D772 Statistical Data Literacy, catalog number MATH 1802, is the three-CU general education course covering statistics, research, data and data representation applied to interpreting data and supporting decision-making. Notice what the description emphasises. It is a literacy course, which means the target is reading statistics critically and presenting them honestly, more than computing them from scratch.
You are being trained as a consumer of statistics
Most quantitative courses train producers: here is a formula, apply it to this data set. A literacy course trains the other role, the one everybody actually occupies. You are handed a chart in a meeting, a figure in an article or a dashboard from a vendor, and you have to decide whether to believe it and what to do about it.
That reframes what you should be practising. The productive drill is not computing a standard deviation by hand. It is picking up any published statistic and running four questions across it: who was measured, how were they selected, what exactly was counted, and compared to what. A figure that survives all four is usable. Most do not survive the second.
Selection is where the interesting failures live. A survey of people who volunteered answers a different question from a survey of a random sample. A dashboard filtered to completed cases hides the incomplete ones. An improvement measured only among those who stayed says nothing about those who left. None of that is detectable from the number itself, which is why the questions have to be asked deliberately.
Data representation is the second half of the course
The catalog names data representation explicitly, and it carries real weight. Presenting data is a set of decisions with consequences, and this course scores whether you make them defensibly.
The rules that matter most are unglamorous. Chart type follows the comparison: categories compare with bars, distributions show as histograms, two continuous variables belong on a scatter plot, and change over time gets a line. A vertical axis that does not start at zero exaggerates differences, which is acceptable only when you say so. Colour should carry meaning, not decoration, and never be the only channel carrying it. Every axis needs a label and a unit, and every chart needs a source line.
The deeper rule is that a display makes a claim. A chart of two bars claims those two things are comparable. A trend line claims the pattern continues. A percentage claims a denominator. Before building any display, write the sentence you want a reader to say afterwards, then check that the display actually earns that sentence.
Turning competencies into a work plan
WGU keeps competency detail and assessment information inside your Course of Study, not in the public catalog, so start there. Where a performance assessment exists, each scored aspect is judged on its own three-point scale, a score of 2 in each is what passes the task, and there is no averaging between aspects. Where the instrument is a proctored objective assessment, your preassessment result tells you where the hours belong.
A worked plan with numbers. Four strands: data types and study design, descriptive summaries, representation and display, and interpretation for decisions. Over four weeks at eight hours a week that is thirty-two hours, or eight per strand. Spend two hours per strand on the concepts, four on exercises, and two on a critique: take a real published chart, write two hundred words on what it does well and what it hides. Four critiques is a portfolio of exactly the reasoning the course scores, and the practice transfers directly into any written deliverable.
Where the task is written, budget by aspect. Five scored aspects across a 1,400-word submission gives about 250 words each after an opening and a close. Interpretation aspects deserve 350, taken from aspects that only ask you to describe the data set, which are complete once they are accurate.
A structure for a data-informed recommendation
Where task directions set a format, follow them exactly. Otherwise this order carries a data argument cleanly from source to decision.
| Section | Contents | What a reader checks |
|---|---|---|
| Decision at stake | The choice this analysis is meant to inform | Whether the data could actually change the decision |
| Data provenance | Source, collection period, how cases were selected | Whether the sample matches the population you are talking about |
| Variable inventory | Each variable with its type and what it measures | Whether the methods used suit the types |
| Summary | Centre, spread and shape, with the display that fits | Whether the summary hides a skew or an outlier that matters |
| Display | One chart per claim, labelled, sourced, axis stated | Whether the chart earns the sentence it sits under |
| Reading | What the data supports, in plain language | Whether the language overstates what a sample can show |
| Recommendation | The action, with the uncertainty stated | Whether the recommendation would change under a different reasonable reading |
The provenance row is the one that separates a literacy course from a computation course. A perfectly computed statistic from a badly selected sample is worse than no statistic, because it carries authority it has not earned. Two sentences on where the data came from protect everything downstream.
Presenting numbers without misleading anyone
- Give the sample size next to every proportion. Sixty percent of five is not a finding.
- Report centre and spread together. A mean with no measure of variation invites the reader to imagine a precision the data does not have.
- Prefer the median where a distribution is skewed, and say why you chose it. Income, wait times and length of stay are almost always skewed.
- Label axes with units and state any truncation. A chart whose axis starts at ninety percent can be honest, but only if it says so.
- Avoid stacking too many series into one display. Two comparisons in one chart is usually one too many.
- Cite the data source the way you would cite a paper, with enough detail that someone else could retrieve it.
The habit that most improves this kind of work is writing the counter-reading. After stating what the data shows, write one sentence describing how a reasonable sceptic would read the same numbers. Addressing it turns a report into an argument, and it is the clearest possible evidence of statistical literacy rather than statistical procedure.
What Competent looks like in D772
WGU records outcomes as Competent or Not Competent, with no letter grades and no ordinary grade point average. Performance assessment work can be revised and resubmitted with no penalty for the earlier version, so a return costs days inside a six-month flat-rate term rather than standing.
Work that passes on the first read tends to show:
- Data provenance stated before any conclusion is drawn.
- Variable types identified and methods matched to them.
- Displays that fit the comparison and carry labels, units and a source.
- Language calibrated to what the data can support, keeping association separate from cause.
- A recommendation that names an action and states its uncertainty.
Where your version of the course carries a proctored objective assessment, the rule is absolute. Proctored assessments are yours to sit. We prepare with diagnostics, drills, worked data critiques and timed practice, and we give an honest readiness read. We do not sit or assist during an assessment and we never ask for portal credentials.
Six mistakes that cost time in D772
- Computing before questioning the data. A literacy course scores the questions you asked about the source at least as heavily as the arithmetic.
- Reporting a mean for a skewed distribution. It is not wrong arithmetic, it is a misleading summary, and misleading is the failure mode this course is built to catch.
- Building a chart before deciding the claim. The display exists to support one sentence. Without that sentence, chart choice is arbitrary.
- Using percentages without denominators. The number is uninterpretable and the omission reads as either carelessness or persuasion.
- Turning an association into a recommendation. Data can support an action without establishing a mechanism, and the sentence has to say which one you are relying on.
- Leaving out uncertainty entirely. A recommendation with no stated confidence reads as overclaiming, and overclaiming is the exact opposite of the competency.
How support works on this course
Send your competency list, any preassessment result and the task directions from your Course of Study. What comes back is aspect-mapped: a model report with provenance stated, displays matched to claims, language calibrated to the evidence, and a recommendation that names its uncertainty, plus a walkthrough of the reasoning so you can build the next one yourself.
Terms at WGU run six months at a flat rate, so what lowers effective cost per course is how many you close inside a term. Data literacy is one of the general education courses that pays off well beyond its three CUs, because every later course that asks you to use evidence is asking for the same four questions about the data behind it.
Questions students ask about D772
Is D772 the same as MATH 1802?
Is this a maths course or a research course?
How does D772 relate to D582?
Data report due and the chart does not prove the claim?
Send the rubric and directions. You get a model report with provenance, displays matched to claims, and a recommendation that states its uncertainty.
Where D772 sits in WGU's programs
The July 2026 catalog places this code in 8 current WGU programs. 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.