4. Make major design decisions so that the design answers important evaluation questions for intended users. Consider design options and choose those most appropriate for the evaluation’s purposes.

  • Pure or mixed methods design: Determine whether the evaluation will be purely qualitative or a mixed method design with both qualitative and quantitative data.
  • Units of analysis: No matter what you are studying, always collect data on the lowest level unit of analysis possible; you can aggregate cases later for larger units of analysis. Below are some examples of units of analysis for case studies and comparisons.
    • People-focused: individuals; small, informal groups (e.g., friends, gangs); families
    • Structure-focused: projects, programs, organizations, units in organizations
    • Perspective/worldview-based: People who share a culture; people who share a common experience or perspective (e.g., dropouts, graduates, leaders, parents, Internet listserv participants, survivors, etc.)
    • Geography-focused: neighborhoods, villages, cities, farms, states, regions, countries, markets
    • Activity-focused: critical incidents, time periods, celebrations, crises, quality assurance violations, events
      Time-based: Particular days, weeks, or months; vacations; Christmas season; rainy season; Ramadan; dry season; full moons; school term; political term of office; election period
      (Note: These are not mutually exclusive categories)
  • Purposeful sampling strategies: Select information-rich cases for in-depth study. Strategically and purposefully select specific types and numbers of cases appropriate to the evaluation’s purposes and resources. Options include:
    • Extreme or deviant case (outlier) sampling: Learn from unusual or outlier program participants of interest, e.g., outstanding successes/notable failures; top of the class/dropouts; exotic events; crises.
    • Intensity sampling: Information-rich cases manifest the phenomenon intensely, but not extremely, e.g., good students/poor students; above average/below average.
    • Maximum variation sampling: Purposefully pick a wide range of cases to get variation on dimensions of interest. Document uniquenesses or variations that have emerged in adapting to different conditions; identify important common patterns that cut across variations (cut through the noise of variation).
    • Homogeneous sampling: Focus; reduce variation; simplify analysis; facilitate group interviewing.
    • Typical case sampling: Illustrate or highlight what is typical, normal, average.
    • Critical case sampling: Permits logical generalization and maximum application of information to other cases because if it's true of this one case, it's likely to be true of all other cases.
    • Snowball or chain: Identify cases of interest from sampling people who know people who know people who know what cases are information-rich, i.e., good examples for study, good interview subjects.
    • Criterion sampling: Pick all cases that meet some criterion, e.g., all children abused in a treatment facility; quality assurance.
    • Theory-based or operational construct sampling: Find manifestations of a theoretical construct of interest so as to elaborate and examine the construct and its variations, used in relation to program theory or logic model.
    • Stratified purposeful sampling: Illustrate characteristics of particular subgroups of interest; facilitate comparisons.
    • Opportunistic or emergent sampling: Follow new leads during fieldwork; taking advantage of the unexpected; flexibility.
    • Random purposeful sampling (still small sample size): Add credibility when potential purposeful sample is larger than one can handle; reduces bias within a purposeful category (not for generalizations or representativeness).
      • Sampling politically important cases: Attract attention to the evaluation (or avoid attracting undesired attention by purposefully eliminating politically sensitive cases from the sample).
    • Combination or mixed purposeful sampling: Triangulation; flexibility; meet multiple interests and needs.
  • Determine sample size: No formula exists to determine sample size. There are trade-offs between depth and breadth, between doing fewer cases in greater depth, or more cases in less depth, given limitations of time and money. Whatever the strategy, a rationale will be needed. Options include:
    • Sample to the point of redundancy (not learning anything new).
    • Emergent sampling design; start out and add to the sample as fieldwork progresses.
    • Determine the sample size and scope in advance.
  • Data collection methods: Determine the mix of observational fieldwork, interviewing, and document analysis to be done in the evaluation. This is not done rigidly, but rather as a way to estimate allocation of time and effort and to anticipate what data will be available to answer key questions.
  • Resources available: Determine the resources available to support the inquiry, including:
    • financial resources
    • time
    • people resources
    • access, connections
QEC Home
4