Field Journal

One exchange a week, turned into a rule someone else could run, then tested and revised. Eleven Sundays; your best ten count.

Weight: 30%. Turned in: weekly, 11 times

Submission instructions: the hand-in process and whether uploads are available.

Once a week, catch the protocol doing something, and write down the rule.

The Field Journal is the course’s main instrument. Every lecture gives you a part of the protocol to look for, every lab gives you practice looking, and every Sunday you turn one real exchange into a rule you can test.

It starts in week 2 and runs to week 12: eleven submissions, each worth 3%. Your best ten count, so one week is yours to drop.

What an entry looks like#

Each entry has four parts, the shape of a good engineering retrospective:

1. What happened. One concrete exchange: who (by role, not name), where, and what was said and done.
2. The rule you inferred. Stated so someone who wasn’t there could run it.
3. The test. A second time you checked the rule, including if it failed.
4. The rule now. Revised or held, and why.

Checking a rule does not require changing your behaviour: you can observe a second exchange while choosing to run or decline a subroutine. A prediction about what would have happened is worth recording, labelled as a prediction; it does not replace the observed second instance.

A short example, from week 4’s material:

What happened. Tuesday, group meeting of five. I asked “can everyone do Thursday 2pm?” Two people said “yes” straight away. One said “…yeah, should be fine” after a pause. That person didn’t come on Thursday.

The rule. In this group, a yes after a noticeable pause means “probably not, but I don’t want to say so in front of everyone.”

The test. Friday, I asked the same person one-to-one, by message: “Are Thursdays usually good for you? Totally fine if not.” Reply: “Honestly Monday is better.” That was follow-up on the first episode. At Friday’s group meeting, I asked whether people could stay ten more minutes. Another member paused, then said yes. I checked: “Do you need to go? Fine if you do.” They said, “No, I was adding up the time,” and stayed for the ten minutes.

The rule now. The second delayed yes did not predict a refusal. I’ll check availability once, with an explicit exit, rather than decode the pause as a no. Neither episode tells me whether being in front of the group caused the pause.

What you submit#

One file per week, containing at least one entry. Write as much as you need: fifty words or five thousand are both fine. There is no word limit and no minimum.

Grammar, spelling and presentation are not marked. Talk your observations through with an AI as much as you like. The writing must be yours.

When it’s due#

At 23:59 Canberra time on each Sunday of teaching weeks 2 to 12. Nothing is due over the two-week break. The full list of dates is above.