The question between slides

Listening to a whole class gives you the feeling you understood it, and the only thing that tells understanding apart from having listened is trying to recall it without looking. So when a slide ends, before moving to the next, Mina asks you one thing.

It is not a test and there is no score: Mina cheers if you got it and looks puzzled if you did not, and that is all. You turn it off with Questions, in the viewer settings. If the question is not ready yet, the slide ends the way it would without one: you are never made to wait for it. In silent mode, which has no end of audio, it comes up when you move to the next slide by hand.

Four forms, taking turns

Picking between options is recognising; filling a gap is producing the exact word; answering the open question is saying the whole thing; telling the slide back out loud is giving it again. Just one of the four, over forty slides, would always be the same exercise.

The first three take turns by slide number. Telling it back comes up on one slide in four, and since four is not a multiple of three, it lands evenly across the other three: it never keeps taking the place of the same one. It needs a microphone and uses your plan's transcription, so with no microphone, or with the month's transcription used up, the one that was due comes up instead.

Telling it back in your own words

You tap Start talking, tell the slide the way you would tell it to someone, for up to four minutes, and tap That is all. Mina tells you which ideas you covered: «You explained 3 of 5 ideas», plus two or three sentences with what you left out, and the frame points at the parts of the slide you said nothing about. If you covered everything, «You said all of it». Below, you get what you said, transcribed. No grade and no score: all you need to know is what you missed.

What you say is transcribed in the cloud, and the recording is not kept. The text is compared against the ideas in the script, the same paragraphs you listened to, not against the words on the slide. Explaining well means not repeating: if the idea is about coupling and you say «how much one module depends on another», that idea is explained. Counting matching words would mark it as missed.

Naming is not explaining: saying the term without saying what it means does not count. And when in doubt it does not count in your favour, because being left with one extra idea to review is no big deal, while being told you explained something you did not is.

A question that fails the check is not shown

The question is about what was explained, and the answer has to be on the slide. Both conditions are in the request to the model, but neither is taken for granted: what comes back is checked in code and anything that fails is thrown away. The check makes sure the answer is on the slide, that no two options say the same thing, that the right one is not the longest of all, that the missing word appears only once in the sentence, and that the question makes sense without the slide in front of you («the chart above» does not count). Options are shuffled, so the right one does not always land first. If a multiple choice or fill in the gap question fails, an open one is tried once; if that fails too, the slide ends with no question. A cover page, a photo or a slide with very little text gets none.

Better to ask nothing than to ask just anything: a made up answer you walk away with is worse than not being asked at all.

With the answer in view, answering is just reading

While the question is on screen, the script goes away and the slide blurs. Once you answer, the slide comes back with the frame on the spot where that answer lives, which is what turns an «I did not» into a «so it was right there».

With the keyboard, the up and down arrows move through the question's buttons, the numbers 1 to 4 pick an option, Enter does what the visible button does and Escape skips it. The left and right arrows still change slides. The other shortcuts are in Listening to explanations.

What you missed comes back later

When you say you did not have it, the part of the slide where the answer was gets noted, just like the paragraph that was playing when you asked Mina something without leaving the viewer. That is where Practice starts next time. You do not need to flag anything by hand: missing a question and asking about something already tell it what was shaky. Telling it back in your own words notes nothing: what the model thinks of what you said is not your decision.