Crime updates qualitative research

Through my Human Centered Design and Engineering (HCDE) master’s degree, I worked with a research team to investigate the impact of civilian-reported crime on people living in Seattle. We conducted this work through a two phase research project.

Timeline

September 2023 to December 2023

Team structure

4 user researchers

My role

During this project my main roles included: 

  • Conducting netnography and user interviews

  • Leading the creation of note taking, analysis materials, and communication templates

  • Creating content for our presentation deck

I come from a content design background, so I played a large role in making sure anything we shared with participants was clear and concise.

Objective

Understand the impact that crime updates have on people in the Seattle area. 

The process

Phase 1

Research questions

  • What are the motivations for community-sourced crime reporting?

  • How does bias show up in community-created crime updates?

  • How are posts similar and different by platform?

Methods

We conducted netnography to gain a high-level understanding of crime updates across platforms and what emotions they evoke. We review crime reporting posts and responses across platforms, including Reddit, Nextdoor, and Citizen. Next, we used thematic analysis to identify similarities and differences between responses.

Participants

We reviewed posts on Next Door, Citizen, and Reddit, since these are the top applications in this emerging trend of real-time communal crime reporting used this criteria to identify relevant posts for our project. The post aligned with this criteria:

  • Posted within the week of Nov 16, 2023 to Nov 23, 2023.

  • Limited to Seattle

To stay within the timeframe of our project, we identified the top 5 crime-related posts. 

Findings

We identified these main insights from our phase 1 research:

Granularity: Differences in the types of responses depend on the granularity of the community. More granular, less anonymous platforms have more of a community feel and direct emotional engagement with the situations.

Proximity: Proximity to the crime affects the responses. With location-specific alerts, users generally reacted more fearful and inquisitive of the crime since it might endanger their safety in the moment. 

Anonymity: More anonymity leads to less accountability. Anonymity influences the responses of participants. The closer your username is to your identity, the more accountable you become. 

Support: Platforms introduce novel way of supporting a fellow netizen. We observed a novel way in which netizens support each other via “algorithm-gaming”, a tactic that did not exist previously in offline communication.

Phase 2

Research questions

  • How do crime civilian-sourced alerts affect the emotions of people in the Seattle area?

  • What aspects of crime alerts lead to negative impacts on mental health? 

  • How can individuals' mental health be supported while they utilize crime alerts?

Methods

For this phase, we conducted interviews to dive deeper into participant thoughts and actions. We were able to get insights that weren’t possible during netnography. Initially, we sent out a screener to identify potential participants’ approximate engagement with the platforms. After recruitment, we conducted five interviews to learn in-depth about each participant’s perspective. Then, we conducted thematic analysis from transcribed interviews.

Participants

We used this criteria when recruiting participants:

  • Uses a crime alert platform with an engagement aspect such as NextDoor, Citizen, etc.

  • We wanted to observe the community engagement aspects, so we excluded tools like UW Alerts and the traditional newspapers

  • Regularly uses it every 2 weeks

  • Lives in Seattle because Seattle has a higher crime rate

Findings

We identified these main insights from our phase 2 research:

Participants look to top comments

The accuracy and truth of communal updates is likely to be distorted since it promote sarcastic and negative remarks. Traditional news sources are vetted, while communal crime updates lack fact checking


Humor and sarcasm were used as tools

Online commentary often featured dismissive language or presented crime as entertainment, creating a disconnect with real-world unease.

Participants seek local, realtime safety information

Being closer to a crime made participants more likely to read/check an update and also display anxious behaviors.

Dark practices exploit helplessness

Situations where people use Citizen can be life-or-death scenarios involving violence or firearms.

Platforms are withholding crucial information from consumers as a means of monetizing their app and increasing revenue.


Recommendations

These are my recommendations to companies who create and develop crime updates platforms.

Design:

  • Make users identify with their actual name to promote accountability

  • Provide mental health resources to help people recover from the anxiety they experience

  • Add platform moderators who can check for quality and accuracy of crime updates

  • Stop promoting paid subscriptions during vulnerable moments (e.g., live crime streaming). Find other moments to promote paid subscriptions.

Future research:

  • Need to validate the impacts of both social discourse and informational public safety platforms. A survey would be an effective next step.

  • We advise further research exploring how dark patterns are utilized on other platforms.

Impact

After implementing design updates to the crime platforms, I’d recommend follow-up user interviews to understand if users are:

  • More confident in the accuracy of posts

  • Less uncomfortable with the promotion of the paid service

Quantitative measures that could also be tracked, include:

  • User sentiment: How are users feeling after they read crime updates?

  • User retention: After reading crime updates, are users staying on the platform?

What I learned

We had limited time to conduct our research. If I conducted this research again, there are a couple of things I’d do differently, including:

Scoping down: Initially our research plan scope was large with the amount of time we had. I think we could have been even more focused by only reviewing one or two platforms.

Recruiting a range of participants: To get additional perspective, I think we could have also recruited interview participants who actively post and participate in community crime reporting.