Case C, literature-search tutorial

• The rule taught is stated plainly and correctly: be specific, use no punctuation or tags for a simple topic search. A learner with no background can follow it.
• The worked example is concrete. A real two-concept query, typed and submitted, which is the right way to teach a procedure.
• The gap is scope. One search pattern in 46 seconds. A learner who needs to combine three concepts, or filter by date, or find nothing and try again, has been shown the first move and left there.

Traced to: slides 2–4

• Correctly served, arguably. A 46-second refresher is a reasonable format for someone who needs one reminder, and the graded item is the useful part.
• The distractors will land for this learner specifically: anyone who has used a database that rewards phrase quoting or truncation carries that habit across, and this item targets exactly that transfer error.
• No indication of when the interface last changed. The course is built from 151 screen captures of a live third-party system. A practitioner cannot tell whether the screens match what they will see, and screen-capture tutorials of external systems age badly and silently.

Traced to: the graded item, the image set

• A fixed 1024×768 canvas with responsive: false. The project scales rather than reflows, so on a 375-pixel viewport the entire course shrinks to 37% of its designed size, text included.
• The content is 151 screen captures of a search-results interface. Those are the elements that survive scaling worst, dense small text inside an image, which cannot reflow at all.
• The audience makes this worse than it looks. This teaches a look-it-up skill to people who look things up in the moment, frequently on the device in their hand. The task is mobile; the course excludes mobile.

Traced to: project metadata, image set

• The instructional text is short, imperative and concrete, "type this, then click search", which is close to ideal for a non-native reader.
• The snag is a keyboard instruction embedded in feedback: the correct-answer response ends "Click anywhere or press 'y' to continue." The letter is arbitrary, appears once, and is not signposted anywhere else. A reader parsing more slowly is the most likely to miss it and stall on a screen that looks finished.
• Four narration audio files with no caption track, so a reader who would use captions to support comprehension has nothing to read along with.

Traced to: slide 5 feedback, media folder

• The player exposes a full control bar including play, pause, back, forward and a captions toggle, which is a reasonable base.
• The captions toggle has nothing to show. The package contains four narration files and no caption or transcript content anywhere. The caption machinery in the player engine is present, the caption data is not. A learner who turns captions on gets an empty overlay.
• Open question: whether the simulated type-and-click interaction on slides 2–4 can be completed without a pointing device cannot be determined from static inspection. It is the one finding in this report that needs an instrumented run to settle, and we are not going to guess at it.

Traced to: playbar configuration, runtime data, media folder

This learner succeeded, which fails the course.

• Success is recorded for viewing 75% of six slides, five slides. The graded question sits on slide five.
• The question must be answered to advance, but not answered correctly: three options, unlimited retries, "Try again." on failure. Exhaustive clicking resolves it.
• The result never reaches the record regardless. The success criterion is view, not score, so the 80% pass mark has no path to the LMS.

Traced to: contentReporting block, quiz block, slide list