Oklahoma Marriage Initiative (OMI): A Process Evaluation. Approach to Qualitative Data Collection


The 163 respondents interviewed were workshop leaders and participants in a wide range of sectors (described below).  They included 18 OMI management staff, 8 members of the OMIs research advisory group and curriculum developers, 24 state or local managers at the public and private agencies that deliver OMI services, 42 workshop leaders, and 71 workshop participants.  Respondents were from high schools, welfare agencies, adoptive/foster parent services, state prisons, juvenile offender programs, Cooperative Extension Services, and Head Start programs, as well as members of the faith community, mental health counselors, and others delivering services in the general community.  Most of these respondents were located in either the Oklahoma City or Tulsa metro areas, although some were in rural areas (such as correctional centers). 

Timing.  Qualitative data were collected in the field on three occasions.  Early in the project (late 2005), a first site visit was conducted to explore the broad implementation approach and refine the evaluation design.  A second visit, in April 2006, gathered information from Oklahoma Department of Human Services (DHS) staff, who fund and oversee the initiative, and from OMI staff at Public Strategies, Inc. (PSI), the firm responsible for day-to-day operations.  In October, November, and December 2007, the study team conducted three week-long visits to numerous institutions, communities, and individuals implementing OMI services to ascertain their perspectives on how services are provided and received.  During these visits, we convened meetings with management staff at institutions, held group interviews with workshop leaders in various sectors who facilitated OMI services, and conducted focus groups with people who had participated in workshops in several different OMI sectors. 

Each semi-structured interview and focus group was conducted in person by a team of two experienced researchers, except for interviews with research advisory group members, which were conducted by telephone.  Each interview and focus group was guided by a detailed protocol and lasted approximately 60-90 minutes.  Responses to questions were documented with detailed notes and audio recordings. 

Triangulation.  By exploring specific topics from the perspectives of multiple sources, the study team was able to develop a detailed and nuanced understanding of the OMI.  For instance, similar questions were asked of respondents who were representatives of organizational management, workshop leaders, and workshop participants, to gain their different perspectives.  This strategy was followed in each of the selected sectors. 

Selection of sectors.  The OMI is or has been active in a wide array of sectors, representing diverse institutions and populations (see Table I.1).  The study classified these sectors under two overarching heads: institutional and community.  The five institutional sectors are education, corrections, health, social services, and the military.  The five sectors classified as community-based are faith, counseling, mental health services, community services and other, and the Hispanic and Native American communities.

We focused on sectors that are especially active today, but also included several agencies or sectors that have been active in the past but are less so now.  Such sectors were included so the study could identify lessons from OMI experiences in efforts that were not sustained or did not grow as expected, in order to inform new and developing state and community initiatives.

Sampling strategy.  Our plan for identifying, selecting, and contacting respondents for the interviews and focus groups relied on a combination of purposeful identification and representative sampling.  Purposeful identification was most appropriate when information and insights could come only from certain individuals with particular roles or knowledge, such as DHS officials, senior OMI staff, curriculum developers, members of the research advisory group, and lead staff at public and private agencies that deliver OMI services.  In contrast, a representative approach was more appropriate in collecting information that would be used to characterize the experiences or attitudes of large numbers of workshop leaders and participants.

To select workshop leaders, we first identified trained leaders in each sector and agency from the OMI management information system.  In this process, we stratified leaders by sector and activity level, in order to include both leaders who had conducted many workshops and those who had conducted fewer or none.  This method was intended to help us address questions about why some workshop leaders are trained but do not conduct workshops.  In selecting leaders for the interviews, it was necessary to consider geographic location as well as our desire to triangulate interviews from different levels within a given agency or sector.  Selected leaders were mailed an invitation to attend an interview or focus group, and asked to dial into MPRs toll-free number to register.  After obtaining information about availability from interested respondents, we contacted them again to advise them of the date, place, and time, and provided multiple reminders.  Interviews and focus groups were usually held at a neutral location, such as a local library or community center, unless that was infeasible.  

Table I.1.
Major OMI Service Delivery Sectors
Sector Agency Sponsor Target Population Workshop Leaders Curriculum Adaptation(s)
Education OK Dept. of Career Technology High school students Family and Consumer Sciences teachers Connections-PREP®
OSU Cooperative Extension Services Adult students in community education OSU educators PREP®
Corrections OK Dept. of Corrections (DOC) Prison inmates and their partners/spouses Prison chaplains and trained inmates PREP®, Within My Reach
OK Association of Youth Services (OAYS) Adolescent first offenders and their parents Trained, sometimes licensed facilitators and OAYS contractors PREP®
Health OK Dept. of Health, Child Guidance Parents OK Dept. of Health professional psychologists and clinicians PREP®
Social Services OK Dept. of Human Services (DHS) TANF recipients Career Development Specialists Within My Reach, PREP®
OK Dept. of Human Services Adoptive/foster parents/grandparents, parents of special-needs children OK Dept. of Health professional psychologists and clinicians ENRICH, PREP®
Community Action Agencies or Head Start Agencies Head Start parents Head Start staff PREP®
Military Army, Air Force, and National Guard bases Members of the military and their partners/spouses; base and post employees Family advocacy and family support staff; chaplains;  employee assistance counselors PREP® or Christian PREP®
Faith N/A Church and synagogue congregation members and general public Pastors, other clergy, and lay people PREP®; Christian PREP®; Jewish PREP®
Counseling; Mental Health Services N/A Private clients and general public Various licensed professionals, such as counselors, social workers, and psychologists. PREP®; Christian PREP®
Community Services and Other N/A Employees and general public Various volunteers PREP®
Hispanics N/A Hispanic/Latino adults Volunteers who focus on Hispanic families PREP® for Latinos; Christian PREP® for Latinos
Native Americans N/A Native American adults Volunteers who focus on Native Americans PREP®; Christian PREP®
Note:  Institutional sector means services are typically provided by agency staff as part of their regular jobs.  Community/volunteer sector means services are usually provided by individual volunteers on their own time.

Obtaining a strictly random sample of workshop leaders and participants was not always possible.  Some self-selection inevitably occurred, as some invited leaders and participants responded and others did not.  At state prisons and juvenile offender programs, we selected sites to visit and interviewed workshop leaders at those sites.  At two prisons, the chaplain who led the workshops brought to the participant focus groups some prisoners who had not been invited.  Due to the wide geographic distribution of teachers in high schools, we capitalized on a statewide conference during a study site visit, and invited teachers attending the conference to participate in our discussion group.

View full report


"report.pdf" (pdf, 726.12Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®