What this part is for
Part 2 mapped what criminology already knows about how people use criminal justice technology. This part maps the holes. Each entry below is a UX research-shaped question that, in our reading of the literature, has not been answered in a way a designer, a vendor, or a procurement officer can actually use. The bar is the same bar we used for the bibliography: a question only earns a slot if a determined researcher could move it forward in a single quarter, with one user type, one method, and a defined decision on the other side. If we could not write a feasibility note in two sentences, we cut it.
We did not try to be comprehensive. Eighteen is enough to cover the nine subject areas at two questions each without padding. The order inside each cluster reflects our own ranking of "if you only did one of these, do this one" — readers should treat that order as a starting point, not a verdict.
One user type, one method, one quarter, one decision. If we cannot describe all four in two sentences, the question is not ready.
How to read an entry
Every problem below carries four labels: the user, a method that fits the question (not the only method that could work), a feasibility score, and a seed citation that points into the bibliography. The feasibility score is rough: tractable means a single researcher could ship a credible study in a quarter, moderate means two quarters with access negotiation, and hard means the access problem is most of the work.
Inmate tablets and in-cell technology
1 · The kiosk-to-tablet handoff for newly admitted people
Most jails introduce tablets and kiosks during the first seventy-two hours, when a person is dealing with intake, withdrawal, and disorientation. The literature shows tablets are used heavily once people are settled (Open Campus 2023), but the onboarding moment itself is undocumented. A short field study of intake-period tablet use would tell vendors which screens to redesign first.
2 · The family payer experience
Tablet economics depend on family members loading funds from the outside, yet the family-payer journey — finding the right kiosk, navigating fees, recovering from failed transactions — has barely been studied as a UX object (McKay 2022). A diary study with ten paying families would surface friction that no in-facility study can see.
Parole and probation check-in apps
3 · False-positive remediation in GPS supervision
GPS check-in apps generate alerts that are mostly noise; officers triage them in seconds (Armstrong & Freeman 2011). The supervisee side of a false-positive — being called, accused, or violated for a signal drop — has no published UX account. Interviews with twelve people who have lived through one would yield a redesign brief on the spot.
4 · App-mediated reminders and missed appointments
Reminder apps reduce failures-to-appear in field experiments (Chohlas-Wood 2025), but the mechanism — which message, when, on which surface — has not been disentangled. A factorial usability test with parole clients would let an agency choose a message template with evidence rather than instinct.
Body-worn cameras: the officer side
5 · The activation decision under load
Activation rates collapse in high-stress encounters (Huff 2022; Lawrence 2019). What no study has done is co-design the activation control itself with officers — bigger button, audible cue, automatic triggers from holster sensors — and test it on a shift. The decision on the other side is a procurement spec.
6 · Footage review as a daily workflow
Departments accumulate hundreds of hours of footage per officer per month. How supervisors actually triage it — what they watch, what they skip, what they flag — is the bottleneck in every BWC program, and the literature is mostly silent (Myhill & Hohl 2023). A think-aloud study of review sessions would map the cognitive load directly.
Predictive policing and risk-assessment tool use
7 · What officers ignore on the screen
Officers selectively use the parts of a predictive tool that match their priors and ignore the rest (Brayne 2017; Sandhu & Fussey 2020). A structured screen-recording study during ride-alongs would quantify which UI elements get attention and which get bypassed — directly actionable for vendors who keep adding features no one looks at.
8 · Risk-score legibility for the person being scored
Risk-assessment outputs are designed for judges and analysts, not for the defendants whose lives they shape. A small usability study with defendants and their attorneys would test whether the score, the factors, and the appeal path are intelligible — a precondition for any due-process argument (Stevenson 2018).
Court e-filing and self-represented litigants
9 · The mobile e-filing experience
State e-filing portals are still designed for desktop, yet self-represented litigants increasingly file from a phone in a courthouse hallway (Sandefur 2019). A mobile-first usability sweep across five state portals would generate a comparable redesign brief for every court system in the country.
10 · Clerk-side workarounds
Clerks routinely build spreadsheets, sticky notes, and shadow systems around e-filing platforms (18F 2021). A contextual-inquiry study of three clerks' offices would surface the official software's most expensive UX failures faster than any vendor audit, because the workarounds are the failures, rendered visible.
Victim notification
11 · Notification timing and trauma
VINE and equivalents notify victims of offender movements, but the timing — late at night, during work, alongside other alerts — is rarely studied as a UX variable (Campbell 2018). A diary study with twenty victims would yield a timing-and-channel policy rooted in lived experience rather than vendor defaults.
12 · Notification accuracy from the operator side
The operator UI that staff use to set up, edit, and cancel notifications is the least-documented surface in the entire VINE ecosystem (McEwin 2002). A heuristic walk-through with operators in two jurisdictions would surface error-prone flows that show up downstream as missed or wrong notifications.
Tip lines and anonymous reporting
13 · The composition step on a school tip line
School-safety tip lines accept hundreds of thousands of tips a year, but the UI that prompts a student to write the first sentence has never been A/B tested in the open literature (Espelage 2021). A field experiment varying prompt wording could move signal-to-noise meaningfully with no new investment.
14 · Tip-room triage as a craft
Tip-line operators apply tacit rules to escalate or dismiss reports. Those rules are not written down (Hsieh 2022). A short ethnography of one tip room would let a designer build the triage UI around the work that is already happening, instead of against it.
Digital-evidence tools
15 · The disclosure-portal experience for defense
Prosecutor-built disclosure portals are designed around the prosecutor's workflow; the defense side is an afterthought (Miller 2022). A usability study with public defenders would surface specific failure points — search, previewing, bulk download — that have measurable effects on case prep time.
16 · Chain-of-custody friction in cross-org handoffs
Most digital-evidence errors live at the seams: lab to officer, officer to prosecutor, prosecutor to court (Hitchcock 2016). A service-blueprint study following ten cases end-to-end would map the points where custody metadata gets re-typed, lost, or silently overwritten.
Jail visitation platforms
17 · The call itself, not the call frequency
Video-visitation research is rich on whether and how often visits happen (Vera 2024), and thin on what happens inside a call: dropped audio, frozen video, lost minutes, awkward handoffs between children and incarcerated parents. A small in-home observation study with families would put numbers on quality where there are only volume counts today.
18 · The remote-visitation choice for staff
Corrections officers gatekeep the room, the schedule, and the device that hosts a visit. Their workflow has not been the subject of a published UX study (Tartaro 2017). A shift-along observation in one facility would surface staffing-pattern bottlenecks that determine whether the policy promise of "more visits" is real on a Tuesday.
Continue to Part 4 — Careers →