D019 Data Literacy and Evidence-Based Practices is recorded as EDUC 5291 in the catalog and is worth 3 competency units. It asks you to identify an educational problem, recognise which types of data speak to it, generate and analyse those data, and draw conclusions that improve learning across a K-12 setting. D019 and EDUC 5291 are the same course. The thing that decides how the course goes for you is not statistics. It is whether you can state a problem narrowly enough that data could ever answer it.
Why the problem statement is the whole course
Students lose more time in this course to a badly framed problem than to any analytical difficulty. Reading achievement is low is not a problem statement; it is a mood. Sixth grade students entering below benchmark in reading fluency have not closed the gap by spring, while seventh grade students in the same intervention have, is a problem statement, because you can already see which data would confirm or kill it.
A leader-level data course also expects you to work across data types rather than only test scores. Achievement data tells you what students can do. Demographic data tells you who they are and where the pattern concentrates. Perception data from surveys and interviews tells you what people believe is happening. Process data about scheduling, attendance, staffing and program participation tells you what the school actually did. Conclusions built from one type alone are the standard weakness, because achievement without process cannot explain anything and perception without achievement cannot prove anything.
The evidence-based practices half changes what counts as a good recommendation. It is not enough to propose something sensible. The proposal has to be traceable to research that studied a comparable population, and you have to be able to say honestly how comparable it was.
Every task closes as Competent or Not Competent, with no letter grade and no ordinary grade point average involved. The 3 competency units describe the share of a six month flat rate term the course represents.
Reading the aspects as an analysis pipeline
The scoring detail for D019 lives in your Course of Study rather than the public catalog, so start by counting your rubric's aspects. Each is scored on its own and every one has to reach a 2 for the task to pass. Data work fails this test in a specific way: students spend their length on the analysis they enjoyed and leave the limitations aspect in a single sentence.
The budget, worked. Take a rubric with seven scored aspects and directions pointing at roughly 2,800 words. Reserve 200 for an opening that names the setting and the problem and 160 for a close, which leaves 2,440 of scored body. Flat, that is 348 words per aspect. Weighted, give 520 to the analysis aspect where the actual work is shown, 430 to the conclusions aspect where the reasoning either holds or does not, and 298 to each of the remaining five. That sums to 2,440.
Build the tables and figures before you write the prose. In data writing the exhibits are the spine, and prose written first almost always describes exhibits that turn out not to say what you assumed.
A shape for a school data investigation
Follow the directions where they prescribe a format. Where they do not, this sequence keeps each scored move visible instead of buried inside a narrative.
| Stage | What it establishes | The test it has to pass |
|---|---|---|
| Problem of practice | The specific gap, in one sentence, bounded by grade, subject and time | Could data in principle prove this wrong |
| Data inventory | What exists already, what has to be generated, and which type each source is | At least two data types represented |
| Collection method | How anything new was gathered, from whom, over what window | Someone else could repeat it |
| Analysis | The comparisons made, shown in tables or charts with the counts underneath | Every figure has a denominator |
| Findings | What the data support, stated separately from what you believe | No finding without a visible exhibit behind it |
| Limitations | Sample size, response rate, missing data, confounds you cannot rule out | Named specifically, not waved at |
| Evidence-based response | The practice you would adopt, with research support and a monitoring measure | Traceable to a study, not to a hunch |
Denominators deserve their own habit. Twelve students failed is not information. Twelve of nineteen is a crisis and twelve of four hundred is noise, and evaluators reading an analysis aspect notice the difference immediately.
Choosing comparisons that survive a second reader
Almost every conclusion in school data work rests on a comparison, and the comparison you choose decides the answer before any arithmetic happens. Four choices come up constantly and each one has a trap in it.
The first is cohort against cross-section. Comparing this year's fourth graders to last year's fourth graders compares two different sets of children, and a difference can be nothing more than the difference between two groups. Following the same children from third grade into fourth answers a question about growth. Both are legitimate; writing one and claiming the other is not.
The second is aggregate against disaggregate. A school average moves slowly and hides everything interesting. The pattern that a leader can act on almost always appears when the same measure is broken out by subgroup, by teacher section, by attendance band or by how long a student has been enrolled. Mobility in particular explains an enormous amount of apparent underperformance, and a building that never separates continuously enrolled students from newcomers is measuring its own enrolment churn.
The third is proficiency counts against growth. The share of students meeting a benchmark is the number a district reports; how far each student moved is the number that tells you whether teaching worked. A cohort can improve substantially and move nobody across a cut score, and a leader who only tracks the cut score will conclude that an intervention failed when it did not.
The fourth is baseline. If there is no measure of where things stood before the change, no comparison afterwards means anything. When your investigation covers something already underway, say so and reconstruct the closest baseline you honestly can, then note in the limitations what that reconstruction cost you.
Sourcing, provenance and student privacy
Data in a school investigation carries obligations that a literature paper does not, and the aspects covering ethics or appropriate use are easy marks that students throw away.
- Aggregate before you publish. Report by grade, cohort or subgroup at a size where no individual can be reconstructed, and drop cells that get too small to protect.
- State provenance for every figure. Which system produced it, for what date range, and who pulled it. A number without an origin cannot be evaluated.
- Distinguish measurement from truth. A benchmark assessment measures performance on that assessment. Whether that equals reading ability is an argument you have to make, not assume.
- Choose research that matches your setting. A study of a suburban high school does not automatically govern a rural elementary intervention, and saying so strengthens rather than weakens your case.
- Keep perception data honest. Report the response rate alongside the result every single time, because a survey with few responses supports a much smaller claim than students usually make from it.
- Cite in the style the directions require, and cite at the point of claim rather than collecting citations at the end of a paragraph.
If your investigation would involve gathering anything new from students or staff, check with your program and your school about permissions before you collect. Coursework analysis and formal research have different requirements, and the time to learn which one you are doing is before the data exists.
What separates Competent from a return
Because aspects score independently, submissions in this course usually return for one of four reasons: a conclusion that outruns the data, a missing limitations discussion, an analysis with no exhibit, or a recommendation with no research behind it.
- Every claim in the findings section can be traced to a specific table, chart or quoted response.
- Correlation language stays correlational. If you did not isolate a cause, do not write one.
- The recommendation names a measure that would show it working, and a date by which you would look.
- The limitations section names things that could genuinely have changed the answer.
- Nothing identifiable about a student, family or employee appears anywhere in the document.
Performance assessment work at WGU can be revised and resubmitted without a grade penalty, so returns cost calendar rather than credit. In a flat rate six month term that calendar is the constraint that decides how many courses close, which is why the limitations paragraph is worth writing properly the first time. Where this course also carries a proctored objective assessment in your plan, our position is fixed: preparation only, never sitting or assisting during an assessment, and never any request for portal credentials.
Data set in front of you and no problem statement?
Send the D019 rubric and what you can access. You get a bounded problem of practice, an exhibit plan and the limitations list drafted before you analyse.
Six mistakes that stall D019
- Framing a problem you cannot test. If no realistic data would ever contradict your statement, the whole investigation has nowhere to go.
- Using one data type. Achievement numbers alone cannot say why, and a school data investigation is scored on explanation as well as description.
- Charting everything. Six figures that each make the same point crowd out the one comparison that mattered. Choose the exhibit that answers the question and cut the rest.
- Writing causal verbs after correlational work. Caused, produced and drove are the words that get findings sections returned.
- Treating limitations as an apology. Named honestly, they read as expertise. Skipped, they read as someone who did not notice.
- Recommending a practice with no research trail. The course is called evidence-based practices, and the evidence has to be visible for the aspect to be met.
Three questions students send about D019
Is D019 the same course as EDUC 5291?
How is D019 different from D179 Data-Informed Practices?
Do I need statistics software for this course?
Where D019 sits in WGU's programs
The July 2026 catalog places this code in 1 current WGU program. Open a program page for the complete standard path and term positions. The live Degree Plan remains authoritative after transfer credit, substitutions, and mentor planning.
The assessments, one by one
The public catalog does not publish this course's PA/OA identity or task count. WGU Tutors publishes at most one PA manual per course and only from a WGU-controlled public rubric. Until that source exists, PA help begins from the student's real Course of Study and OA support remains preparation only.