You are on page 1of 6

Alex Caviness

Annotated Bibliography
Friend, Zach. "Predictive Policing: Using Technology to Reduce Crime." FBI. 9 Apr.
2013. Web. Accessed 15 Oct. 2016.
In this FBI article, Zach Friend discusses the benefits predictive policing brought to different
police departments. He describes how the Santa Cruz Police Department, short on resources,
used algorithms to predict future crimes from the time and location of prior crimes. They
contacted a professor to write the program which allows[a] them to deploy shifts at whichever
time an area is more likely to have specific crimes. The algorithm was based on one for
earthquake aftershock prediction and incorporated general policing knowledge along with
crime research. All new crimes are added as data points to the algorithm to improve the
current accuracy of crime hot spots. Officers patrol these hot spots when not otherwise
obligated, and frequency of visits to hot spots is recorded. Predictive policing [b]reduced the
number of burglaries in Santa Cruz over the first six months, and it was proven to be effective
in a controlled experiment by the Los Angeles Police Department. The enhanced maps with
hot spots improve on the intuition of veteran cops, but still, predictive policing faces
opposition from those afraid it will eliminate the human and skill aspects of polici[c]ng.
Police departments in six states have implemented predictive policing and [d]been pleased
with the reductions in crime it has brought. The source is credible because it comes from a
government agency (the FBI) website and cites data from different police departments.[e] It
aims to educate [f]on the nature of predictive policing and exult its usefulness in the reduction
of crime. However[g] because it was published by the FBI, it is a little biased towards police
and glosses over the arguments against predictive policing.[h] Like the other sources, it

defines predictive policing, but the article focuses on its benefits and current use as opposed
to its unclear constitutionality and future potential. This article helped me define predictive
policing and understand how it is practically used. I will use this source in establishing what
predictive policing is[i] and how it has been implemented up to now.[j]

Hvistendahl, Mara. Crime forecasters. Science, vol. 353, no. 6307, 2016, pp.1484-1487. Web.
Accessed 18 Oct. 2016.
This article in the journal Science attempts to answer if predictive policing actually works
and, if so, if it is biased. Mara Hvistendahl writes about how predictive policings seemingly
objective nature is seen to improve relations with minorities by police but to cover up racial
bias with science by civil rights groups.[k] After all, the algorithms used to predict crimes are
based on the data from a biased system. Most programs use seismology algorithms to
determine [l]where aftershock crimes will occur after an initial crime. In Chicago, they use
data to compose a list of likely victims and offenders since most gun violence is caused by
a small number of people. However, this raises the issue of focussing on possible offenders
which will likely on [m]deepen the prejudice against minorities by the police. An assessment
of the list showed that supposed victims were no more likely to actually be victims, but far
more likely to be arrested than average citizens. Most encouraging results come from
historical data, but that does not imply a causal relationship. Furthermore, most departments
are not willing to perform a controlled experiment to evaluate predictive policing
effectiveness because it would take years when the average police chief lasts only three. The
key to making predictive policing work, the author says, is combining statistical methods of
policing with others. In ot[n]her words, data alone cannot solve all of policings issues. The

article is credible as it is published in a peer-reviewed journal and cites various police


departments and a RAND[o] re[p]port. It does seem to focus more on the problems with
predictive policing, but the data on those problems are largely backed by the RAND report.
The article shed morel light on current programs and problems than the other sources and
will be used primarily in demonstrating the bias in predictive policing.[q][r]

Kaste, Martin. "Can Software That Predicts Crime Pass Constitutional Muster?" NPR. 26
July 2013. Web. Accessed 15 Oct. 2016.
This NPR article by Martin Kaste is about the predictive policing as[s] its constitutionality as
reasonable suspicion. [t]He describes how the predictive policing technology in Los Angeles
and Seattle uses past data to predict property crime and gun violence. The technology can
predict high crime risk in areas other than the one in which a crime recently occurred. One
fear with predictive policing is Minority Report-styled pre-crime, where someone could be
arrested for a crime they were going to commit. The key difference here is that the algorithm
predicts where a crime is probable, not who will commit it. There remains a constitutional
issue, however. Reasonable suspicion is required for police to stop someone, but it is
unclear if a computer-generated hot spot on a map counts as reasonable suspicion. Most
police departments are urging against it, but the issue will likely remain ambiguous until
decided by the courts. This article is from NPR, a reliable news source commonly regarded
as unbiased.[u] It does a good job of addressing both sides of constitutionality debate without
leaning towards either one. It presents an overview of predictive policing to address concerns
people have about its scope and use. The article focuses on the constitutionality of predictive
policing in its current form. It helped me understand what a lot of the issue some people have

with predictive policing is. I will use this source in explaining the controversy in the present
state of predictive policing.[v][w]

Mayer-Schnberger, Viktor, and Kenneth Cukier. Big Data: A Revolution That Will Transform
How We Live, Work and Think. London: John Murray, 2013, pp. 157-163. Print.
In the book Big Data, the authors discuss the ways in which big data will revolutionize
society, from tracking deadly infections to finding the best time to buy airplane tickets. In
Chapter 8: Risks, they describe the various negative consequences big data could cause,
including those associated with predictive policing. They explicate[x][y] how predictive
policing could end up violating civil rights if went unchecked. Big data is already used in
most state parole boards to help determine if parole should be granted to a prisoner by
predicting his likeliness of committing another crime. This, however, is a start towards a
more dangerous path as police departments seek to identify who will commit a crime of
predicting a crime and making the criminal-to-be culpable for a crime he had not yet
committed. This act would violate the principles of justice and the presumptions of
innocence foundational to our legal system, however well intentioned it was. Besides, data
will never be perfect and there is no guarantee a warning from the computer was not a false
alarm. The[z] ties back to the problem[aa] that big data is all about correlations and is incapable
of establishing causation. The source is credible because it was written by two of foremost
experts in the data analytics field. Having worked in the industry, they have a good scope
for big datas potential and drawbacks. It is a popular work, however, and it may have been
written for entertainment to the extend of overselling the drama of its conseque[ab]nces. It
addresses the negative repercussions possible in the future of predictive policing more than

the other sources. This source helped me understand predictive policing in the larger context
of big data. I will use this source to connect predictive policing to big data and demonstrate
the negative possible future of predictive policing.[ac]
**my comments** - Sam
Overall, nice job. I felt like I understood the sources and you did a nice job of discussing
validity and bias. I found that stylistically you went into the explanation of the sources as
if it were your own, which is fine but I was forgetting that it was from them, not you. Its
really probably a nonissue but I just noticed it. Do what you will. Otherwise, one of your
sources was missing a possible uses part and there were a few parts where I wasnt
quite sure what you meant. No major problems anywhere though! Also, your choice of
sources seems very well-rounded, but did we need a website?

[a]In the lines above, verbs are in past-tense (brought, used, contacted); I would
change "allows" to "allowed" just to ensure parallel structure.
[b]has
[c]this is important yay
[d]"and *have* been pleased..."
[e]This is some good stuff.
[f]Make sure to include who it serves to educate, or the intended audience
[g]need comma
[h]Love how you mentioned the bias!
[i]you should definitely use the statistics
[j]Overall, very thorough job! You summarized, evaluated, and established the
usefulness of the source to your research perfectly. Just be careful with your verb
tenses. You switched from past to present often. For example, you wrote, "The
algorithm *was* based on one for earthquake aftershock prediction and *incorporated*
general policing knowledge along with crime research. All new crimes *are* added as
data points to the algorithm to improve the current accuracy of crime hot spots."
Would it be, "All new crimes *were* added..."? Just check on that; I may be wrong!
Other than that, the source seems to be perfect for your topic!
[k]Okay, so I read this sentence a few times. Is it supposed to be, "Mara Hvistendahl
writes about how predictive policings seemingly objective nature is *NOT* seen to
improve relations with minorities by police, but to cover up racial bias with science by
civil rights groups."
[l]is there an extra i here?
[m]Is the "on" supposed to be there? The sentence makes sense without it as well!
[n]This is all what the article is saying?
[o]Is this an acronym? If it is, I would spell it out.
[p]should I know what this is?
[q]are you going to use it?

[r]Great job of comparing the source to the other sources you have! I would conclude
with a sentence about HOW examining the bias in predictive policing will fortify your
overall argument.
[s]Instead of "as", is it supposed to be "and"?
[t]maybe it's early but this sentence isn't making sense to me
[u]Fabulous.
[v]this is good
[w]Slay. This explanation of the source thoroughly conveys what the source covers and
its value to your work. Make sure to include the intended audience.
[x]stylistic choice... explicate seems pretentious but I know it's how you write
[y]Ouch. Are you calling my writing style pretentious?
[z]This
[aa]I liked how you connected your info. back to the overarching issue.
[ab]should be "extent" but this is a very valid point
[ac]GREAT CONCLUSION.

You might also like