n defining the goals and desired outcomes for the Agency Inventory of Services (Agency Survey), the committee addressed several questions. Should the survey be a comprehensive assessment of specific victim service agencies, including what they are providing, to what degree, and to what type of victim? To determine service trends and gaps, should the survey provide a general picture of what is available to victims of crime in the city of Denver? What is the sample group? Should all or only a few representative agencies be studied? If a few, how will those agencies be selected? Most important, what accomplishments were hoped for with these service data?
The Agency Survey Committee took its direction from the goals and objectives of the Denver VS2000 project. In this case, the goal was to establish a seamless, interdisciplinary system of service delivery for victims of crime, including victims from previously underserved or unserved populations. The assessment objective for this goal was to evaluate the current delivery system for gaps, the quality of service delivery, linkages between agencies, and service to underserved groups. Ideally, the resulting data would inform the development of the essential components of a victim service model, recommendations for new program development, and recommendations for realignment of existing services. As such, the committee determined that the assessment should not be an evaluation of each distinct victim service agency. Rather, the assessment should provide a general picture or snapshot of Denver's victim services environment, highlighting gaps in services that could be prioritized and addressed by the model victim services network.
The decision to focus on general service trends rather than individual agency performance was important for another reason. Some victim services providers receiving the results of this research would also determine the changes to be made in the victim service delivery system based on the research they designed and implemented. Some victim service providers were also subjects of the survey. The committee asked service providers to participate in this project, which would take substantial time and attention away from their jobs of providing direct services for an unforeseen length of time. With limited resources available to them, victim service providers often work in relative isolation from one another and may be in a position of competing for funds. In the light of these factors, the Agency Survey Committee decided that a survey that seemed to be scrutinizing the performance of individual agencies would be counterproductive to an environment of mutual trust, collaboration, and commitment to the goals of the project.
The committee determined the survey's target population to be the agencies and programs providing victim services in Denver. To get a clear picture of service gaps and understand who was using services, the committee decided to gather information from as many programs as possible, including criminal justice-based programs and allied professional agencies with victim service components, such as hospitals.
Another parameter of the Agency Survey was that it include both objective and subjective data. Objective data included such items as services provided, number of clients served, client demographics, and number of staff. Alone, these data do not reveal unmet needs or gaps in services until compared with a list of predetermined needs. To obtain information about service gaps, the survey asked service providers where they believed the gaps to be. The survey asked what services were needed but not provided, not just for the victims at their agencies, but for any crime victim in Denver. These subjective data relied on the perceptions and beliefs of service providers. Though the committee knew such information was potentially biased, it included it in the survey to harvest the wealth of anecdotal and firsthand information that service providers have about victims and the services available to them.
When designing the survey, the committee considered factors that would affect its return rate. These factors included the length and complexity of the survey, its importance to the respondent, the respondent's motivation for completing and returning it, and the respondent's perception of an obligation to complete and return it. The Denver VS2000 Agency Inventory of Services is a lengthy 10-page document with a fairly complex format. It includes check boxes, grids, scales, and space for written comments. The survey designers tried to balance the need for comprehensive information with ease of completion.
Although the survey's primary objective was to reveal gaps in service and weaknesses in interagency linkages, it also included questions on other types of information such as agency mission, distribution of staff and board, service fees, hours, funding sources, reporting rates, and client demographics. The design was based on an assumption that service providers are people who often deal with paperwork and data gathering and would not be daunted by a complex, lengthy survey. Further, the motivation to complete the survey was assumed to be high because it was targeted primarily to participants in the VS2000 project, who had collectively identified the necessity of a needs assessment.
Categories of Information in the Agency Survey
The Agency Survey collected information in four categories: agency information, client information, service information, and agency linkages.
Agency information collected in the survey included date of inception, mission, involvement in interagency protocols, funding sources, number and ethnic distribution of the staff and board of directors, service area, accessibility, service fees, and marketing of services. The survey asked questions that inventoried the characteristics of the agencies that were providing victim services in Denver. These questions were included with the thought that the information might prove valuable at some point during this 5-year project. The committee designed the survey with the assumption that having as much information as possible about existing services would be helpful when developing a model service network. This portion of the survey did not address service gaps.
The survey asked for client information including gender, racial, and ethnic distribution; age; disability; and income level. The survey also asked respondents what percentage of clients reported the crime to police and what the respondents thought were the primary reasons for nonreporting. The survey designers tried to make the survey easy to complete and easy to analyze. Wherever possible, the survey provides a choice of answers with boxes to check or lists of characteristics with spaces to fill in numbers or percentages. These documents made the survey easier to analyze, reduced the amount of writing required of the respondent, and increased the likelihood that each question would be answered.
The section of the survey on service information was formatted into a table or grid. This design captured relevant information about each type of service available to victims of crime, such as how many victims are using it, who is the target population, and what services are needed but not provided. The survey designers developed a comprehensive list of 55 different services available to crime victims. The table format allowed the survey to collect a range of information on each distinct service offered by an agency or program, while keeping the survey a reasonable length.
The Agency Linkages Chart is another important part of the survey. Sometimes victims "fall through the cracks" of a service network due to weaknesses in communication and collaboration among victim service agencies. The Agency Linkage Chart was designed to reveal the general interagency environment. In the survey, "linkage" is defined as "a high or appropriate level of communication, resource, and information sharing between agencies; a clear knowledge of each other's services; a referral mechanism; and trust that referred clients will be treated appropriately and have relatively easy access to quality services." The Agency Linkage Chart contains a list of 80 victim service programs and allied professional and community groups. Respondents were asked to measure the level of their agency's linkage with each group or agency.
Implementation of the Agency Survey
The Agency Inventory of Services and a cover letter explaining the purpose of the survey were mailed to 99 victim service programs and allied professionals. Respondents were given 1 month to complete the survey and return it to VS2000 staff. VS2000 staff placed reminder calls to each of the agencies to remind them of the deadline and ask if they had any questions about the survey. Of the 99 programs, approximately 20 told VS2000 that the survey did not apply to them because they did not provide direct services to victims. Of the remainder, 47 surveys (59 percent) were completed and returned to VS2000. Data were computed using EpiInfo statistical software.
Results of the Agency Survey
Analysis and summary of the data collected by the survey focused on areas considered most relevant to the VS2000 project. These areas were client and staff demographics, reporting rates for different types of victims, agency linkages, and service gaps. For the most part, the survey results were not surprising and were in accord with anecdotal information and findings of previous surveys. However, some of the survey findings were interesting. Client and staff demographics indicated that overall services reached a broad range of victim types. By contrast, racially and ethnically diverse clients, staff, and board members fell into the 0-25 percentile range, which is lower than the percentage in the general population. The survey also revealed that, compared with other types of crime victims, reporting rates were shown to be significantly lower (10-26 percent) for adult survivors of sexual abuse, women with disabilities, domestic violence victims, sexual assault victims, non-English-speaking victims, and gay, lesbian, bisexual, and transgendered victims.
Analysis of agency linkages found that the top 10 agencies and groups requiring the most linkage were those serving American Indians, new immigrants, and non-English-speaking victims, as well as the Department of Social Services, hospitals, the faith community, and community leaders. Surprisingly, most clients selected information and referral as a needed but unavailable service, whereas most agencies responded that they provided this service. Discussion with service providers on the VS2000 working teams revealed that they felt information and referral were not being done as well as expected. Specifically, interagency knowledge was lacking and referrals were often not appropriate. Other service gaps identified by survey included emergency shelter, case management, on-scene crisis intervention, legal assistance, and 24-hour hotlines for deaf and non-English-speaking victims.
Evaluation: Suggested Changes to Agency Survey Content and Method
After analyzing the survey results and obtaining feedback about the process from service providers, VS2000 staff determined that some changes in survey content and method could have resulted in a better return rate with more concise, meaningful information.
Several survey respondents felt the survey was too long, and the tables and charts were intimidating and hard to understand. Although many respondents did not share this opinion, the VS2000 staff felt it was safe to assume that respondents would more likely complete a shorter survey. The survey information on agency characteristics was not summarized and has not been used by the VS2000 project during its 5 years of existence. Considering that the primary goal of the survey was to measure services and reveal service gaps, the agency information was extraneous and should have been omitted.
One of the tables, the Service Information Table, measured the services provided by a particular agency and the general gaps in services for all victims by all agencies, from the providers' perspective. Some respondents thought they were to indicate only the gaps they thought existed within their own agency or only for the types of victims they served. They were reluctant to list gaps because they thought it might reflect poorly on their agency or it might result in an increased expectation for their agency-to expand or to change services in ways that would be unrealistic or outside the agency's mission. The staff realized that including some simple, open-ended questions would have been a good idea, even though they would have complicated the analysis process. Simple questions such as "How do you think services for victims of crime could be improved?" would have generated valuable information and avoided confusion about how a question should be answered.
Another chart that measured linkages between victim service programs indicated which programs had excellent or sufficient linkage with the network of agencies and which did not. This information was helpful in illustrating which agencies might need to do more outreach, networking, and cross-training. Respondents also began requesting analyses that would show how each agency rated itself. Although this would be valuable information for each agency to have, providing the information would violate the confidentiality of the responders. If a statement of consent had been included in this section of the survey, sharing this information with the requesting agency could have been allowed.
OVC Bulletin, October 2000
Denver Victim Services 2000 Needs Assessment