Reducing negative and increasing positive stress
Difficulty reads as threat, and the wrong kind of pressure lands on the graded work
The seminar got tense, you invented a ground rule on the spot, and it read as a rebuke of whoever happened to be talking.
What it is
Two moves in opposite directions. Reducing negative stress removes ambiguity that was never load-bearing: rubrics and exemplars before the work is due, clear framing of what kind of difficulty a task involves, agreed language for disagreement. Increasing positive stress puts productive pressure where it does work, in low-stakes structured argument, and takes it off the graded submission.
The fellowship supplied this as poster bullets rather than a procedure, plus a flat prohibition on ever telling a student they cannot do something.
Run it
- Name which kind of hard it is before students start. Hard because nobody has written this genre, hard because the field has not settled it, or hard because the algebra is fiddly. The first two are worth sitting in, the third is not, and students cannot tell them apart from inside the task.
- Front-load rubrics and exemplars so the ambiguity is gone before the work starts rather than explained afterwards.
- Agree the language in week one, not in week nine.
- Put the arousal where it does work. A clock and a time-boxed group task are useful. The same pressure on the graded submission is not.
- Replace high-stakes surprises with low-stakes checks. Pop quizzes do the opposite of this.
Say this
Spoken before a hard task
“This next piece is hard in a specific way. It is not that the computation is fiddly. The field has two defensible answers and you are about to pick one and defend it. If you feel stuck in the next twenty minutes, that is the task working rather than you failing it. What I want at the end is your choice and your reason, not the right answer.”
Language contract, syllabus text, week one
“We will disagree in this seminar, often about methods and sometimes about topics that are not only methodological. Three things hold.
Disagree with the claim and name it. ‘That estimate assumes no unmeasured confounding, and I do not think that holds here’ is a contribution. ‘That’s a bad paper’ is not.
Use the field’s terms precisely, and say when you are not sure of one. Nobody here knows every term of art, including me, and asking what a word means is never a cost.
You may pass. On any question, in any session, without a reason.”
The third rule has to be real. If you write “you may pass” and then work the room until the quiet student answers, you have taught them the other two are also decorative.
Techniques
Write, then discuss
On any charged item, five to eight minutes of private, uncollected writing before anyone speaks. Ask for what they currently think and one thing that would change their mind. Say explicitly that you are not collecting it.
Spoken instruction
“Before anyone talks, take eight minutes and write two things: what you currently think, and one thing that would change your mind. I am not collecting this and nobody else will see it. When we come back you will have a position to speak from, so the discussion starts from twelve positions instead of from whoever is quickest.”
Decide before class whether you are acknowledging a topic or opening the floor on it. Those are different, and choosing in the moment means choosing badly.
Structured debate
The positive-stress half. Pick a disagreement that is genuinely live, assign positions a week ahead so nobody argues what they already believe, and require one page: the strongest version of the assigned argument, the single best objection to it, and what evidence would make them concede. Grade the objection, not the argument.
Assigned the week before
“You have been assigned a position. You may not have been assigned the one you hold. Bring one page: the strongest version of your assigned argument, the single best objection to it, and what evidence would make you concede. You will be graded on the objection, not on the argument.”
In my courses
Rubrics used in EPI 501. Planned for EPI 601, every week.
I will reduce negative stress by providing clear rubrics and exemplars for all major assignments before they are due. I will increase positive stress through structured in-class debates on methodological controversies (e.g., ‘Should we match or adjust?’), where students take assigned positions and must defend them with evidence. Low-stakes formative assessments (exit tickets, self-checks) will replace high-stakes pop quizzes.
The three debates named in the plan are “should we match or adjust”, “is the p-value useful”, and “additive or multiplicative scale for interaction”, each attached to the session where that content lands. The prompt came from EPI 501 survey data showing workload concerns, so the response for 601 is to front-load rubrics and exemplars and cut ambiguity-driven stress rather than to reduce the work. Used before the fellowship, in the form of rubrics.
Evidence
No study supports any of this directly, and the fellowship supplied none.
Write-then-discuss was the only move on the fellowship’s difficult-conversations poster marked with agreement. Its mechanism is the private-first step that also opens any structured activity, which has better support.
The nearest studied relative of the debate is constructive controversy, which has support but comes from K-12 settings and was evaluated largely by the people who developed it (Johnson & Johnson, 2009).
Naming which kind of hard a task is has no source at all. It is mine. The argument is structural: a student who knows which kind of difficulty they are in can decide whether to push or ask, and one who does not defaults to assuming the difficulty is them.