Week 8 - Mixed Methods

Reasons for Using Mixed Methods

  • In the late 1980s, publications emerged defining mixed methods research across disciplines.
  • Greene, Caracelli & Graham (1989) identified five reasons for conducting mixed methods studies:
    • To triangulate the data:
      • Collecting both quantitative and qualitative data simultaneously.
      • Comparing and contrasting the findings.
      • Merging the data.
    • To complement the data:
      • Seeking elaboration, enhancement, illustration, and clarification.
      • Mixing results of one method with another.
    • To develop the data:
      • Using results of the first method to develop or inform research using the other method.
    • To understand paradoxes or contradictions:
      • Following up on findings that need further explanation.
    • To expand the data:
      • Extending the breadth and range of a study.
      • Using different methods for different components of the study.

Types of Mixed Methods Designs

  • Main types of mixed methods designs:
    • Convergent parallel design
    • Explanatory sequential design
    • Exploratory sequential design
    • Embedded design

Convergent Parallel Design

  • Used to obtain different but complementary data on a topic.
  • Purpose is to combine the strengths of each type of data.
  • Both quantitative and qualitative data are collected simultaneously.
  • Example: A local community health center conducts a healthcare needs study.
    • Combines questionnaires, in-depth interviews, and focus groups.

Explanatory Sequential Design

  • A two-phased study.
  • Phase 1: Collection of quantitative data to address study questions.
  • Phase 2: Qualitative data is obtained to explain or build on the initial quantitative results.
  • Example: Study on the retention of older healthcare workers (Hodgkin et al., 2017).
    • Research Question: ‘What are the organizational and social factors that impact on the retirement intentions of health care workers who are aged 55 years and over?’
    • Phase 1: Participants (n=299) completed a survey.
      • Measures: demographic variables, retirement intentions, an effort reward imbalance measure, and a general health measure.
    • Phase 2: A smaller sample (n=17) participated in in-depth interviews.
      • Explored retention and retirement intentions to help explain some of the quantitative findings.

Exploratory Sequential Design

  • Begins with and prioritizes qualitative data.
  • Results of the qualitative stage are used to develop the quantitative stage.
  • Based on the premise that exploration of an issue or concept is required first.
  • Useful in developing theories or concepts when measures or instruments are not available.
  • Example: Dellemain, Hodgkin & Warburton (2017) used exploratory sequential design to develop a practice theory for rural case management.
    • The authors argued that although a theory on case management has been developed, there has been little research into the impact of rurality on this type of community work, particularly in the Australian context.
    • Design selected was a qualitative dominant, sequential, exploratory mixed-method design, the aim of which was to develop community-based rural case management practice theory.

Embedded Design

  • In health research, one dataset alone cannot fully explain patterns in the data.
  • Mixed-method approach offsets limitations by adding different types of data for a supportive secondary role.
  • Example: Study of the workflow and work patterns of Australian residential aged care facilities (Hodgkin, Warburton & Savy, 2012).
    • Research was concerned with accurately reporting and documenting the activities undertaken by the healthcare workforce (e.g. division 1 nurse, division 2 nurse, allied health practitioner and ward clerk).
    • Used a structured observation technique over a two-hour period to document each role and the time is taken to do each activity (quantitative data).
    • Supplemented by qualitative data in the form of structured interviews with key personnel.
    • Provided crucial contextual data to help explain contextual factors, such as staff shortages, qualifications held by staff, and facility layout affecting tasks.