Content analysis has been used in several studies of social work doctoral dissertations (see Jenkins, Cherry, Nishimoto, Alvelo, & Okert, 1982; Harrison & Thyer, 1988; Brun, 1997; Dellgran & Hojer, 2001; Shek, Lee & Tam, 2007; Horton & Hawkins, 2010). One prominent theme in these studies was the paucity of social work dissertations that address outcomes of social work interventions (Harrison & Thyer, 1988; Horton & Hawkins, 2010). This, and questions about other types, topics, and study populations found in social work dissertations are addressed in this study.
Methods:
This study involves analysis of social work dissertations written in the United States between 1998 and 2000, and between 2008 and 2010. Since the late 1990s, electronic copies of doctoral dissertations are easily accessed on-line from the data base of the University of Michigan (UMI). All dissertations that were available in full text form were included in this study. Five MSW-level graduate assistants downloaded documents and entered data and six doctoral students were involved in data coding and analyses. Because the codes were based on manifest content (i.e., each author's characterization of the type of study, topic, and population), inter-rater reliability was high (93-96%).
Results:
A total of 1451 dissertation abstracts were reviewed, 761 from 1998-2000 and 690 from 2008 -2010. Of the 1451 dissertations, 56.7% were quantitative exploratory or descriptive studies, 19.7% were qualitative exploratory or descriptive studies, and 7.8% were mixed methods exploratory or descriptive studies. Other research types included instrument design and evaluation (1.4%) , case studies (1.1%) and historical research (.8%). Over both time periods, only 8.6% of dissertations were experimental/explanatory or evaluation designs. In the dissertations completed in 1998-2000, 10.7% were experimental/explanatory/evaluation designs; this proportion dropped to 6.0% in the 2008-2010 data. Extensive use of experimental/evaluation designs was clearly linked to particular schools. Other patterns were similar across both time periods. The majority of the dissertations (71.7% in the first three years and 76.3% a decade later) were about social phenomena and problems; other studies looked at social work practice (14.2%; 11.3%), policy (3.3% for both time periods), and social work education (2.9%, 2.6%). The most common phenomena and problems addressed across both time periods were mental health/mental illness (9%), child welfare (5.4%), substance abuse (3.2%), and health and health care (2.7%). Other common issues were adoption, disabilities, domestic violence, HIV/AIDS, violence, and welfare reform. Social workers were the identified population in 6.5% of the dissertations. Other populations commonly studied were adolescents and children, African Americans, caregivers, immigrants, and single mothers.
Conclusions and Implications:
Social work dissertations completed in the United States between 1998 and 2000 and between 2008 and 2010 addressed many topics of concern to social work practitioners and educators. Unfortunately, few examined outcomes of social work interventions or employed the kinds of methods that would add to the knowledge needed for evidence-based practice. Doctoral program directors and dissertation chairs need to encourage research on practice.
