The “Data Detective” Approach: Spotting Preparation Gaps with Built-In Statistics

Educator preparation programs do not need to perform complex statistical protocols to uncover meaningful insights about candidate performance. In fact, some of the most powerful program decisions come from simple, consistent analysis of the data programs already collect.

With the EPiC™ Support Dashboard, faculty and program leaders can take on the role of “data detectives”—using straightforward built-in descriptive statistics to identify patterns, pinpoint preparation gaps, and highlight areas of program strength. The goal is not to become statisticians. The goal is to make data usable, visible, and actionable.

When programs approach their data this way, they move beyond surface-level reporting and begin to uncover the narrative behind candidate performance.

Start with the Big Picture: Detecting Outliers in Program Data

One of the simplest ways to begin is by examining data for outliers, which can distort mean scores and standard deviations across key assessment areas.

When faculty review scores across EPiC rubrics or observation markers in the EPiC Support Dashboard, the presence of any outlier representing extreme values is automatically shaded either positive or negative. Though outliers may skew data, it is still prudent for the EPP to investigate why. Was the outlier due to an oversight in scoring or a candidate's actual performance?

Candidates with exceptionally high scores can be used to identify "exemplars", whereas candidates with surprisingly low scores might highlight a need for targeted mentorship or support.

Looking Deeper: Exploring Relationships Through Correlations

Once outliers are identified and addressed, programs can take the next step by examining how different aspects of candidate performance relate to one another.

Even simple analyses across data sets in the EPiC Support Dashboard can reveal important correlations or associations.

For example:

• Do candidates who score highly on lesson planning also demonstrate strong instructional delivery during observations?

• Is there a correlation between candidates’ use of questioning strategies and their ability to engage students in higher-level thinking?

• Do candidates in certain field placements consistently perform differently from others?

These connections help programs move beyond isolated data points and begin to understand how different components of preparation work together.

Instead of asking, “How did candidates perform?” programs begin asking, “Why are we seeing these results?”

Comparing What Matters: Rubrics, Cohorts, and Field Sites

Another powerful strategy is comparison.

By examining performance across:

• EPiC rubrics

• Candidate observations

• Lesson plan preparation

programs can identify where variation exists and what may be contributing to it.

For instance, one cohort may demonstrate stronger performance in differentiation due to changes in coursework or increased emphasis in methods classes. Similarly, candidates placed in certain field sites may show higher levels of student engagement, suggesting strong mentoring or instructional modeling in those environments.

These comparisons, viewed in the EPiC Support Dashboard, help programs isolate variables and better understand which aspects of their coursework are producing the strongest outcomes.

Turning Simple Statistics into Compelling Evidence

Accreditation expectations continue to emphasize the use of data to drive program improvement. Organizations such as CAEP, NASDTEC, and AAQEP expect programs to show not only what data is collected, but how it improves outcomes.

EPiC’s built-in statistical analysis supports this work.

By identifying outliers, making comparisons, and detecting strong correlations, programs can uncover gaps and create clear, credible evidence of continuous improvement through targeted program changes. Data shifts from a reporting requirement to a powerful tool for telling the program’s story.

In the end, the most effective programs are not the ones with the most data, but the ones that know how to use it.