Dedoose Blog

BLOG

Dedoose as a Qualitative Criminal Justice Resource: Gang Desistance and Social Network Data Analysis

5/2/2021
.
We keep telling you Dedoose is capable of benefiting any type of research project, whether quantitative or qualitative. Today, we’ll review one mixed methods project that was supported by a near $1 million grant from the National Institute of Justice in cooperation with the RAND Corporation, Temple University Department of Criminology, Urban Institute, and Columbia Heights/Shaw Family Support Collaborative. In the study, Dedoose was incorporated to handle everything from establishing inter-rater reliability among multiple qualitative coders, to generating charts showing the coding intersections of reasons for exiting criminal groups and social network types of significant individuals in gang members’ lives.  
.

Studying Gang Desistance: Using Dedoose to Understand Why Members Leave Their Groups

.
In a nutshell, there is no “key reason” that a given gang member decides to part from their criminal group and lifestyle. According to Pyrooz and Decker (2011), “push” and “pull” factors are thought to originate from within or outside of the group and cause individuals to leave. Decker and Lauritsen (2011), on the other hand, have long held the perspective that many members simply “grow out” of gang membership over time. Indeed, Pyrooz, Decker and Webb (2014) have pointed out that the typical stint of a gang member is one year or less in duration. Sampson and Laub (2003), who are no strangers within qualitative research circles, have worked to develop well known “turning points”, such as gaining legitimate employment, getting married or having a child. These points in time, they argue, tend to sway members from leaving and appear more often in interviews with former members than others. Additional turning points include leaving a group because of the influence of a significant other in their life, which is an important point for the research discussion at hand, in terms of how Dedoose needed to be used to answer the target questions of the study.
.

Guiding Research Questions:

  1. What “reasons for leaving” are contained within gang member interviews collected from groups in Washington, DC and Philadelphia, PA?
  2. Do former members’ social networks appear to influence their reasons for leaving, as evidenced by the overlap of social networks and reason for leaving codes?
As you can see, Dedoose was needed not only to code gang member interviews, but also to see where codes related to reasons for leaving “ran into” codes created to designate a member’s social network influencer type (ex. family member, friend, boss/employer, invested other who is not biological family, spouse or partner). As a result, at a glance and in real time, the research team was able to piece together which types of influential significant others, tended to be mentioned more or less alongside reasons for leaving their criminal groups.  
.

Understanding Code Overlap With Dedoose: Identifying Reasons for Leaving and Influential Social Networks

Although the interviews were a key portion of the study, prior to the interviews participants were asked to complete a quantitative survey, with the final sample qualifying members as those who had previously met Euro-gang criteria for gang status while actively involved in their group. Participants were between 14-25 years old, with interview sessions lasting between a few minutes to more than an hour for several members. Ninety three percent of participants were male, with 72% of respondents coming from the Philadelphia, PA recruitment site. 
.
During the interview, former gang members were asked to identify their top three reasons for cutting ties with their criminal group. With respect to the data collection process, a social network component was executed across three waves spaced six months apart. In this portion, participants were asked to identify 20 people who they felt the closest to at the time of the survey. These lists were then compared by the team over time to detect changes in those who were recorded. 
.
Prior to data import of quantitative surveys and qualitative interviews into Dedoose, the team created example coding schemes using Decker and Lauritsen’s work as a guide. These lists were then used and shared by multiple coders to identify reasons for leaving in the interviews, including any that might have been mentioned by not specific to the deductive list. Using Dedoose, parent codes were created to designate types of social networks, ranging from Family, to Old Group, to New Group members. These were then split into child codes, such as Antisocial family relations under Family parent codes, or being “cool with” members under Old Group (even after exiting the gang).
.
.

Key Findings of Qualitative Analysis: Top Reasons for Leaving

Upon analysis of the coded interviews, Dedoose’s interface and code co-occurrence charts were used to gain insight into the most frequently cited reasons that gang members in the sample exited their criminal groups. Using exported data from simple Microsoft Word documents, the team charted for each participant which reasons for leaving were mentioned in each interview, along with the first one cited by mention and the reason that was most dense by frequency. 
.
Perhaps more interesting were the revelations from the code co-occurrence chart, which mapped side-by-side which reasons for leaving involved which type of social network member. For example, the code co-occurrence count might naturally be expected to be high between a reason for leaving such as parenthood, and the social network type of biological family member (ex. Son or daughter being mentioned during this component). On the other hand, it was possible for the team to see which type of social network tended to initiate and/or fuel the exit strategy of a given participant. This speaks volumes when it comes to generating evidence for data-driven policy making abilities, for instance. The study ultimately found that family relationships mentioned within the qualitative interviews (ex. not non-familial prosocial networks such as a coach or mentor) were important in the sense that these could be used to guide investment efforts and funding to support at risk youths.
.
.

Theorizing Gang Desistance and Social Networks: Top Reasons

Overall, more than 40 percent of those surveyed listed maturing or growing up as a top reason for leaving. This lends additional qualitative support for Decker and Lauritsen’s perspective, as well as the turning points put forth by Sampson and Laub to a lesser extent. A tie was observed between group changes and influential relationships, which refer to negative or new dynamics in an old group and experiencing the positive impact of someone close to the participant, respectively. 
.
.
Dedoose was able to illustrate how social network types overlapped with members’ reasons for leaving. Several key takeaways were reported as a direct result of being able to embrace this capability. When observing the reasons for leaving -which overlapped with parent codes for family, influential relationships, involvement with the justice system and parenthood- were the most commonly cited reasons for exiting a group. Mentions of members of a New group (ex. not criminal or gang involved) were more commonly associated with codes for justice involvement. This could signal that increasingly violent or police-involved activity experienced by a member can potentially serve as a “push factor” out of the group and into a new, law abiding one, instead. 
.
.
Those who focused most of the interview talking about Old Group (ex. Parent code) members, on the other hand, tended to focus on “the bad things” (ex. Child code) compelling them to leave the group, increasing dynamic changes becoming a problem, experiencing trauma, or becoming involved with the law. As you can see, Dedoose’s code co-occurrence chart enabled the team to eek out multiple layers from a single data source with simple rounds of data coding.  
.
.

What Can You Discover From Your Data With Dedoose? Let’s Find Out.

Regardless of whether you have heard about our tool before or not, we can assure you that even more exciting features, upgrades and options are on the way! We’re certain that our data capabilities are the perfect match for whatever project or research challenge you are facing now. 
.
What are you waiting for? Get in touch with one of our friendly team members (no joke- we’re actually incredibly personable). We can’t wait to uncover the possibilities that can come from teaming up with Dedoose! 
.
NOTE: If you would like to feature a LINK to the copy of this entire presentation that appeared at March 2017 Academy of Criminal Justice Sciences, you can use this: https://docs.google.com/presentation/d/1HHOp9hBOUyEE9jCSoMhqC4b6WRsMCBhwI_lukAkJcRY/edit?usp=sharing   ;
Back