The Study

The Safety Decision Aid is an effort funded by the National Institute of Mental Health through a five year, randomized controlled trial. Users are randomized to either the full decision aid or a control site. Primary outcomes being measured are exposure to violence and mental health.

The Study Logo
Recruitment for the study began in March 2011 with a target sample size of 720 participants. Since the decision aid is web accessible, recruitment efforts are also mostly web-based. Study participants must meet the following eligibility criteria:
  • English or Spanish speaking women over the age of 18
  • Currently in an abusive relationship with an intimate partner
  • Comfortable using a computer
  • Comfortable using the internet
  • Access to a computer that can be used safely
  • Has or is willing to get a safe email account
  • Lives in either Oregon, Missouri, Arizona, or Marlyand

Eligible participants are enrolled for a year long intervention in which they complete the decision aid at Baseline, 3 months, 6 months, and 12 months. With each completion of the decision aid, we hope to see an increase in safety seeking behaviors, better decision making process, improved mental health and reduced exposure to intimate partner violence compared to the control group.

The Team

As researchers in intimate partner violence (IPV) and decision science, we set out to develop and study this web accessible application to deliver tailored solutions for women experiencing abuse. Our team reaches across four states and three countries.

Participating Universities

The study team has coordinating centers located at Johns Hopkins University, Oregon Health & Science University, University of Missouri, and Arizona State University. We have also have domain experts from Kaiser Permanente and George Washington University. International expansion is expected to start in 2011 and 2012 with coordinating centers in New Zealand and Australia.

The Technology

The application utilizes an Analytic Hierarchy Process to present the five relationship priorities to the user who then performs a pairwise comparison of all the priorities. The resulting ranking of priorities drives the rules engine which determines what strategies best match the users' goal.

The Decision Aid Home

In order to deliver tailored strategies to each user, many standard and non-standard instruments were incorporated into the decision aid. With so many components to the decision aid, we focused heavily on usability and simplicity through the use of client side behaviors with jQuery and Flash.

In the usual course of completing the decision aid, the user must first complete the series of instruments, to include the Danger Assessment scale. Before getting any results, the user then completes the pairwise comparisons of the five relationship factors. A comparison must be done for each factor against all other factors. At that point, the user is then presented with the results to the Danger Assessment and their Priorties. Next, we take the combination of the users' priority, their Danger Assessment results, and their desire for their relationship (stay or leave), to deliver a set of standard and tailored safety strategies. For example, if the woman indicates that she wishes to leave the relationship and her priority is her Safety, we suggest that she get an order of protection against her partner if she indicated that she hasn't already done so. This strategy is accompanied by local resources based on the county in which she lives. The number of tailored strategies varies from user to user. Some may get as little as 5 strategies, others may get 25 strategies. Regardless of what is suggested to the user, the user ultimately decides which of the suggested strategies they want to consider including into their Action Plan for safety. The goal is that over time, the user will continue to work towards meeting their Action Plan for safety.