|
|
| |
|
 |
|
 |
Michigan
Outcome Identification Project
The data in this study are predominantly qualitative and exist in the form of notes made by the facilitator, notes (on flip charts) made by the co-facilitator, and audio recordings of the focus group proceedings. The goal of analysis was to identify recurring themes and patterns of responses in each of the domains of inquiry. Once identified, these themes form the basis of the analysis of consensus and difference among stakeholder group opinions.
The strategy for identifying themes and patterns of response across stakeholder groups was to transfer the focus group responses to each question from each focus group onto individual cards. This data was transferred directly from the flip chart sheets written by the co-facilitator and containing the participants priority votes. When necessary, the audio recordings and facilitator notes were used to clarify the wording of a particular response or to better understand the context in which a response was given. This process yielded one card per response for each group in each of the three research domains: 1) who the system should serve; 2) what outcomes the system should achieve; and 3) what program or system-level changes will be necessary to achieve these outcomes.
Responses to the research questions regarding indicators of outcomes and indicators of necessary changes were transferred onto the data cards with the associated outcomes and changes. While efforts were made to make the distinction between outcomes and indicators clear to focus group participants, analysis of the data showed that the indicators identified by group participants were sometimes difficult to measure and outcomes identified by participants might sometimes have better served as indicators. Adjustments were made by the research team in terms of what category a response might be included in. For example, if responses such as improved school attendance or reduced school suspensions were identified by focus group participants as outcomes, they were re-categorized as indicators. In contrast, responses such as academic success and appropriate school behavior were retained as outcomes relating to school functioning.
Three sets of cards resulted from the initial transfer of the data. The data cards were uniformly marked with the name of the stakeholder group, the focus group facilitator, the number of participants, and the date the focus group session was conducted. The cards were also marked with the appropriate research domain, for example "Who" for the priority population research domain, so that cards could be categorized by the appropriate research domain and could always be tracked by who the participants were as well as how the item was prioritized.
Data within each domain was analyzed in a similar fashion. The data cards were read and hand-sorted into categories. It is most important to note that these categories were emergent from the data rather than identified prior to the sorting process. This means that the categories which emerged during the analysis represent the multiple perspectives of the stakeholders who participated in the focus group process.
It was hoped that the focus group process would yield data about how participants prioritized their responses. Using the priority stickers, the priority measure was intended to demonstrate which responses groups gave the highest priorities. During the process of data analysis, however, it was determined that the priority-setting procedure was not adopted with enough consistency across stakeholder groups to allow for generalization about priorities on this basis. Five focus groups used the sticker process to indicate priorities in all three research domains. Some groups chose not to use the priority stickers at all. For example, in a focus group of participants within mental health, the group voiced the belief that all of the outcomes it had identified were important and it was impossible to give one priority over the other. This group developed a statement which broadly defined the target populations and outcome goals it supported and voted unanimously to have this prioritization represent the group. It is also important to note that among groups that participated in the sticker priority process, the number of sticker votes did not correlate with the number of participants in a focus group.
Even when the group participated in the priority-setting process, some individuals chose not to make a priority selection. Because of the uneven use of the priority sticker process, it was decided that the measure of response frequency would be most useful in determining the responses valued by participants and priority votes were not included in the summary tables. For the purpose of discussion of the findings, however, reference will be made to priority votes taken among the groups that participated in this process as well as reference to the discussion of priorities by groups that did not participate in the process.
Return to The Michigan Outcome Identification Project Home Page
 |
 |
 |
CFS Home | CFS Centers & Projects | CFS Publications | CFS News | CFS Faculty & Staff | CFS Divisions Copyright © 2005, Dept. of Child & Family Studies, Louis de la Parte Florida Mental Health Institute -- see terms of use. |
|
To contact us
about this website, write us at cfswebmaster@fmhi.usf.edu To correspond with employees of the department, write to them care of: The Department of Child and Family Studies Louis de la Parte Florida Mental Health Institute University of South Florida 13301 Bruce B. Downs Blvd. Tampa, FL 33612-3807 |