Learn it → Apply it → Save as Skill
An unfinished task creates mental tension that keeps pulling your attention back — even when you’d rather be thinking about something else. Finishing the task releases that tension. So, it turns out, does making a concrete plan to finish it later.
The Dr. Wallace’s Favorite Mental Models is designed to turn useful ways of thinking into reusable tools. Each model gives you a fast visual overview, deeper interactive learning, practical application exercises and an AI-ready Markdown skill you can save for future use.
Instead of merely reading about mental models, you can gradually build your own AI skills library: a collection of portable reasoning skills that can help ChatGPT, Claude, Grok and other AI assistants apply the mental models you value.
Starting a task creates tension that keeps it active in memory and at the edge of attention. The tension eases once the task is actually finished — or once you’ve made a specific, concrete plan for finishing it.
Important: this shows the direction of the effect, not a measured size. The underlying memory-recall finding has an inconsistent replication record — see History & Accuracy below.
The model is most useful when work is broken into interruptible chunks: task switching, learning sessions, negotiations left mid-conversation, or marketing and content that need to hold attention across a gap.
The tension is what does the work. Closing it — by finishing or by planning — is what lets you stop thinking about it.
Starting a task, even briefly, creates goal-directed tension — the mind treats it as “live.”
Work gets paused before completion. The tension doesn’t switch off; it stays attached to that task.
While open, the task keeps nudging at attention and stays more available in mind than tasks already closed out.
Finishing resolves the tension. A specific resumption plan appears to do much of the same job without finishing anything.
Open a use case to see what the model changes.
Interrupt a study session deliberately, mid-topic, rather than always stopping at a clean break. Some evidence suggests interruption keeps material more accessible — though this is one of the less consistently replicated parts of the research, so treat it as a technique worth testing, not a guarantee.
A visible progress indicator (“Profile 64% complete”) turns an abstract task into a concrete open loop — one of the more reliably observed applications. Pair it with a clear, small next step rather than a vague “finish setup.”
Cliffhangers and open questions work by leaving a loop unresolved across an ad break, chapter, or episode. Resolve it clearly, and soon — a loop that never pays off just trains the audience to stop caring.
Before switching away from an interrupted task, write down exactly where you left off and the next concrete step. That plan appears to do much of the work of “closing the loop,” even while the task itself is still open.
If a disagreement can’t be resolved in the room, agree explicitly on what happens next and when. An unresolved point without a plan tends to keep generating friction; one with a plan tends to quiet down, even before it’s actually settled.
Name the open question out loud and set a date to revisit it. Teams, like individuals, seem to hold onto unresolved threads — making the thread visible and scheduled is often enough to let people refocus.
This is an illustrative example designed to teach the method — finish what’s small, plan what isn’t.
Expand for the distinction between the original 1927 finding and what later research actually confirmed.
In the 1920s, Gestalt psychologist Kurt Lewin noticed that a Berlin restaurant waiter could recall the details of orders that hadn’t yet been paid, but forgot them the moment the bill was settled. His student Bluma Zeigarnik turned the observation into a formal experiment: participants worked through a series of simple tasks, some interrupted before completion. Zeigarnik published the results in 1927 in Psychologische Forschung, showing better recall for the interrupted, unfinished tasks.
The explanation draws on Lewin’s field theory. Starting a task creates a state of psychological tension tied to that specific goal. The tension keeps goal-relevant information more accessible in memory. Completing the task discharges the tension; interrupting it leaves the tension — and the accessibility — in place.
No — that’s a related but distinct finding named after Zeigarnik’s colleague Maria Ovsiankina: the Ovsiankina effect, the urge to resume an interrupted activity. Zeigarnik’s original claim was about memory; Ovsiankina’s is about motivation. In the modern literature, the motivation-to-resume finding has held up more consistently than the original memory-recall finding.
Only partly, and this matters for how you use it. A 1968 study by Van Bergen found no significant recall advantage for unfinished tasks. A 2025 meta-analysis of the interruption-recall literature concluded that the classic Zeigarnik memory effect “lacks universal validity,” while the Ovsiankina resume-urge effect held up as a more general tendency. Treat “unfinished things are remembered better” as plausible and historically important, not as settled science — and treat “unfinished things keep tugging at you until resolved or planned” as the sturdier, better-supported part of the idea.
The original recall finding has failed to replicate in several studies. Use interruption as a technique to test for yourself, not as guaranteed science.
“I’ll remember it better” (Zeigarnik) and “I feel driven to finish it” (Ovsiankina) are two different claims with two different evidence bases. Don’t use one to justify the other.
Chronic unresolved loops produce anxiety and fatigue, not just helpful tension. The goal is closure — by finishing or by planning — not an ever-growing pile of open threads.
An interruption with no resumption point doesn’t reliably release tension; it just leaves you stuck. Name the specific next action and when you’ll take it.
Notification badges and manufactured incompleteness exploit the same mechanism to manipulate engagement. Be honest about what’s actually unfinished.
Motivation, fatigue, how personally invested someone is, and when the interruption happens all appear to change the size — and even the direction — of the effect.
a task requires sustained, uninterrupted focus. Deliberately opening loops during deep work fragments attention rather than helping it.
the task is small enough to finish now. Planning a two-minute task is slower than simply doing the two-minute task.
you need guaranteed recall. The memory-boost claim is one of the less consistently replicated findings here — pair interruption with spaced repetition and active recall, which have stronger evidence.
This browser tool structures your thinking. It does not send your information anywhere or perform AI analysis.
Save as Skill converts the model into portable Markdown instructions that teach an AI when and how to apply the Zeigarnik effect, what questions to ask, what evidence to weigh, what mistakes to avoid, and how to structure its advice.
Save the skills you find useful and organize them into your own reusable library. The Markdown is intentionally model-neutral, so it can be used wherever an AI system accepts uploaded files, project knowledge, custom instructions or reusable text context. Platform capabilities differ, so this is a portable skill file rather than a claim of one-click installation.
Not for the task itself — only for the mental tension around it. Masicampo and Baumeister (2011, Journal of Personality and Social Psychology) found that making a specific plan for an unfinished goal reduced the goal’s intrusive pull on attention almost as effectively as completing it, while an unfinished goal with no plan kept intruding on unrelated tasks. The work still needs doing; what changes is whether it keeps interrupting everything else you’re trying to think about.
The original theory ties the tension to genuine goal engagement. A task you’re indifferent to may never generate much tension to begin with, whether it’s open or closed — so the effect is strongest for things you’re actually invested in finishing.
Generally no. The better-supported use runs the other way: notice the loops that are already open — you have plenty — and either close them quickly or give them a specific plan, rather than manufacturing new ones for a memory benefit that isn’t reliably there.