Stage 5 – The exercise
The five tasks were written to be worked through in a class, and they hold up alone with one adjustment: nothing is timed, so take the reproduction in Task 3 as slowly as it needs. The browser lab contains the whole exercise on one page. The reproduction in Task 3 has two hints and a fully worked solution behind collapsible callouts, and the three gaps in Task 5 share one hint, which gives the model syntax for the baseline specification, the one that the preregistration block holds before you change any of its choices.
| Work through | What you do | Timing |
|---|---|---|
| Exercise overview | Re-read the task-by-task section now that you have worked through the modules, and use it as a guide to the rows below. | 20 minutes |
| Your workspace, or the Browser lab | Tasks 1 to 3: run the setup check and confirm that the panel has 270 rows; meet the panel, write the estimand and finish the causal graph; then fit the constrained model and reproduce t = −3.804, the value recorded by Multi100 for analyst C6HJR. | 90 minutes |
| Specification menu | Read the axes, the three levels they sort into, and the claim-alignment convention before you touch Task 4, so that ‘supports the claim’ means the same thing on every axis. | 30 minutes |
| Estimand choices | Set the seven Level 1 choices, read the question that they define, and watch how far the measurement and estimation variants of that one question spread, before you choose yours. | 20 minutes |
| Task 4, in the workspace or on the Browser lab | Try three or four specifications with try_spec(), print them side by side with my_experiments(), then decide which one you would defend, run it once more so it is the most recent call, and submit it with submit_exploration(), which sends the last specification you tried, whatever else is in the log. |
45 minutes |
| Task 5, in the workspace or on the Browser lab | Complete the preregistration block and run prereg() before writing any model code. Then translate the specification in the block into model syntax yourself, run it, and submit with report_result(). |
45 minutes |
| Cheatsheet | Keep it open while you work. If the browser lab stalls, the first two of the four steps listed there are the ones you can take alone. Steps 3 and 4 and the by-hand grid lookup assume a facilitator you do not have. | as needed |
The class chart when working alone. Submitting works as it does in a taught session. Both submit_exploration() and report_result() print a pre-filled form link, and pressing Send puts your dot on the Multiverse page. What is missing is the comparison around which the day is built, with the dots of a whole class arriving on one chart within a few minutes of each other. The upper chart on that page is pre-computed and always complete, showing all 2,520 specifications with the five Multi100 analysts and the published coefficient of the original author (OA) tagged on the curve. The live chart at the foot of the page fetches whatever submissions have arrived and redraws when you press Run Code, and between deliveries there may be very few of them, or none at all. So read the comparisons you can actually make. Your placement in the grid of 840 specifications, which try_spec() and run_spec() print beside your coefficient on the browser lab page, can be read against the five analysts, whose placements run from about the 16th percentile to about the 68th. And my_experiments() prints the small multiverse that you built yourself, which stands in, on a small scale, for the spread of results that a class would produce.
By the end of the stage you have reproduced t = −3.804 yourself and completed the preregistration block before running the analysis that it fixes. The Multiverse page carries two dots of your own, one from the exploration in Task 4 and one from the specification preregistered in Task 5.