/ Single-Case Design Graphs

For teachers

Assign a graph your students build and analyze

Each student gets their own dataset, builds an APA-style graph of it in Excel, decides whether it shows experimental control, and prints a submission report to upload to your course. You get an answer key for every student. No accounts, and nothing is stored on a server.

How it works

  1. Create an assignment below. Pick a graph type (or a mix) and whether students see the correct answers in their report. You get a code and a link.
  2. Share the link in your course. Students enter their name and student ID. Their ID creates a dataset that's theirs alone, so classmates can't copy each other's graphs.
  3. Students build and analyze. They download their data, build the graph by following the lessons (set to their graph type), attach the picture, complete the checklist, answer visual analysis questions, and write an interpretation.
  4. Students submit a PDF of their report, and their Excel workbook if you ask for it (tick the box when you create the assignment, and the student page asks for it). The report ends with a receipt code.
  5. You grade with the answer key and the rubric. Enter a student's ID to see their data, the correct answers, and how their graph should look. Use the receipt code to confirm the answers in the report match their dataset.

Step 1: Create an assignment

After students create their report, should it show the correct answers?
What do students upload?

Step 2: Answer key

Enter the assignment code, your teacher key, and one student ID per line, exactly as students typed them (letter case and extra spaces don't matter).

How private is the key? The teacher key keeps students from looking up their answers on this page. It isn't strong security: everything runs in the browser, so a student who studies this site's code could still work out their answers. For graded work, weight the graph and the written interpretation, which students produce themselves.

Step 3: Check a receipt code

Each report ends with a receipt code such as 102-K7Q2X. It records the student's answers, and its check characters change if the answers, the student ID, or the assignment code are edited. It catches a report that was changed or made for another dataset. It can't stop a determined student from making a new report, so always grade against the answer key.

Check a whole class at once
Separate the two with a comma or a tab. Pasting two columns from a spreadsheet works.

Grading rubric

A suggested rubric for the report. Adjust the points to your course. Each row links to the lessons that teach it.

CriterionFull creditPartial creditTaught inPoints
Data and graph typeOne column per phase or condition; the right graph type; every value plotted correctlyMinor layout or plotting errorsExcel Lesson 1; Design guide2
AxesFloating zero, sensible ranges and ticks, both axis titles, session 0 hiddenOne axis problemExcel Lesson 32
Phase changesLines halfway between sessions, the data path breaks at phase changes, phases labeled above the plot area (or a legend for conditions); dashed criterion lines in a changing criterion designLines or labels misplacedExcel Lessons 4, 5 and 62
APA styleBlack and white, markers that tell apart conditions sharing a phase (the phases of an AB, reversal, multiple baseline, or changing criterion graph share one filled circle), one sans serif font, no gridlines, border, or chart title, 6.5 × 4.5 in (each multiple baseline tier 6.5 in wide and about 2 in tall). Chart size can be checked only in the workbook; if you don't collect it, grade the rest of this row.One or two style problemsExcel Lessons 3 and 72
Visual analysis questionsAll auto-graded questions correct. Students get 2 or 3, depending on their dataset; the answer key lists each student's.3 × the share answered correctly, rounded to the nearest half pointThe six features; Experimental control3
InterpretationDescribes level, trend, and variability, then immediacy, overlap, and consistency, with numbers from the graph; reaches the correct conclusion about experimental control and explains it with the design's logic (for example, untreated tiers staying stable in a multiple baseline, or behavior matching each criterion in a changing criterion design). For alternating treatments and functional analyses, judge separation between conditions instead of change between phases; for a functional analysis, the conclusion is whether it is differentiated and which function it identifies.Correct conclusion with thin support, or good description with the wrong conclusionModel interpretation (reversal); how each design shows control5
Total16

Participation option. Ask students to reach mastery (8 of their last 10 correct) in four sets on the practice page: the mastery check and the three skill sets (the six features, experimental control across designs, and conservative dual-criteria). Then have them print their practice record, which always lists those four sets and says how many are mastered. The record comes from the student's own browser, so use it for participation credit, not as verified evidence.

Privacy. Student names and IDs stay in each student's browser and appear only on the report they print. Nothing is sent to USF.ac or anyone else. If your institution prefers, have students use an anonymous code instead of a student ID.