., average income, education level, gender and race composition) of address, street, and neighborhood would influence the violence persistence or aggravation. Also, since the data is obtained from the 911 system, do you think police resources in an area would play a role in violence aggravation?

[zhiqianc]

Hi Professor O'Brien, thanks for your sharing. My question is: what do you think is the limitation of this kind of multilayered complexity of social geography?

[xin2006]

Hi Professor O'Brien,

Thank you for presenting your work. I am curious about the disorder and crime cases. First, in terms of the type of cases, is there any overlapping in the definition of indices? And I am also wondering if the reported cases in the Parcels, Streets, Tracts is underestimate, especially some cases related to private conflict? I think that some conflicts happened in house will not be reported to police.

[xiaowei-v]

Hi Professor O'Brien, thanks for sharing! I realized there are factors related to neighborhood that might contribute to the results, but as you noted, the dynamics between disorder and crime vary in different neighborhood. My question is that how should future work address this issue.

[DehongUChi]

Hello, Professor O'Brien, thank you for sharing your work with us on how violent crimes persist and aggravate on multiple geographical levels. I am curious if the formation of streets and neighborhoods are changed, in particular, neighborhoods, would change the findings of the study, since the determination of neighborhoods are rather arbitrary in some cases.

[LuZhang0128]

Hi Professor O'Brien, Thank you for sharing this amazing study with us! Since you are using the recorded cases, I wonder if the unreported cases will be significant and thus bias the results. If so, to which direction would you expect the bias to be? Thanks!

[yutaili]

Thanks for sharing your work on criminology and how crime aggravates and persists at streets and neighborhoods. I wonder how do you consider the crime that has not been reported in this case? And do you think the broken window theory contributes to the persistence of crime in some neighborhoods? Looking forward to your talk tomorrow.

[ChongyuFang]

Hello Prof. O'Brien, Thank you for sharing your fantastic research with us. I'd like to know how your research can be applied to other US cities? Look forward to your talk tomorrow!

[hazelchc]

Hi Professor O'Brien, Thank you for sharing your interesting piece of work with us! I'm curious about the policy implications of the paper. What does it imply when action is greatest at addresses? Do you think similar results may also be found in other cities? Thanks!

[Hongkai040]

Hi Professor O’Brien, Thank you for sharing your work with us. This work is especially meaningful thinking of shocking violent crimes that happened recently. I'd be more interested in the dynamics of how different forces shape violent crimes. For example, is there any way to incorporate police patrolling, local income, etc., into the framework, analyzing how they influence violent crimes at different levels ?

[linhui1020]

Hi Professor O'Brien, Thanks for presenting your work in our program. I am curious about additional factors that may impact the outcome variable of interest. How this methods are advantaged compared with other methods? How unreported crime rates would influence the crime pattern?

[YLHan97]

Hi Professor O’Brien, Thank you for sharing your work with us. Here are some questions after reading Well-Being and Equity in Communities and their Places:

• What is the relationship between well-being and equity in communities and their places, according to the author?
• How does the concept of pointillism relate to the design and planning of cities?
• How does the author define the connection between community spaces and the well-being of individuals?

[zihua-uc]

Hi Prof. O'Brien,

I wonder if there are other interesting factors that could affect crime, such as how homogenous the neighborhood is in terms of race, and religion, or the average socioeconomic status in the neighborhood.

Furthermore, crime is an interesting variable to study, but here the paper defines crime as whether crime has occurred, or as counts. However, a gun shooting incident that kills 20 people is very different from an armed robbery with no casualties. It would be interesting to see what results show when we take into account the severity of crime.

[yujing-syj]

Hi Professor O'Brien, thanks so much for sharing this interesting research to us! Violent crime is definitely a very important topic that we all care about especially in Chicago. You mentioned that you calculated six measures of physical disorder, social disorder, and violent crime from administrative records for all parcels in Boston. Do you have any relative research that are focusing are Chicago since the features of crime could be different from place to place? Also, what do you think is the largest implementation of this research in reality? Whether it is feasible?

[helyap]

Hi Professor O'Brien,

What are the ethical implications of your predictive modeling that are primarily focused on the location/spatial factors? Especially given the controversial backdrop of BWT that has essentially encouraged decades of what has been deemed to be overly aggressive, violent, and racist overpolicing, is there concern that this framework abstracts away the social and systemic elements of spaces and communities to ideas of objective variables that can supposedly serve as evidence for police intervention and their implementation of social order?

[erweinstein]

Hi Professor O'Brien,

I second (third? fifth?) my fellow students who want to know how well your approach would work (and what the results would be) when applied to cities other than Boston.

I would also be very interested in seeing different types of crime chosen for the analysis. As you note, "6.1% of parcels had at least one instance of public violence in 2011 and 1.3% had at least one gun-related event..." Perhaps larger datasets and those involving other cities would enable you to have a better choice of which type of crime to focus on for the analysis. There is certainly a larger issue, hardly unique to your work, that the most serious violent crimes, like gun-related incidents, are often the most-discussed by non-academics, but with very few events (and fewer still accounting for their distribution across "cells"/levels of analysis), they are the most difficult about which to draw statistically meaningful conclusions.

[Toushirow1]

Hi Professor O'Brien, Thank you for sharing your research. I was surprised by the connection between crime and geography, which I have never thought about. I am wondering if increasing the predictive parameter largely increases the accuracy of the analysis. Is necessary for a social scientist to increase model size as AI engineers do in the field of foundation models? Thank you!

[Coco-Jiachen-Yu]

Hello Professor O'Brien, thank you so much for sharing your work with us! The data source was restricted in Boston area. To what extent do you think the results will generalize to other cities that might have a different characterization of population (e..g., race, age, SES, etc)? Also, can you explain the reason why you limited location data to only the residential areas and excluded other busier or public areas such as downtown and schools? Does type of crime differ at different locations? How did exclusion of those location data affect the results?

[bningdling]

Hi Professor O’Brien, thank you for sharing your work with us!

Since this paper primarily concentrate on residential data, I was wondering do you think the findings of this paper would apply to business addresses as well. What are the essential characteristics of residential data that differences it from other data? Thank you very much!

[WonjeYun]

Hello Dr. O'Brien, Thank you for sharing your research ideas with us. The research seems to explain the possible factors behind the emergence of crime behavior at a given location. I wonder if the research could be used in a dynamic analysis of a specific region depending on the change of features in that area. If the analyzed features of a location could have impact on I wonder if the same features could be derived from data more further in the past.

[awaidyasin]

Thanks for your work, professor. I had a question regarding simultaneity concerns in the theories presented. Both theories implicitly carry a causal interpretation i.e., disorder leads to crime. Does the multi-level model setup in Equations (1)-(4) somehow take into account possible reverse causality? My understanding (from reading the setup) is that it is a static system, and the coefficients may not have causal meaning.

[mintaow]

Hi Professor O’Brien,

Thank you for sharing your work with us! It is very interesting that public violence and gun-related events in one geographic location could potentially have an influence on other areas, highlighting the interaction as well as the persistence between communities. I am particularly curious about your perspective on how the proximity or clustering factor (e.g. geographic distance, or certain racial group neighborhoods/communities) play in the interaction between communities in violence and aggravation.

[LynetteDang]

Dear Prof. O'Brien, Thank you for sharing your work with us. I am wondering if you ever take a step further to investigate how the proximity or clustering factor play a part not just in crime but recidivism. Looking forward to your talk tomorrow!

[cgyhumble0612]

Hi Professor, thank you for sharing such an interesting topic. I would like to know if it's possible to do the same researches in the higher level: violence and disorder at global geographic scales? What are the pros and cons for that? Look forward to your sharing this Thursday~

[zbchen0129]

Hi Professor O’Brien,

Thanks for your impressive work and sharing it with us! I believe your research would help communities a lot. I have a few questions. Could you explain more about the meaning of geographic scale? Besides, how do you measure the different types of disorder you mentioned? Do you think the models you use can be applied to Chicago? Thanks!

[sushanz]

Hello Professor O’Brien,

thanks for sharing! It will be great if you can explain more about how these geographic scale theories in the research as well as how you combined them in analysis