Learn Investigate whose data you see and what a summary hides.
Start with everyday situations, predict an outcome, and test your ideas. You need basic arithmetic, percentages and graph reading; explanations introduce additional ideas where needed. No coding or account is required. Every investigation has school, college and workplace extensions—you can choose any of them.
Audit records, choose honest summaries, design a fair comparison and explain what sample uncertainty does and does not support.
Open any topic directly. Review functions and graphs whenever useful.
The original two-investigation starter remains available. Choosing the full course adds its published lessons to your selected scope; your earlier work stays saved. Skipping does not count as completion or understanding.
Free exploration: no completion scope selected yet. Project reports and independent evidence are tracked separately.
Choose a unit, then open a lesson on its own page. This outline lists the sequence; it does not combine the full lessons into one page.
Identify selection bias, duplicate events and missing observations.
○ Not marked complete · Included · Practice check available
Can a large survey still give the wrong answer?
An original synthetic campus has 100 commuters and 100 residents. Support for a later opening time is 80% among commuters and 20% among residents. Change invitations and response rates.
○ Not marked complete · Included · Practice check available
Which rows are observations and which are recording errors?
Six original synthetic service records contain one repeated event ID, one missing duration and one impossible negative duration. The trace shows every row.
Choose summaries and chart geometry that preserve the meaning of evidence.
○ Not marked complete · Included · Practice check available
What changes when the slowest service disappears from the log?
Five original synthetic service waits start at 10, 12, 14, 16 and a controllable final value, in minutes. Inspect the mean, median and missingness.
○ Not marked complete · Included · Practice check available
Did the data change, or just the visible baseline?
Two synthetic dashboard scores are shown as bar heights relative to a controllable lower axis bound. The numerical measurements remain visible.
Separate treatment effects, composition differences and sampling uncertainty.
○ Not marked complete · Included · Practice check available
Can easier tasks make a weak treatment look better?
An original synthetic task study has easy-task success 80% and hard-task success 40%. Treatment changes both rates by the same chosen number of percentage points.
○ Not marked complete · Included · Practice check available
How much more data is needed to halve random uncertainty?
Inspect a normal-approximation interval for an invented independent binary sample. The percentage is an estimated sample proportion; this is not a medical or policy recommendation.
A fictional service claims it became faster because its dashboard mean fell. Build an evidence brief using a cleaned record audit, a distribution summary and a comparison design. Preserve the raw data and distinguish a changed record from a changed service.
Save a report, compare captured evidence, and attempt three new cases. A completed report records submitted work; the rubric supports a separate human review of its quality.
Loading saved experiments…
Progress is saved in this browser when storage is available. Review your experiments and export a backup in My Work.