April 8, 2026 • 32 mins
I talk a lot about crime analysis and analyzing crime data, but I realized that I’ve never actually talked to a crime analyst yet. To correct this, I’m talking with Carlee Ruiz, a former crime analyst and policing researcher, discussing how she bridges the gap between academic research and real-world law enforcement through her platform, Police Research Hub.
Carlee highlights how valuable evidence-based insights are often inaccessible or underused, and explores the growing role of data, AI, and technologies like real-time crime centers in policing. The conversation also examines the balance between anecdotal experience and empirical evidence, emphasizing that context and implementation matter. Ruiz further discusses recruitment and retention challenges, pointing to leadership and organizational culture as key drivers of officer turnover.
Carlee Ruiz is a former crime and intelligence analyst with the Modesto Police Department and now works in research and technical assistance for law enforcement agencies at RTI International. She saw firsthand that much of the research useful to officers wasn’t easily accessible, which inspired her to create a website that translates peer-reviewed studies into practical, actionable insights for policing.
Resources
Follow the Jeff-alytics Podcast:
- Instagram: @RealTimeCrimeIndex
- LinkedIn: Real-Time Crime Index
- Website: The Jeff-alytics Podcast
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_01 (00:00):
I'm Jeff Asher and this is the Jeffalytics Podcast.
(00:04):
For most people, research aboutpolicing lives in academic
journals, buried behindpaywalls, written in a way that
almost guarantees it won't beused by the people making
decisions in the field.
But that research actuallymatters.
It shapes how departments thinkabout crime, how they deploy
resources, and what strategieswork or don't.
Carly Ruiz has made it hermission to close that gap.
She's a former crime analyst,turned researcher, and the
(00:27):
creator of the Police ResearchHub, where she translates
academic studies into insightsthat policymakers, law
enforcement professionals, andinterested members of the public
can actually use.
In this episode, we talk aboutwhat crime analysts really do,
why data visualizations playsuch a critical role in decision
making, and how emerging toolsare changing the landscape in
criminal justice.
We also get into the limits ofthat research, the tensions
(00:49):
between data and livedexperience, and what the
evidence actually says aboutimportant topics like
surveillance technology andpolice recruitment and
retention.
Let's get started.
My guest today is Carly Ruiz.
Carly, thanks for joining me.
SPEAKER_00 (01:02):
Hi, hi Jeff.
How are you doing?
SPEAKER_01 (01:04):
I'm great.
As you can see, uh I am freezingin my house.
So I'm I'm giving up the gamethat uh not only do I not have
great insulation, apparently,but uh apparently I have a low
tolerance for heat.
So we're in a giant, giantsweatshirt right now.
Um so now that I've alreadytotally gotten off track, uh
Carly, what is your background?
(01:24):
What what brings you here today?
SPEAKER_00 (01:26):
I guess the biggest thing that brings me here today
is that I host a little websitecalled Police Research Hub,
where I take peer-reviewedresearch, policing research that
usually stays up with the ivorytower and the academics and
bring it down to make itactually actionable and bring
insights from that to lawenforcement or law enforcement
(01:46):
adjacent profession.
So a little bit of mybackground, I was a crime and
intelligence analyst with anagency here in California for
four years.
And I think I was in almostevery single department there at
one point where I worked intheir RTCC, I worked in
investigations, I worked in thechief's office, I worked in
their intelligence area.
But a lot of times I was the onethat was tagged for a like, hey,
(02:09):
we're doing this project, canyou see if it's working?
Um, and you know, doing kind ofthese like low-level research
projects.
And that's kind of really whereI shined and really had a great
interest in.
So I wanted to take that to thenational stage.
So I joined RTI as a justiceresearcher in their policing
research program, where I do alot of projects around like
fatal, non-fatal shootings forincreasing clearance rates,
(02:31):
helping with police retention,peer support programs.
I also evaluate organized retailtheft, you know, violence
intervention, you know, allthese little things.
If it involves a law enforcementagency, I will probably hop on
it because I love working withlaw enforcement.
From there, I especiallyrealized both from my experience
within working in the agency andthen kind of working outside as
(02:53):
a researcher, is that a lot ofthis like really good academic
research that has really, youknow, key findings that can be
really helpful for lawenforcement is just not
accessible.
It's either locked behind apaywall or just in somewhere
that really most law enforcementwouldn't even think to look when
they're trying to think of, oh,how do we solve this problem?
And so I wanted to be able tobring that out out of those
(03:16):
journals and bring it intosomewhere that's accessible and
usable and actionable for lawenforcement.
SPEAKER_01 (03:21):
I'm embarrassed to say that you are the first crime
analyst or former crime analystthat I've had on the podcast to
this point.
SPEAKER_00 (03:28):
I'm surprised.
SPEAKER_01 (03:29):
Um I I just thought this is a bit of a tangent, but
can you just sort of describelike what is a crime analyst?
What does they what does a crimeanalyst do for a police
department?
I I realize some people in theaudience may not just be aware
of this as a profession.
SPEAKER_00 (03:43):
Yeah, definitely.
Have the honor.
Um so yeah, crime intelligenceanalysts.
So if the role will varydepending on the law enforcement
agency, but overall, it kind offits into two buckets, I would
put.
So there is the crime or thecrime part to where you look at
the crime stats.
A lot of times it's looking atweekly, bi-weekly, monthly, um,
helping prepare for Comstatmeetings or ILP meetings, you
(04:06):
know, seeing what's up, what'sdown, looking at the individual
crimes to see if there's amethod that's the same across
these crimes, maybe they'reconnected and kind of doing a
deep dive in there.
And then also sometimes doingthis like research background,
looking in the data and lookingat what's the problem and if we
can find a solution, is oursolution working?
And then there's the kind ofmore what I would call the
intelligence side where a lot oftimes you'll work with the
(04:29):
investigators.
So a lot of times like homicideinvestigators, you're right
where with them, pulling up,doing workups, so doing a deep
dive into a person's backgroundor an address or a car and
trying to help them find leadsto solve their case, or even
helping that at the patrollevel, especially in a real-time
crime center.
So it's really a crimeintelligence analyst, you're
you're deep in the data, you'redeep in the computer, either
(04:51):
looking at crime stats orlooking in someone's background
and you know, trying to likesupport the department that way.
SPEAKER_01 (04:57):
All right.
That tangent aside, yeah.
I wanted to get back to thepolice research hub and and talk
a lot about this because it'ssuch a great tool.
It's it's such a great resource.
And you kind of answered like,what is it?
Where did the genesis come from?
But who, when you built this,who is sort of the client?
Who do you see as the personusing this, or is it not just
(05:18):
one person that you kind ofbuilt this in mind for?
SPEAKER_00 (05:20):
When I built this, I started trial testing this on
LinkedIn.
Like my first one was postingIan Adams study on that, hey, AI
report writing software is notas efficient as you think it is.
Um, and that got a lot oftraction and I realized with law
enforcement, they weren't awareof that study.
And, you know, there waspositive people who were
appreciative of being posted andothers who were not.
(05:42):
But, you know, it kind of showedlike, oh, law enforcement,
especially, you know, is notaware of these, you know,
studies that maybe academics areaware that have come out.
And so most of the time I re Iframe it for a law enforcement
officer, especially with Iusually provide resources
towards the end of like whetherit's a training or whether it's
(06:03):
a guide.
So a lot of times I frame itprobably more for the sergeant
or higher um command stafflevel.
But I also, I mean, anyone ofany level, and even civilians,
um, can read this.
I've I've done some things ondata that could help with the
crime analysts, and I've donesome, you know, some things with
you're maybe someone like avictim advocate coming from
(06:25):
you're still part of maybe apolice department or out like in
that sphere, uh, but maybe notworking directly with the police
department can be uh useful.
But a lot of times it's I writeit more towards the, hey, you're
as a law enforcementprofessional, you know, this is
something you should be aware ofor know, or here's a resource
for.
SPEAKER_01 (06:43):
So you mentioned Ian, who is a recent guest of
the pod as well, whichimmediately brings to mind how
do you come up with the researchand is there a role for AI in
not necessarily doing what youdo because you bring your
expertise, but in in makingthings easier for identifying
research for any part of this,or is this all a, you know, a
(07:04):
human needs to sort of bringthis to bear?
SPEAKER_00 (07:06):
It depends.
So as far as the AI piece, Imean, a lot of times with the
police departments that I'veworked with in my my own
experience, they usually havesomething where it's like, we
have a problem.
Like, let's say, like, oh, we'reseeing a large increase in calls
related to homelessness.
What do we do to solve thisproblem?
And a lot of times right nowit's like, well, what is our
neighboring agency doing?
(07:27):
You know, who have I talked to?
Which is a great way to getinformation of like who's doing
what, but it's not a reallygreat way to get like what has
evidence-based found to work andhow you can implement that.
So I guess I mean AI could beuseful in being able to at least
give you the research articlesor at least get you, hey, there
is an evidence base on thisparticular topic that could be
(07:50):
helpful when you're developingyour program, you know, and
pulling articles for you.
But I would still recommendreading them because sometimes
it does like to fabricatecitations.
And so I would make sure that,you know, hey, give me the
actual article.
Don't just give me the answersbecause it might give you
something that's incorrect.
But it it could definitely giveyou a nice starting point of
like, I have this question,what's out there?
(08:11):
And then from there you can jumpfrom other to other things.
SPEAKER_01 (08:14):
When you're putting all this together, having done
this now for a little while,what are the biggest challenges
of taking what can be reallyinaccessible research?
I mean, you know, academics canwrite in a way that makes it
almost impossible for anyonethat's not themselves to
understand it to an audience,like, especially if your target
is police leadership, that likeis the complete polar opposite
(08:35):
of that sometimes.
SPEAKER_00 (08:38):
No.
And I mean, sometimes I'm like,whew, I'm reading this and I'm
just like, man, I am struggling.
Um, but I would say a lot oftimes, especially the data
pieces, like I've done a couplepieces on like predictive data
or AI, if you're not in thatrealm, can be difficult to
understand.
You know, for a lot of people,AI is still this black box of
(08:59):
mystery.
And I don't, you know, you putin something and then it puts
something, and I don't know howyou did it.
So for especially for academicresearch articles, it could be
difficult to understand what itdoes, but there's some really
good, interesting things comingout about predictive analytics
that can be useful to lawenforcement, but also is very
key to understand how it doesthat.
I do ask when people subscribeto my website, what topics do
(09:23):
you want to hear?
And using data is my mostpopular like topic that mostly
law enforcement who subscribewant to hear about.
So it is something that they'revery interested in and want to
learn more about, you know,using data and tech and tools,
but you know, and I think thosearticles especially are the most
difficult to understand as alayperson.
SPEAKER_01 (09:41):
What kind of feedback have you gotten on all
of this?
SPEAKER_00 (09:44):
As far as feedback, most of it's been positive as
far as bringing light to some uhfindings that have, you know,
that people were not aware ofbefore.
I add a lot of storytellingelements, or if I have personal
experience, I'll add that andpeople, you know, like that kind
of realism in there.
You know, I have got, especiallywhen whenever I bring up AI,
(10:04):
there's always a little bit oflike the debate of, you know,
usefulness or, you know, is ithere, take our jobs, you know,
like there's always some sort ofkind of squabble uh there.
But most of the time thefeedback's been very positive.
There is a few resources outthere that do something similar.
I know the um like the appliedpolicing briefs, Dr.
Radcliffe has his podcasts wherehe interviews people as well.
(10:26):
So there is resources out therelike like mine, but there's not
enough, I would say.
And so I think people are veryhungry for this kind of
information.
SPEAKER_01 (10:35):
Yeah, absolutely.
So can we can we dive into whatyou found?
I'm putting you on the spot.
I don't know that you have anencyclopedic memory of all of
the things you've written, buthave any of the research topics
that you've covered reallyjumped out as like either
unexpected or have you thinktranslating it to a wider
audience has had a biggerimpact?
SPEAKER_00 (10:56):
I think, okay, so one I will use I always
surprised when I'll thinksomething like, I don't know,
this will be popular, but I wantto get this out there and then
it explodes.
One, I did a post on datavisualization and what's the
best way to display graphs in aComstat meeting, you know?
And something as a crimeanalyst, I was very like die on
(11:18):
this hill to, you know, let'snot do week-to-week comparison,
what's up, what's down.
I think this was one of ReneeMitchell's articles doing a
study of like, hey, let's showthese different types of data
visualization graphs to a lawenforcement leadership and see
if they can make decisions, whatwhich graph helped them make the
better decision in terms ofresponse?
(11:39):
And had did they understand itbetter?
Instead of like, let's just donormal percent change graphs,
let's do one that shows themedian for this time and a
standard deviation above andbelow, you know, to show, hey,
crime does fluctuate, but hey,let's oh, this is above our
norm.
This is about out of thestandard one standard deviation.
Maybe we should do somethingabout that.
(12:01):
And, you know, with that kind ofgraph, leadership was able to at
least deploy a better solutionto address that increase in
crime versus just using a normalpercent change.
And which maybe this is not themost exciting finding, but for
me, but like that's somethingthat that I don't think law
enforcement is aware of of like,yeah, there is a study on what's
(12:24):
the best data visualization touse to help you make decisions.
And maybe that there's betterways and there's not so bad, you
know, not so good ways.
And that, I mean, that wasprobably one of my more popular
posts.
A lot of people were commentingon it and sharing it and
reposting it and showing thatlike this is of some great
interest.
And I don't think a lot of lawenforcement was aware that
there's other ways to do thisthan just doing percent change.
SPEAKER_01 (12:45):
Sometimes I'll post a graph that'll be like percent
change because that's it's justeasiest sometimes that it's for
the public audience.
And if you'll be like, well, youshould really use z-scores.
Like, if you can like trying todescribe that to an audience, I
mean it's very difficult.
SPEAKER_00 (12:59):
How do you balance it to I made the mistake as when
I was a crime analyst, I I toldmy chief I was like, oh, and the
poison z-score is this, and thenthey never let me forget about
it for years on end.
You know?
I mean, it was like they wereimpressed, but also like you
know, sometimes you don't usethe terminology and just keep
(13:20):
the stats language out and tryto keep it as plain language as
possible so they understand.
SPEAKER_01 (13:24):
Yeah, it it sounds fancy.
Yes, not really, but one one ofthe things, uh just again going
by like some of the researchthat's available, uh a couple of
them, especially the recentones, talk about sort of the
role of real-time crime centers,the role of cameras.
Have you uh like synthesized abunch of these?
And um, I'm just looking at likeuh there was one on I I forget
(13:48):
the name of it, but ha theactual like review of real-time
crime centers, one onprioritizing burglary cases that
have CCTV.
Have you thought about likesynthesizing some of these?
Have you put all of themtogether into like a larger
analysis in your head?
Or how does that work?
SPEAKER_00 (14:03):
Yeah, so at least with the RTCC camera ones, I
worked for a long time in anRTCC, so like that's a that's a
really good background I canpull from as far as like I know
how these work and how whatwould be useful for a law
enforcement agency to know aboutRTCCs.
I have an interest.
I do want to so like what yousaid, like pull together all the
resources on one topic, um, dolike basically like a literature
(14:27):
review like on RTCCs, likewhat's all the research out
there.
I do have a goal of doing that.
Maybe RTCC would be my firsttopic.
It's just it takes a long timeto be able to find all the
research that's out there, pullit together in a in a cohesive
form that doesn't make it into a50-page report.
You know, I I have been workingon that and I do hope to
(14:48):
eventually post at least uh acouple of those a year, or maybe
like one a quarter, uh, of like,hey, this is all the current
research on this one topic, uh,you know, AI report writing or
uh, you know, RTCC camera-basedtopic.
That is a future goal, notcurrently what I have right now.
SPEAKER_01 (15:07):
Can you sort of give me your impressions of sort of
RTCCs and kind of that synthesisof what the research shows and
not just RTCCs, but the theseare real-time crime centers, um,
but just like surveillancecameras and like the role that
surveillance cameras can playfor police departments with all
of this experience?
Have you sort of do you have ageneral idea of like what's the
best way to balance success andand privacy concerns and all
(15:31):
that?
SPEAKER_00 (15:31):
As far as so for real-time crime centers, they're
still relatively new in terms ofthe evidence base.
Um we are seeing some researchcome out that does show as far
as like case clearance-wise,they can help increase, you
know, the solving of cases,which logically that makes
sense.
There's more cameras, you'remore likely to capture evidence.
With it being staffed, your eyeson the scene faster.
(15:54):
And you know, the sooner yourespond to the scene, the more
likely you are to solve cases.
Especially that in that case,RTCs really shine.
There's a lot of other metricsthough, the RTCCs like, you
know, whether it's officersafety, you know, awareness of
what's going on in yourcommunity.
It's really hard to capture withdata.
Um, and so that's that's aharder point of like when you're
(16:17):
building an RTCC, you reallyhave to figure out what do you
want this to do for your agencyand know that is and understand
that metric might not be easy tocapture.
There is evidence that arecoming out that at least helping
solve cases, they can do that.
Now, crime reduction has variedresults, but there's not a lot
(16:38):
of research out there yet onevaluation of RTCC.
Now, cameras have been around alot longer than RTCCs, and
again, more on that case, theydo help with case clearances.
Now, crime reduction, I thinkthere is varied results
depending, you know, if thecamera is visible, it can't work
as a deterrent, but you know,does that push it off to other
(16:59):
areas?
It it kind of depends on theresearch in the area.
Um, but I think overall, youknow, real-time crime centers,
license plate readers, puttingall those technology pieces
together is, you know, there ismore evidence and it's starting
to grow that like, you know,this can be very helpful in
solving cases, in gettingsituational awareness in in your
(17:22):
community and what's happeningin your community.
SPEAKER_01 (17:24):
Looking more broadly, how do you respond when
sort of the anecdote leadsrather than the research?
When you run into people, you'reexplaining what the research
shows, and they say that theirlife experience or their opinion
says X and actually the datasays Y.
How do you have that discussion?
SPEAKER_00 (17:42):
Yeah, so I mean, always validate the antidote.
I mean, people have their ownexperiences, and especially with
policing, officers have lived itand experienced it.
And so, you know, theiranecdotes are are valid in what
they're experiencing.
And I also, with differenttechniques, different technology
works differently in differentcities.
(18:03):
What works in one city might notwork in another, or the city has
to adjust because theircommunity is not the same as
Chicago is, you know, Chicagoand Oakland are not the same.
And so what might work in onemight not work in the other.
And so it works for them becauseof their personality and who
they interact with, and maybethey have rapport, where in this
(18:23):
other area, like other city, theofficer doesn't have the same
rapport and it doesn't work.
And so that is like I I hesitateto shut down anecdotes because
things will differ.
And also the field of policingresearch is relatively new.
And so we haven't, you know, onestudy might say this doesn't
work, but if we look at it in adifferent context, it might
(18:43):
work.
And so to shut down anythingmeans like I I hesitate to to do
that.
SPEAKER_01 (18:49):
Is there a question or a data set or an answer that
you it isn't answered somethingthat isn't answered by policing
research that you really wishwas available?
Like is there something thatyou're just dying to write about
and think about, or some thingsthat you're dying to write about
and think about, and it justthere's just nothing there?
SPEAKER_00 (19:06):
I mean, okay, so I I do a lot right now.
I have a big project on AI andpolicing.
So it's very new.
So we don't have a lot.
Um so loved more of like a lotof this technology, more of how
is especially like dataanalytics, integrated data
analytics, how does you know uhthat help law enforcement if you
(19:28):
bring all these data sourcestogether and you know, give them
LPR and their CAD data and theirRMS and their neighboring
agencies and you know, OTSINTand all that stuff, you know,
does that really help them solvecases?
There's nothing on that now, youknow, and so and a lot of it it
has a lot of potential.
And as an analyst, I'm like, Ican see the potential of it, but
(19:51):
you know, there's no research onit.
And then also like license platereaders, there's not a lot of
license, there's not a lot ofresearch out there on them
because a lot it's verydifficult to identify like every
agency who dubs on LPR, I alwaystry to say, hey, you know, if
you make an arrest, make a notethat it was because you pulled
(20:11):
it from LPR, it was an LPR hit.
And that's always been difficultto to uh you know get across to
agencies of like track that infoso you can prove that LPR works
for you.
It helps you, you know, withyour auto theft cases or with
you know your warrants andwhatnot.
I mean, I even talked to acouple guys at Flock Safety who
are like trying to get agenciesto track this data, and it's
(20:34):
it's very difficult even fromtheir end.
So that's something also withLPR and AI and how that's
changing the policing field, weneed so much more research on.
SPEAKER_01 (20:45):
Sort of at the other end of the spectrum, how do you
handle research that you knowit's not obvious what the what
the answer is, what the findingis, or potentially it's a really
interesting finding, but it'spotentially not the strongest
finding, or even potentially abad finding.
Like obviously there's a widerange of quality and research.
How how do you sort of workthrough that?
SPEAKER_00 (21:08):
What comes to mind is um so Jerry Radcliffe ran a
study in Philadelphia.
SPEAKER_01 (21:14):
Also also a former podcast.
SPEAKER_00 (21:17):
Uh he he ran a study in Philadelphia where they had
um, I think it was like withintheir transit unit, they signed
a um like a social worker withan officer.
And especially when this was inthe height of like alternative
policing programs, instead ofhaving the officers show up, you
have someone else show likesomeone that's of either a
(21:38):
psychiatric care social workershow up.
And I think, yeah, like his hisfinding was like you know,
officers who like the officeronly actually had better
results, or they're about thesame result as an officer with a
social worker.
And when you look at it on thesurface, you're like, oh, so
(21:59):
social workers.
Don't really work.
But when you dive deeper intohis study, he, you know, brings
out the point of, well, okay,these actually these officers
who are working have beentrained in like multiple things
of like with like a socialworker.
Like many had a social workerbackground, and they've been
trained in like how to interactwith people who are, you know,
(22:20):
having mental episodes, who aredealing with addiction and
drugs.
So they've been speciallytrained.
And then the social worker side,there was a lot of turnover
because the way the position wasadvertised, you know, they
didn't realize how how much workwould go into it.
They just had a really hard timekeeping these social workers
employed in these shifts.
(22:41):
And so it's kind of like, well,does that speak to the, you
know, social workers don't work?
Or does it speak more to, well,these officers are specially
trained so they do better, butalso these social workers
weren't really given a good shotbecause they had a lot of
turnover, they didn't have agreat amount of training.
And so, you know, so that's likea mixed finding where it's
(23:02):
brought out of like maybesomething we can learn from this
is when you're implementingsomething, make sure it's given
its best foot forward.
And so that's something I Iemphasize a lot, and sometimes
in implementation research thatI talk about is if you're
implementing a solution,consider all these factors.
Because sometimes a lot of timesI see with law enforcement,
they'll just let's just startsomething, get it going, run,
(23:24):
run with it.
And then they there's a lot ofissues that pop up, and then
they say, well, it didn't work.
And you're like, well, it mightnot have worked because it
wasn't implemented well.
And so that's you know, alwayscome kind of a consideration
when we see mixed findings arenot working is how well was it
implemented?
How can we better implement it?
You know, was enough forethoughtand consideration put into it at
(23:45):
the at the beginning for it to,you know, actually have a
chance.
SPEAKER_01 (23:49):
Which sort of brings up the question of how do you
how do you communicate to youraudience that like, here's what
we found, this was a study, itit's not like every experience,
the you know, your results mayvary, sort of thing, in terms of
this is what the research shows,but the research is not perfect.
How do you how do youcommunicate that to people that
don't they say, oh, somebody didresearch, it must be true.
SPEAKER_00 (24:11):
So for the instance for the uh Radcliffe study, I
did display the results of like,okay, you know, they're either
like the same or the officersactually did a little better.
But I did go into as a part ofreading the full study, is going
into, hey, these were thedownsides of like basically kind
of what I just told you of likethe social workers had a lot of
(24:35):
high turnover, they didn't haveclear enough job description,
not a lot of training, and theseofficers have had a lot of
training that more than generalofficers do.
So I brought it a point of like,if you're interested in doing
this program, even though it waskind of a a null, that here's a
ways that maybe you can havemore success with it.
I also write a lot of differentone studies looking at
(24:55):
implementation science and kindof like stepping up like law
enforcement of like if you'retrying to do a reform on a
certain thing, hey, this is agood case study of like how to
do like how to do reform, how tohow to implement a solution and
do it in the right way the firsttime.
Talk more about how they did itversus the findings.
(25:16):
Um I I kind of we'll talk moreon that.
SPEAKER_01 (25:19):
The the last kind of like big area I want to get into
is sort of near and dear to myheart, and I think you've got
some experience with this, whichis um police recruitment and
retention.
And I'll sort of set the stagefor the audience.
Um, I'm sure you know this, butthe vast majority of police
departments lost a ton ofofficers between 2019 and that
sort of 2022 to 2023 period, andare really struggling to start
(25:44):
to grow.
And I have a substack that I'lleventually publish that it looks
like they maybe started to growa little bit in 2025, but still
most big police departments arewell below where they were six
years ago.
And obviously, from a umresource allocation standpoint,
how many detectives you can puton the streets, how your
response times will look, howyour clearance rates will look,
(26:05):
all of these things areimpacted.
Your service delivery, policedepartments as service delivery
organizations are very muchimpacted by the fall of
officers.
I swear there's a question inthis.
Um What have you found?
What are the the key issuesfacing departments?
How can they start to grow andstem the tide of resignations
(26:26):
and of and improve retention?
Is is there research findings inthis?
SPEAKER_00 (26:31):
Yeah.
So there is a lot of recentstuff that's come out in terms
of policing retention and even alittle bit of what I've been
involved in.
I mean, from what I've read andfrom what some of the projects
that I've done around policeretention, the big number one is
leadership.
Is for one of my projects, I dida deep dive in like Glassdoor
(26:51):
and uh Reddit data and lookingat police retention and what
officers were saying in regardsto why are they leaving or you
know, why, you know, why whythey're considering and leaving
and leadership, the way thatleadership treats them,
especially when there's anincident that happens, and then
a lot of times officers feellike, oh, so-and-so was thrown
(27:12):
under the bus as a scapegoat, orI feel like because I received a
complaint, now I'm being putthrough the ringer, I'm I'm
guilty before proof and youknow, I have to prove my
innocence.
That kind of feeling was atleast what I've been seeing is a
big driver of uh retention.
I know I've posted a couplestudies that show that, like,
(27:33):
you know, one of the big pointsfor like officer wellness is
like, oh, PTSD, you know, you'rebeing exposed to very traumatic
incidents, especially as a lawenforcement officer, you know,
that's driving that.
And while that PTSD is a majorcomponent of officer wellness
that needs to be considered andthat needs to be um, you know,
given resources for andunderstanding for, that these
(27:55):
organizational factors of howyour leadership is, how your
scheduling is, how how safe youfeel talking to your sergeant or
to your lieutenant, you don'tfeel like your organization
supports you, you're way morelikely to leave.
And also way more likely to feellike your PTSD symptoms like are
exaggerated.
There's a lot of research outthere that a lot of times it's
(28:16):
not just like, oh, it's becauseof what happened, they can't
handle what's happening out inthe field.
It's actually because they don'thave the support of their
leadership behind them, but thesupport of their organization
behind them that's driving themout or driving them to have, you
know, burnout.
Good thing is is that can bechanged.
I definitely had a few, a fewarticles on that.
SPEAKER_01 (28:35):
Going, you know, full into opinion mode.
Do you think that it's possiblefor police departments to grow
substantially, or is this justnot an environment that's
conducive to that and we have tolearn to live with what we have?
SPEAKER_00 (28:47):
Um I mean, I would say like mass hiring, I mean
never is really that good of athing, but the way the recruit
matters.
And I think as far asrecruitment-wise, especially
with like the efforts of 30 by30, which is if you're not
aware, is you know, for in theUS to get if you're a 30 by 30
(29:10):
agency, you're trying to get 30%of your recruits um to be women
by 2030.
There's a lot of different likerecruitment efforts are towards
that way.
It's a little harder to say ifthere is like a growing mass
that you'll be able to get tothe numbers you were in 2018.
I know some departments are,some departments are back to
where they were.
And definitely the departmentsthat I've spoken to have either
(29:32):
done some wellness initiativesor really made sure to keep
their guys and focus onretention.
And retention bonuses don'tnecessarily work, but you know,
to make sure that their officersare heard and more of a focus on
that versus recruitment.
But if you're down on yourofficers, I mean, you need to
focus on recruitment and try tofocus on diversity recruitment.
(29:54):
And there is a few studies outthere, at least, that show,
depending on like the videosthat you have, what type of
traits you show will change howwhat people who will apply.
SPEAKER_01 (30:06):
So if you're show, you know, showing a lot of the
the SWAT to your yourpromotional video of everybody
smiling and doing their hardwork.
SPEAKER_00 (30:13):
So, you know, there is studies out there if like if
you show, you know, officers whoare helping the community and
yeah, they are doing thesmiling, like playing with the
kids, but also, you know,helping the community members
and showing that kind of front,you're more likely to get more
women.
Um, versus if you're showing thethe SWAT machine and the
helicopter and the guns andlike, oh, we're gonna beat the
(30:36):
bad guys, you're less likely toget women.
And so, like, there's differenttechniques depending on who
you're trying to recruit for.
SPEAKER_01 (30:42):
Well, that's fascinating.
And uh it's all great.
Um, policeresearchhub.com is thewebsite, right?
SPEAKER_00 (30:49):
Yes, yes.
SPEAKER_01 (30:50):
Yeah.
So so what's next?
SPEAKER_00 (30:53):
My goal is to keep continuing posting, finding all
these the research articles thatat least I think I find
interesting, and I knowhopefully other people will find
interesting and keep going withthat.
I'm I'm also pretty prevalent onLinkedIn as well, and I will
post questions to the group andsee who's what's it people are
interested in.
And uh hopefully eventuallywe'll be able to do those deep
(31:14):
dives on a single topic becausethat's like my next big goal
that I want to do on that.
SPEAKER_01 (31:19):
It's wild how like each little community has found
different websites now in 2026,2025.
So, yeah, it's a vibrant policeresearch community on LinkedIn.
So and you're one of the moreprolific posters in talking
about the research, so it'sgreat to follow.
All right, Carly, thank you somuch for joining the show and uh
keep up the great work.
It's it's fabulous.
(31:40):
Uh, police researchhub.com.
SPEAKER_00 (31:42):
Thank you for having me.
SPEAKER_01 (31:44):
Thanks for listening to the Jeffalytics Podcast.
Be sure to subscribe and tolearn more, head on over to
ahdatalytics.com for moreinformation and previous
episodes.
If you like what you heard,please leave a glowing review,
which will help others todiscover the show.
Till next time, I'm Jeff Asher.
I'm Jeff Asher and this is the Jeffalytics Podcast.
(00:04):
For most people, research aboutpolicing lives in academic
journals, buried behindpaywalls, written in a way that
almost guarantees it won't beused by the people making
decisions in the field.
But that research actuallymatters.
It shapes how departments thinkabout crime, how they deploy
resources, and what strategieswork or don't.
Carly Ruiz has made it hermission to close that gap.
She's a former crime analyst,turned researcher, and the
(00:27):
creator of the Police ResearchHub, where she translates
academic studies into insightsthat policymakers, law
enforcement professionals, andinterested members of the public
can actually use.
In this episode, we talk aboutwhat crime analysts really do,
why data visualizations playsuch a critical role in decision
making, and how emerging toolsare changing the landscape in
criminal justice.
We also get into the limits ofthat research, the tensions
(00:49):
between data and livedexperience, and what the
evidence actually says aboutimportant topics like
surveillance technology andpolice recruitment and
retention.
Let's get started.
My guest today is Carly Ruiz.
Carly, thanks for joining me.
SPEAKER_00 (01:02):
Hi, hi Jeff.
How are you doing?
SPEAKER_01 (01:04):
I'm great.
As you can see, uh I am freezingin my house.
So I'm I'm giving up the gamethat uh not only do I not have
great insulation, apparently,but uh apparently I have a low
tolerance for heat.
So we're in a giant, giantsweatshirt right now.
Um so now that I've alreadytotally gotten off track, uh
Carly, what is your background?
(01:24):
What what brings you here today?
SPEAKER_00 (01:26):
I guess the biggest thing that brings me here today
is that I host a little websitecalled Police Research Hub,
where I take peer-reviewedresearch, policing research that
usually stays up with the ivorytower and the academics and
bring it down to make itactually actionable and bring
insights from that to lawenforcement or law enforcement
(01:46):
adjacent profession.
So a little bit of mybackground, I was a crime and
intelligence analyst with anagency here in California for
four years.
And I think I was in almostevery single department there at
one point where I worked intheir RTCC, I worked in
investigations, I worked in thechief's office, I worked in
their intelligence area.
But a lot of times I was the onethat was tagged for a like, hey,
(02:09):
we're doing this project, canyou see if it's working?
Um, and you know, doing kind ofthese like low-level research
projects.
And that's kind of really whereI shined and really had a great
interest in.
So I wanted to take that to thenational stage.
So I joined RTI as a justiceresearcher in their policing
research program, where I do alot of projects around like
fatal, non-fatal shootings forincreasing clearance rates,
(02:31):
helping with police retention,peer support programs.
I also evaluate organized retailtheft, you know, violence
intervention, you know, allthese little things.
If it involves a law enforcementagency, I will probably hop on
it because I love working withlaw enforcement.
From there, I especiallyrealized both from my experience
within working in the agency andthen kind of working outside as
(02:53):
a researcher, is that a lot ofthis like really good academic
research that has really, youknow, key findings that can be
really helpful for lawenforcement is just not
accessible.
It's either locked behind apaywall or just in somewhere
that really most law enforcementwouldn't even think to look when
they're trying to think of, oh,how do we solve this problem?
And so I wanted to be able tobring that out out of those
(03:16):
journals and bring it intosomewhere that's accessible and
usable and actionable for lawenforcement.
SPEAKER_01 (03:21):
I'm embarrassed to say that you are the first crime
analyst or former crime analystthat I've had on the podcast to
this point.
SPEAKER_00 (03:28):
I'm surprised.
SPEAKER_01 (03:29):
Um I I just thought this is a bit of a tangent, but
can you just sort of describelike what is a crime analyst?
What does they what does a crimeanalyst do for a police
department?
I I realize some people in theaudience may not just be aware
of this as a profession.
SPEAKER_00 (03:43):
Yeah, definitely.
Have the honor.
Um so yeah, crime intelligenceanalysts.
So if the role will varydepending on the law enforcement
agency, but overall, it kind offits into two buckets, I would
put.
So there is the crime or thecrime part to where you look at
the crime stats.
A lot of times it's looking atweekly, bi-weekly, monthly, um,
helping prepare for Comstatmeetings or ILP meetings, you
(04:06):
know, seeing what's up, what'sdown, looking at the individual
crimes to see if there's amethod that's the same across
these crimes, maybe they'reconnected and kind of doing a
deep dive in there.
And then also sometimes doingthis like research background,
looking in the data and lookingat what's the problem and if we
can find a solution, is oursolution working?
And then there's the kind ofmore what I would call the
intelligence side where a lot oftimes you'll work with the
(04:29):
investigators.
So a lot of times like homicideinvestigators, you're right
where with them, pulling up,doing workups, so doing a deep
dive into a person's backgroundor an address or a car and
trying to help them find leadsto solve their case, or even
helping that at the patrollevel, especially in a real-time
crime center.
So it's really a crimeintelligence analyst, you're
you're deep in the data, you'redeep in the computer, either
(04:51):
looking at crime stats orlooking in someone's background
and you know, trying to likesupport the department that way.
SPEAKER_01 (04:57):
All right.
That tangent aside, yeah.
I wanted to get back to thepolice research hub and and talk
a lot about this because it'ssuch a great tool.
It's it's such a great resource.
And you kind of answered like,what is it?
Where did the genesis come from?
But who, when you built this,who is sort of the client?
Who do you see as the personusing this, or is it not just
(05:18):
one person that you kind ofbuilt this in mind for?
SPEAKER_00 (05:20):
When I built this, I started trial testing this on
LinkedIn.
Like my first one was postingIan Adams study on that, hey, AI
report writing software is notas efficient as you think it is.
Um, and that got a lot oftraction and I realized with law
enforcement, they weren't awareof that study.
And, you know, there waspositive people who were
appreciative of being posted andothers who were not.
(05:42):
But, you know, it kind of showedlike, oh, law enforcement,
especially, you know, is notaware of these, you know,
studies that maybe academics areaware that have come out.
And so most of the time I re Iframe it for a law enforcement
officer, especially with Iusually provide resources
towards the end of like whetherit's a training or whether it's
(06:03):
a guide.
So a lot of times I frame itprobably more for the sergeant
or higher um command stafflevel.
But I also, I mean, anyone ofany level, and even civilians,
um, can read this.
I've I've done some things ondata that could help with the
crime analysts, and I've donesome, you know, some things with
you're maybe someone like avictim advocate coming from
(06:25):
you're still part of maybe apolice department or out like in
that sphere, uh, but maybe notworking directly with the police
department can be uh useful.
But a lot of times it's I writeit more towards the, hey, you're
as a law enforcementprofessional, you know, this is
something you should be aware ofor know, or here's a resource
for.
SPEAKER_01 (06:43):
So you mentioned Ian, who is a recent guest of
the pod as well, whichimmediately brings to mind how
do you come up with the researchand is there a role for AI in
not necessarily doing what youdo because you bring your
expertise, but in in makingthings easier for identifying
research for any part of this,or is this all a, you know, a
(07:04):
human needs to sort of bringthis to bear?
SPEAKER_00 (07:06):
It depends.
So as far as the AI piece, Imean, a lot of times with the
police departments that I'veworked with in my my own
experience, they usually havesomething where it's like, we
have a problem.
Like, let's say, like, oh, we'reseeing a large increase in calls
related to homelessness.
What do we do to solve thisproblem?
And a lot of times right nowit's like, well, what is our
neighboring agency doing?
(07:27):
You know, who have I talked to?
Which is a great way to getinformation of like who's doing
what, but it's not a reallygreat way to get like what has
evidence-based found to work andhow you can implement that.
So I guess I mean AI could beuseful in being able to at least
give you the research articlesor at least get you, hey, there
is an evidence base on thisparticular topic that could be
(07:50):
helpful when you're developingyour program, you know, and
pulling articles for you.
But I would still recommendreading them because sometimes
it does like to fabricatecitations.
And so I would make sure that,you know, hey, give me the
actual article.
Don't just give me the answersbecause it might give you
something that's incorrect.
But it it could definitely giveyou a nice starting point of
like, I have this question,what's out there?
(08:11):
And then from there you can jumpfrom other to other things.
SPEAKER_01 (08:14):
When you're putting all this together, having done
this now for a little while,what are the biggest challenges
of taking what can be reallyinaccessible research?
I mean, you know, academics canwrite in a way that makes it
almost impossible for anyonethat's not themselves to
understand it to an audience,like, especially if your target
is police leadership, that likeis the complete polar opposite
(08:35):
of that sometimes.
SPEAKER_00 (08:38):
No.
And I mean, sometimes I'm like,whew, I'm reading this and I'm
just like, man, I am struggling.
Um, but I would say a lot oftimes, especially the data
pieces, like I've done a couplepieces on like predictive data
or AI, if you're not in thatrealm, can be difficult to
understand.
You know, for a lot of people,AI is still this black box of
(08:59):
mystery.
And I don't, you know, you putin something and then it puts
something, and I don't know howyou did it.
So for especially for academicresearch articles, it could be
difficult to understand what itdoes, but there's some really
good, interesting things comingout about predictive analytics
that can be useful to lawenforcement, but also is very
key to understand how it doesthat.
I do ask when people subscribeto my website, what topics do
(09:23):
you want to hear?
And using data is my mostpopular like topic that mostly
law enforcement who subscribewant to hear about.
So it is something that they'revery interested in and want to
learn more about, you know,using data and tech and tools,
but you know, and I think thosearticles especially are the most
difficult to understand as alayperson.
SPEAKER_01 (09:41):
What kind of feedback have you gotten on all
of this?
SPEAKER_00 (09:44):
As far as feedback, most of it's been positive as
far as bringing light to some uhfindings that have, you know,
that people were not aware ofbefore.
I add a lot of storytellingelements, or if I have personal
experience, I'll add that andpeople, you know, like that kind
of realism in there.
You know, I have got, especiallywhen whenever I bring up AI,
(10:04):
there's always a little bit oflike the debate of, you know,
usefulness or, you know, is ithere, take our jobs, you know,
like there's always some sort ofkind of squabble uh there.
But most of the time thefeedback's been very positive.
There is a few resources outthere that do something similar.
I know the um like the appliedpolicing briefs, Dr.
Radcliffe has his podcasts wherehe interviews people as well.
(10:26):
So there is resources out therelike like mine, but there's not
enough, I would say.
And so I think people are veryhungry for this kind of
information.
SPEAKER_01 (10:35):
Yeah, absolutely.
So can we can we dive into whatyou found?
I'm putting you on the spot.
I don't know that you have anencyclopedic memory of all of
the things you've written, buthave any of the research topics
that you've covered reallyjumped out as like either
unexpected or have you thinktranslating it to a wider
audience has had a biggerimpact?
SPEAKER_00 (10:56):
I think, okay, so one I will use I always
surprised when I'll thinksomething like, I don't know,
this will be popular, but I wantto get this out there and then
it explodes.
One, I did a post on datavisualization and what's the
best way to display graphs in aComstat meeting, you know?
And something as a crimeanalyst, I was very like die on
(11:18):
this hill to, you know, let'snot do week-to-week comparison,
what's up, what's down.
I think this was one of ReneeMitchell's articles doing a
study of like, hey, let's showthese different types of data
visualization graphs to a lawenforcement leadership and see
if they can make decisions, whatwhich graph helped them make the
better decision in terms ofresponse?
(11:39):
And had did they understand itbetter?
Instead of like, let's just donormal percent change graphs,
let's do one that shows themedian for this time and a
standard deviation above andbelow, you know, to show, hey,
crime does fluctuate, but hey,let's oh, this is above our
norm.
This is about out of thestandard one standard deviation.
Maybe we should do somethingabout that.
(12:01):
And, you know, with that kind ofgraph, leadership was able to at
least deploy a better solutionto address that increase in
crime versus just using a normalpercent change.
And which maybe this is not themost exciting finding, but for
me, but like that's somethingthat that I don't think law
enforcement is aware of of like,yeah, there is a study on what's
(12:24):
the best data visualization touse to help you make decisions.
And maybe that there's betterways and there's not so bad, you
know, not so good ways.
And that, I mean, that wasprobably one of my more popular
posts.
A lot of people were commentingon it and sharing it and
reposting it and showing thatlike this is of some great
interest.
And I don't think a lot of lawenforcement was aware that
there's other ways to do thisthan just doing percent change.
SPEAKER_01 (12:45):
Sometimes I'll post a graph that'll be like percent
change because that's it's justeasiest sometimes that it's for
the public audience.
And if you'll be like, well, youshould really use z-scores.
Like, if you can like trying todescribe that to an audience, I
mean it's very difficult.
SPEAKER_00 (12:59):
How do you balance it to I made the mistake as when
I was a crime analyst, I I toldmy chief I was like, oh, and the
poison z-score is this, and thenthey never let me forget about
it for years on end.
You know?
I mean, it was like they wereimpressed, but also like you
know, sometimes you don't usethe terminology and just keep
(13:20):
the stats language out and tryto keep it as plain language as
possible so they understand.
SPEAKER_01 (13:24):
Yeah, it it sounds fancy.
Yes, not really, but one one ofthe things, uh just again going
by like some of the researchthat's available, uh a couple of
them, especially the recentones, talk about sort of the
role of real-time crime centers,the role of cameras.
Have you uh like synthesized abunch of these?
And um, I'm just looking at likeuh there was one on I I forget
(13:48):
the name of it, but ha theactual like review of real-time
crime centers, one onprioritizing burglary cases that
have CCTV.
Have you thought about likesynthesizing some of these?
Have you put all of themtogether into like a larger
analysis in your head?
Or how does that work?
SPEAKER_00 (14:03):
Yeah, so at least with the RTCC camera ones, I
worked for a long time in anRTCC, so like that's a that's a
really good background I canpull from as far as like I know
how these work and how whatwould be useful for a law
enforcement agency to know aboutRTCCs.
I have an interest.
I do want to so like what yousaid, like pull together all the
resources on one topic, um, dolike basically like a literature
(14:27):
review like on RTCCs, likewhat's all the research out
there.
I do have a goal of doing that.
Maybe RTCC would be my firsttopic.
It's just it takes a long timeto be able to find all the
research that's out there, pullit together in a in a cohesive
form that doesn't make it into a50-page report.
You know, I I have been workingon that and I do hope to
(14:48):
eventually post at least uh acouple of those a year, or maybe
like one a quarter, uh, of like,hey, this is all the current
research on this one topic, uh,you know, AI report writing or
uh, you know, RTCC camera-basedtopic.
That is a future goal, notcurrently what I have right now.
SPEAKER_01 (15:07):
Can you sort of give me your impressions of sort of
RTCCs and kind of that synthesisof what the research shows and
not just RTCCs, but the theseare real-time crime centers, um,
but just like surveillancecameras and like the role that
surveillance cameras can playfor police departments with all
of this experience?
Have you sort of do you have ageneral idea of like what's the
best way to balance success andand privacy concerns and all
(15:31):
that?
SPEAKER_00 (15:31):
As far as so for real-time crime centers, they're
still relatively new in terms ofthe evidence base.
Um we are seeing some researchcome out that does show as far
as like case clearance-wise,they can help increase, you
know, the solving of cases,which logically that makes
sense.
There's more cameras, you'remore likely to capture evidence.
With it being staffed, your eyeson the scene faster.
(15:54):
And you know, the sooner yourespond to the scene, the more
likely you are to solve cases.
Especially that in that case,RTCs really shine.
There's a lot of other metricsthough, the RTCCs like, you
know, whether it's officersafety, you know, awareness of
what's going on in yourcommunity.
It's really hard to capture withdata.
Um, and so that's that's aharder point of like when you're
(16:17):
building an RTCC, you reallyhave to figure out what do you
want this to do for your agencyand know that is and understand
that metric might not be easy tocapture.
There is evidence that arecoming out that at least helping
solve cases, they can do that.
Now, crime reduction has variedresults, but there's not a lot
(16:38):
of research out there yet onevaluation of RTCC.
Now, cameras have been around alot longer than RTCCs, and
again, more on that case, theydo help with case clearances.
Now, crime reduction, I thinkthere is varied results
depending, you know, if thecamera is visible, it can't work
as a deterrent, but you know,does that push it off to other
(16:59):
areas?
It it kind of depends on theresearch in the area.
Um, but I think overall, youknow, real-time crime centers,
license plate readers, puttingall those technology pieces
together is, you know, there ismore evidence and it's starting
to grow that like, you know,this can be very helpful in
solving cases, in gettingsituational awareness in in your
(17:22):
community and what's happeningin your community.
SPEAKER_01 (17:24):
Looking more broadly, how do you respond when
sort of the anecdote leadsrather than the research?
When you run into people, you'reexplaining what the research
shows, and they say that theirlife experience or their opinion
says X and actually the datasays Y.
How do you have that discussion?
SPEAKER_00 (17:42):
Yeah, so I mean, always validate the antidote.
I mean, people have their ownexperiences, and especially with
policing, officers have lived itand experienced it.
And so, you know, theiranecdotes are are valid in what
they're experiencing.
And I also, with differenttechniques, different technology
works differently in differentcities.
(18:03):
What works in one city might notwork in another, or the city has
to adjust because theircommunity is not the same as
Chicago is, you know, Chicagoand Oakland are not the same.
And so what might work in onemight not work in the other.
And so it works for them becauseof their personality and who
they interact with, and maybethey have rapport, where in this
(18:23):
other area, like other city, theofficer doesn't have the same
rapport and it doesn't work.
And so that is like I I hesitateto shut down anecdotes because
things will differ.
And also the field of policingresearch is relatively new.
And so we haven't, you know, onestudy might say this doesn't
work, but if we look at it in adifferent context, it might
(18:43):
work.
And so to shut down anythingmeans like I I hesitate to to do
that.
SPEAKER_01 (18:49):
Is there a question or a data set or an answer that
you it isn't answered somethingthat isn't answered by policing
research that you really wishwas available?
Like is there something thatyou're just dying to write about
and think about, or some thingsthat you're dying to write about
and think about, and it justthere's just nothing there?
SPEAKER_00 (19:06):
I mean, okay, so I I do a lot right now.
I have a big project on AI andpolicing.
So it's very new.
So we don't have a lot.
Um so loved more of like a lotof this technology, more of how
is especially like dataanalytics, integrated data
analytics, how does you know uhthat help law enforcement if you
(19:28):
bring all these data sourcestogether and you know, give them
LPR and their CAD data and theirRMS and their neighboring
agencies and you know, OTSINTand all that stuff, you know,
does that really help them solvecases?
There's nothing on that now, youknow, and so and a lot of it it
has a lot of potential.
And as an analyst, I'm like, Ican see the potential of it, but
(19:51):
you know, there's no research onit.
And then also like license platereaders, there's not a lot of
license, there's not a lot ofresearch out there on them
because a lot it's verydifficult to identify like every
agency who dubs on LPR, I alwaystry to say, hey, you know, if
you make an arrest, make a notethat it was because you pulled
(20:11):
it from LPR, it was an LPR hit.
And that's always been difficultto to uh you know get across to
agencies of like track that infoso you can prove that LPR works
for you.
It helps you, you know, withyour auto theft cases or with
you know your warrants andwhatnot.
I mean, I even talked to acouple guys at Flock Safety who
are like trying to get agenciesto track this data, and it's
(20:34):
it's very difficult even fromtheir end.
So that's something also withLPR and AI and how that's
changing the policing field, weneed so much more research on.
SPEAKER_01 (20:45):
Sort of at the other end of the spectrum, how do you
handle research that you knowit's not obvious what the what
the answer is, what the findingis, or potentially it's a really
interesting finding, but it'spotentially not the strongest
finding, or even potentially abad finding.
Like obviously there's a widerange of quality and research.
How how do you sort of workthrough that?
SPEAKER_00 (21:08):
What comes to mind is um so Jerry Radcliffe ran a
study in Philadelphia.
SPEAKER_01 (21:14):
Also also a former podcast.
SPEAKER_00 (21:17):
Uh he he ran a study in Philadelphia where they had
um, I think it was like withintheir transit unit, they signed
a um like a social worker withan officer.
And especially when this was inthe height of like alternative
policing programs, instead ofhaving the officers show up, you
have someone else show likesomeone that's of either a
(21:38):
psychiatric care social workershow up.
And I think, yeah, like his hisfinding was like you know,
officers who like the officeronly actually had better
results, or they're about thesame result as an officer with a
social worker.
And when you look at it on thesurface, you're like, oh, so
(21:59):
social workers.
Don't really work.
But when you dive deeper intohis study, he, you know, brings
out the point of, well, okay,these actually these officers
who are working have beentrained in like multiple things
of like with like a socialworker.
Like many had a social workerbackground, and they've been
trained in like how to interactwith people who are, you know,
(22:20):
having mental episodes, who aredealing with addiction and
drugs.
So they've been speciallytrained.
And then the social worker side,there was a lot of turnover
because the way the position wasadvertised, you know, they
didn't realize how how much workwould go into it.
They just had a really hard timekeeping these social workers
employed in these shifts.
(22:41):
And so it's kind of like, well,does that speak to the, you
know, social workers don't work?
Or does it speak more to, well,these officers are specially
trained so they do better, butalso these social workers
weren't really given a good shotbecause they had a lot of
turnover, they didn't have agreat amount of training.
And so, you know, so that's likea mixed finding where it's
(23:02):
brought out of like maybesomething we can learn from this
is when you're implementingsomething, make sure it's given
its best foot forward.
And so that's something I Iemphasize a lot, and sometimes
in implementation research thatI talk about is if you're
implementing a solution,consider all these factors.
Because sometimes a lot of timesI see with law enforcement,
they'll just let's just startsomething, get it going, run,
(23:24):
run with it.
And then they there's a lot ofissues that pop up, and then
they say, well, it didn't work.
And you're like, well, it mightnot have worked because it
wasn't implemented well.
And so that's you know, alwayscome kind of a consideration
when we see mixed findings arenot working is how well was it
implemented?
How can we better implement it?
You know, was enough forethoughtand consideration put into it at
(23:45):
the at the beginning for it to,you know, actually have a
chance.
SPEAKER_01 (23:49):
Which sort of brings up the question of how do you
how do you communicate to youraudience that like, here's what
we found, this was a study, itit's not like every experience,
the you know, your results mayvary, sort of thing, in terms of
this is what the research shows,but the research is not perfect.
How do you how do youcommunicate that to people that
don't they say, oh, somebody didresearch, it must be true.
SPEAKER_00 (24:11):
So for the instance for the uh Radcliffe study, I
did display the results of like,okay, you know, they're either
like the same or the officersactually did a little better.
But I did go into as a part ofreading the full study, is going
into, hey, these were thedownsides of like basically kind
of what I just told you of likethe social workers had a lot of
(24:35):
high turnover, they didn't haveclear enough job description,
not a lot of training, and theseofficers have had a lot of
training that more than generalofficers do.
So I brought it a point of like,if you're interested in doing
this program, even though it waskind of a a null, that here's a
ways that maybe you can havemore success with it.
I also write a lot of differentone studies looking at
(24:55):
implementation science and kindof like stepping up like law
enforcement of like if you'retrying to do a reform on a
certain thing, hey, this is agood case study of like how to
do like how to do reform, how tohow to implement a solution and
do it in the right way the firsttime.
Talk more about how they did itversus the findings.
(25:16):
Um I I kind of we'll talk moreon that.
SPEAKER_01 (25:19):
The the last kind of like big area I want to get into
is sort of near and dear to myheart, and I think you've got
some experience with this, whichis um police recruitment and
retention.
And I'll sort of set the stagefor the audience.
Um, I'm sure you know this, butthe vast majority of police
departments lost a ton ofofficers between 2019 and that
sort of 2022 to 2023 period, andare really struggling to start
(25:44):
to grow.
And I have a substack that I'lleventually publish that it looks
like they maybe started to growa little bit in 2025, but still
most big police departments arewell below where they were six
years ago.
And obviously, from a umresource allocation standpoint,
how many detectives you can puton the streets, how your
response times will look, howyour clearance rates will look,
(26:05):
all of these things areimpacted.
Your service delivery, policedepartments as service delivery
organizations are very muchimpacted by the fall of
officers.
I swear there's a question inthis.
Um What have you found?
What are the the key issuesfacing departments?
How can they start to grow andstem the tide of resignations
(26:26):
and of and improve retention?
Is is there research findings inthis?
SPEAKER_00 (26:31):
Yeah.
So there is a lot of recentstuff that's come out in terms
of policing retention and even alittle bit of what I've been
involved in.
I mean, from what I've read andfrom what some of the projects
that I've done around policeretention, the big number one is
leadership.
Is for one of my projects, I dida deep dive in like Glassdoor
(26:51):
and uh Reddit data and lookingat police retention and what
officers were saying in regardsto why are they leaving or you
know, why, you know, why whythey're considering and leaving
and leadership, the way thatleadership treats them,
especially when there's anincident that happens, and then
a lot of times officers feellike, oh, so-and-so was thrown
(27:12):
under the bus as a scapegoat, orI feel like because I received a
complaint, now I'm being putthrough the ringer, I'm I'm
guilty before proof and youknow, I have to prove my
innocence.
That kind of feeling was atleast what I've been seeing is a
big driver of uh retention.
I know I've posted a couplestudies that show that, like,
(27:33):
you know, one of the big pointsfor like officer wellness is
like, oh, PTSD, you know, you'rebeing exposed to very traumatic
incidents, especially as a lawenforcement officer, you know,
that's driving that.
And while that PTSD is a majorcomponent of officer wellness
that needs to be considered andthat needs to be um, you know,
given resources for andunderstanding for, that these
(27:55):
organizational factors of howyour leadership is, how your
scheduling is, how how safe youfeel talking to your sergeant or
to your lieutenant, you don'tfeel like your organization
supports you, you're way morelikely to leave.
And also way more likely to feellike your PTSD symptoms like are
exaggerated.
There's a lot of research outthere that a lot of times it's
(28:16):
not just like, oh, it's becauseof what happened, they can't
handle what's happening out inthe field.
It's actually because they don'thave the support of their
leadership behind them, but thesupport of their organization
behind them that's driving themout or driving them to have, you
know, burnout.
Good thing is is that can bechanged.
I definitely had a few, a fewarticles on that.
SPEAKER_01 (28:35):
Going, you know, full into opinion mode.
Do you think that it's possiblefor police departments to grow
substantially, or is this justnot an environment that's
conducive to that and we have tolearn to live with what we have?
SPEAKER_00 (28:47):
Um I mean, I would say like mass hiring, I mean
never is really that good of athing, but the way the recruit
matters.
And I think as far asrecruitment-wise, especially
with like the efforts of 30 by30, which is if you're not
aware, is you know, for in theUS to get if you're a 30 by 30
(29:10):
agency, you're trying to get 30%of your recruits um to be women
by 2030.
There's a lot of different likerecruitment efforts are towards
that way.
It's a little harder to say ifthere is like a growing mass
that you'll be able to get tothe numbers you were in 2018.
I know some departments are,some departments are back to
where they were.
And definitely the departmentsthat I've spoken to have either
(29:32):
done some wellness initiativesor really made sure to keep
their guys and focus onretention.
And retention bonuses don'tnecessarily work, but you know,
to make sure that their officersare heard and more of a focus on
that versus recruitment.
But if you're down on yourofficers, I mean, you need to
focus on recruitment and try tofocus on diversity recruitment.
(29:54):
And there is a few studies outthere, at least, that show,
depending on like the videosthat you have, what type of
traits you show will change howwhat people who will apply.
SPEAKER_01 (30:06):
So if you're show, you know, showing a lot of the
the SWAT to your yourpromotional video of everybody
smiling and doing their hardwork.
SPEAKER_00 (30:13):
So, you know, there is studies out there if like if
you show, you know, officers whoare helping the community and
yeah, they are doing thesmiling, like playing with the
kids, but also, you know,helping the community members
and showing that kind of front,you're more likely to get more
women.
Um, versus if you're showing thethe SWAT machine and the
helicopter and the guns andlike, oh, we're gonna beat the
(30:36):
bad guys, you're less likely toget women.
And so, like, there's differenttechniques depending on who
you're trying to recruit for.
SPEAKER_01 (30:42):
Well, that's fascinating.
And uh it's all great.
Um, policeresearchhub.com is thewebsite, right?
SPEAKER_00 (30:49):
Yes, yes.
SPEAKER_01 (30:50):
Yeah.
So so what's next?
SPEAKER_00 (30:53):
My goal is to keep continuing posting, finding all
these the research articles thatat least I think I find
interesting, and I knowhopefully other people will find
interesting and keep going withthat.
I'm I'm also pretty prevalent onLinkedIn as well, and I will
post questions to the group andsee who's what's it people are
interested in.
And uh hopefully eventuallywe'll be able to do those deep
(31:14):
dives on a single topic becausethat's like my next big goal
that I want to do on that.
SPEAKER_01 (31:19):
It's wild how like each little community has found
different websites now in 2026,2025.
So, yeah, it's a vibrant policeresearch community on LinkedIn.
So and you're one of the moreprolific posters in talking
about the research, so it'sgreat to follow.
All right, Carly, thank you somuch for joining the show and uh
keep up the great work.
It's it's fabulous.
(31:40):
Uh, police researchhub.com.
SPEAKER_00 (31:42):
Thank you for having me.
SPEAKER_01 (31:44):
Thanks for listening to the Jeffalytics Podcast.
Be sure to subscribe and tolearn more, head on over to
ahdatalytics.com for moreinformation and previous
episodes.
If you like what you heard,please leave a glowing review,
which will help others todiscover the show.
Till next time, I'm Jeff Asher.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Breakfast Club
The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy, Jess Hilarious, And Charlamagne Tha God!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.