Once the data has been collected, analyze it. Data analysis is dependent upon what type of data you gathered. At times, you may choose to review more than one type of data to assess your program or service. For example, you may choose to use a survey and a focus group.

As data is analyzed, keep in mind its limitations. The true impact of a program or service may not surface at the time of your collection. Some impacts take months, semesters, even years, to evolve. It is highly unlikely that an assessment will prove that a single program caused the desired outcomes. It may be that the program, in conjunction with the many other experiences students encounter on a daily basis, encouraged the skills associated with a defined element of student success.

Finally, understand that an assessment project will likely not answer all of your questions. Assessment can be viewed as a photograph, a snapshot in time. Just as your hairstyles have likely changed over the years and you can compare those photos, you can compare assessment snapshots.

Quantitative Data

A few examples of quantitative data include: responses to a survey that asks students to rate their level of agreement with a statement (1=strongly disagree to 5=strongly agree) or a pile of rubrics that rate students on their ability to explain the importance of physical activity.

When analyzing QUANTITATIVE data the general process is:

  1. Clean the Data: Identify and handle missing values, outliers, duplicates, and any errors in the data.
  2. Give the data a “onceover,” noting initial impressions
  3. Four analytic strategies
    • Description (frequencies, percentages, mean, median, mode, range, standard deviation)
    • Differences (participants versus non-participants; do certain participants do better than others?)
    • Change (pre/post)
    • Expectations (do students meet our expectations)
  1. Alone, neither measure of central tendency (e.g. mean, mode, median) nor measures of variability (e.g. range, standard deviation) tell the whole story. Visualize the Data: Create charts (e.g., histograms, box plots, scatter plots) to visually explore patterns and relationships.
  2. Conduct other useful calculations (e.g. sums, percentages)
  3. Take a step back
    • What do the data tell you about your assessment questions? -- What?
    • What are its implications for policy and/or practice? -- So what?
    • What, if anything, will you change about the program or process? -- Now what?
  1. Other considerations
    • Use online survey design software, Microsoft Excel, or a statistical package like SPSS, STATA, or SAS to make calculations
    • For help with statistical analysis (e.g. statistical significance, confidence intervals, etc.) contact the University Data & Analytics office.

Qualitative Data

Qualitative data may include responses to a survey that asks students to define leadership in their own words, notes, and recording from a focus group, or interviews with students about what they learned through an experience. Analyzing data qualitatively is more time consuming than quantitative methods. Most qualitative data is not analyzed in a single session because of the nature of qualitative data analysis. Those assessing qualitative data must set aside time to review the data, observe elements or themes that seem to “stand out” among other pieces of data, and consider their thoughts about the data. Data is typically reviewed several times to ensure it is fully understood. Good analysis is dependent upon knowing the data. For assessment, coding qualitative data will be the most likely method of qualitative analysis. For this type of analysis, it is helpful to follow some steps with the understanding that going back and forth between steps is normal in the qualitative process.

When analyzing QUALITATIVE data the general process is:

  1. Organize the data
  2. Give the data a “onceover,” noting initial impressions
  3. Re-read the data and categorize
    • You can determine the categories ahead of time, allow the categories to emerge from the data, or both
    • You may end up with subcategories
    • Categorizing data is an iterative process
  1. Determine the relative significant of each category by counting the number of times it occurs
  2. Note the responses that do not fit into the categories
  3. Find compelling quotes to include in your assessment report
  4. Take a step back
    • What do the data tell you about your assessment question?
    • What are the limitations? What are the implications?
    • Does the data lead you to make changes or confirm your approach (or both)?
    • What, if anything, will you change about the assessment process?

*Full text adapted from workbook developed by University of Oklahoma.

* Sources used to compile this information.