In plain words: five short writing habits, on a fixed schedule, for people who work and study for a master's degree. Each one is small: you start with fifteen minutes a week. You keep a dated record, in your own words, of what you decided, what you expected, and what really happened, so you can see where you were right and where you were wrong. AI can help around it. The record itself is yours to write.
Reflection answerable to action, for working master's students in the age of generative AI. Five short reflective practices on a fixed cadence produce a dated record, in your own words, of what you decided, what you expected, and what really happened, so you can see where you were right and where you were wrong. The model describes four phases: before the program, during coursework, during the thesis, and the first years after. This page is written for the person in the middle of it, working now and studying now. The model is the work of Dr Phuong Nguyen. It is described in a preprint on EdArXiv (September 2026) and has been submitted to a peer-reviewed journal.
On this page: The five practices · A worked example · The four phases · Examples by phase · How to start · When it gets hard · What it claims
Five short writing habits, on a fixed schedule, for people who work and study for a master's degree.
Graduate education's structural promise, that academic study and professional work reinforce each other, rarely happens automatically. The Discerning Practice Model addresses this directly. It is not a curriculum, not an assessment framework, and not a learning theory. It is five practices, done on cadence, that produce a dated record, in your own words, of what you decided, what you expected, and what really happened, so you can see where you were right and where you were wrong.
Each practice produces a short piece of writing you keep. Over the program they add up to one record of your own decisions, in your own words, with dates attached. A year in, you have something no transcript and no reference letter can show: what you decided, what you expected, and what really happened.
The practices share one property: each reflection is answerable to action. Answerable to action means two things: that a reflection is incomplete until it commits you to a decision, and that its worth is judged by what that decision produces. The first distinguishes it from reflection written for a reader; the second distinguishes it from reflection a tool can write. A language model can draft a reflection. It cannot make the decision on Monday, be wrong about it, and discover that it was wrong.
The practices are built around the judgment only you can make: writing from memory, predicting before you ask, naming a principle and where it misleads you. Used around the practices, an assistant helps. Used in their place, it leaves a draft you can save, but not a record you have checked against what really happened.
Start with the Friday Synthesis: fifteen minutes a week, and one line is enough in a busy week. Add the others one at a time. At full intensity all five take about two to three hours a week, and they replace less-disciplined versions of things you already do. The discipline is the cadence, not the volume.
See what you keep avoiding
15 min a month
Two short lists: what you take on, and what you avoid.
Your week stays, in your own words
15 min a week
Four short answers about one real moment. In a busy week, one line.
See where you were right, and where you were wrong
Four lines per decision
Before you decide: what you expect, what would prove you wrong, and a date to check.
The rules you really work by, and where they fail
Once a term
Keep the lessons that came up twice, each with its limit.
See what the degree really changed
Every 3 months
15 minutes. What you can do now, and what you still cannot.
You start with the Friday Synthesis. Add the others when you are ready. Names as in the model's preprint. Here the Capability Inventory runs every three months; the preprint sets it once a year.
A short introduction to each of the five practices that make up the Discerning Practice Model. Each one is small, and each one gives you something to read back. Start with fifteen minutes a week. At full intensity, all five take roughly two to three hours per week, replacing less-disciplined versions of activities you are already doing rather than adding new ones.
| Practice | When | What you do | What you keep | What it gives you |
|---|---|---|---|---|
| The Stretch Inventory01 · Experience | Once a month, about 15 minutes | Write what you are taking on, at work and in your studies, and what you are avoiding. | One page a month | See what you keep avoiding. What you avoid, written down, is something you can read back next month. You choose the stretch; nobody assigns it. |
| The Friday Synthesis02 · Synthesize | Every Friday, 15 minutes to start, 30 as the full practice | Answer four questions about your week, in your own words. | Four short paragraphs, dated | Your week stays, in your own words. It turns a busy week into one principle and its limit. The limit is what makes the principle usable. |
| The Decision Hypothesis03 · Apply | Before any decision that matters | Write the decision, what you expect, what would prove you wrong, and the date you will check. | Four lines, then a check-back | See where you were right, and where you were wrong. A prediction written before the result cannot be rewritten after it. |
| The Operating Manual04 · Frame | Once a term | List the principles you really work by, each with the situation where it would mislead you. | Two to three pages, one version per term | The rules you really work by, and where they fail. Comparing two versions shows what you believed then and what you believe now. |
| The Capability Inventory05 · Grow | Every three months, about 15 minutes | Answer three questions: what I do better, what I still cannot do, and what I do on tenure rather than growth. | One page every three months | See what the degree really changed. Week to week nothing seems to move. A look at the record every three months shows what did. |
Each practice runs on its own clock. Nobody starts all five at once: begin with the Friday Synthesis, then add the Decision Hypothesis. The cards below explain each one.
A single page capturing what you are taking on, at work and in any coursework, and what you are avoiding. What you avoid, written down month by month, is a record you can read back.
Four written prompts in prose. What happened this week. What assumption did I make that turned out wrong. What general principle does this illustrate. Where would this principle mislead me. The fourth prompt, the boundary statement, distinguishes a real principle from a slogan. The full practice is thirty minutes; fifteen is the form for students establishing the habit.
Four lines on an index card: the decision, what I expect to happen, what evidence would falsify my expectation, the date I will check back. The check-back date is non-negotiable. Without it, the prediction is hope, not hypothesis.
Two to three pages of the principles you actually operate by, each paired with the situation where the principle would mislead you. The difference between two versions is a dated record of what you wrote then and what you write now.
Three honest questions. What can I now do better than three months ago. What can I still not do that I should. What am I doing on tenure rather than on growth. The third question is the one people tell me they skip. It asks you to separate what you can now do from what came with time in the job.
Written by Dr Nguyen a day late, between conference sessions, in about fifteen minutes. The full story is Reflection 04.
The first answer to question four was a quick sentence that only filled the box. The second answer is the one that makes the principle usable on Monday. That is why question four is there: every principle has boundary conditions, and writing them down is the practice.
Write your own this Friday: the four questions arrive free by email
Three phases run in order: before the program, coursework, and the thesis. The fourth, resonance, runs through all of them and keeps going after graduation. The practices never change. The question they answer does.
From the admission letter to the first class. Usually two to six months.
What goes wrongThe wait feels empty, so it goes on celebration and forum checking. Yet it is the only time you can record how you think before the program changes it.
The questionWhat do I currently believe and how do I currently decide, before the program begins to overwrite me?
What you produceAbout ten hours in total: a first Stretch Inventory, a first Operating Manual, a first Capability Inventory, and three to five Decision Hypotheses.
About twelve months in a two-year program. Six to ten months in an executive program.
What goes wrongExams reward recognizing a framework. Work rewards using it. Fluent AI output makes it easy to feel you know more than you do.
The questionOf what I am studying this week, what will I have actually changed about how I work in three months?
What you produceWeekly Syntheses and Decision Hypotheses tied to what you are studying, and one Operating Manual revision each term.
Usually six to twelve months.
What goes wrongWork and thesis compete for the same evenings. Without a dated record, memory replaces evidence under deadline.
The questionWhat does my dated record of the past year reveal that no literature search will?
What you produceA year of your own dated entries becomes first-hand material for a thesis on a real problem at work.
Through the whole degree, and after it.
What goes wrongClass and work run side by side and never meet. After graduation, you cannot say what the degree changed about how you work.
The questionWhere, this week, did the classroom and the workplace teach me different things about the same problem, and which was right?
What you produceA weekly note of where class and work disagreed, and your answer to which one was right.
Five short illustrations, one for each point in the degree. They are written to show the shape of each artifact. They are not accounts of real people.
A sales manager, admitted to an Executive MBA that starts in four months.
In month nine of the program, these lines are the "before" picture. Nothing else can capture it later.
A brand manager in a pricing module. On Monday, the boss asks for a 20 percent promotion.
He now brings a split proposal instead of a yes or a no.
An operations lead. One supervisor keeps missing the handover meeting.
If the check-back proves the lead wrong, it was attitude after all. The next conversation is firmer, and it rests on a record.
A finance manager in month twelve who must choose a thesis topic.
Dated, first-hand material from her own company, instead of a topic picked from a gap in the literature.
Two years after graduation, interviewing for a country role.
A specific, dated answer, instead of "the MBA broadened my thinking."
People rarely stop because the practice is hard. They stop because they started all five at once, or missed a week and decided they had failed. Start small, and treat a missed week as normal.
What you need: a notebook or a notes app, and fifteen minutes on Friday. Nothing else.
| Practice | The least that still counts |
|---|---|
| Friday Synthesis | One line: what was I wrong about this week? |
| Decision Hypothesis | One a month, four lines. |
| Stretch Inventory | One line: what am I avoiding? |
| Operating Manual | Skip it until the end of term. |
| Capability Inventory | Every three months. Never skip this one. |
Week one starts with the Friday Synthesis: free by email, every Friday
These are the problems people meet while doing the practice, with the fix that usually works. Find yours and try the fix this week.
Copy these into a notebook or a private notes app. Title every entry with its date. Never edit an old entry; add a dated note underneath.
Every Friday
Before the decision
For a large decision you may add two optional lines: my two key assumptions, and the option I rejected and why.
On the check-back date
First Friday of the month
End of each term
Every three months
Use AI around the practice, never inside it. The writing in your own words is the work.
Look for these in your own record, not in how you feel about it.
Where it stands. The model is described in a preprint on EdArXiv (September 2026) and has been submitted to a peer-reviewed journal.
The paper is conceptual. It draws on the learning literature and contains no data from students or classes. The model has not yet been tested at scale, so read it as a careful proposal, not a proven method.
None of the five practices is new on its own. Each has a neighbor in management practice. What the model adds is the way they fit together: private, on a fixed rhythm, each principle paired with its limit, and kept across the whole degree.
| Familiar method | What it shares | What the model adds |
|---|---|---|
| Decision journals | Writing down the reasoning before deciding | One part of a system, not a tool on its own |
| Pre-mortems | Naming what would prove you wrong | Repeated with every decision, not a one-off planning step |
| Stretch assignments | Growth through discomfort | You diagnose your own stretch; nobody assigns it |
| Critical incident reviews | Drawing principles from events | A weekly rhythm, not single incidents |
| Capability portfolios | Evidence of growth | Private and diagnostic, not written for an audience |
| Limit | What it means for you |
|---|---|
| A dated record can still be selective | Keep originals unedited, and compare them with what happened later. |
| Results carry luck and politics | A good outcome does not prove good judgment. Judge the reasoning at the time as well as the result. |
| Privacy can clash with a sponsor | Where an employer expects to see the work, the model offers no mechanical fix. The person keeping the record manages it. |
| It takes time | Two to three hours a week at full intensity. The minimum version exists for the weeks that do not have them. |
| "Judgment" is hard to measure | A fair test would look for visible behavior: naming uncertainty before deciding, stating what would change your mind, and changing it when that evidence arrives. |
The Discerning Practice Model stands on three foundations: thirty years of country leadership, nine cohorts of classroom observation, and the adult-learning tradition cited throughout this site. The programs listed here are not its source. They are where the gap between attending a program and practicing a discipline first became visible, firsthand, from inside the room. What the model extracts from each is the repeatable practice the event itself could not sustain. The practices were built where the work happens.
Built specifically for senior Asian executives transitioning into country and regional leadership. Strengths-based assessment, executive coaching, and developing leadership presence in cross-cultural commercial environments. Influenced the Operating Manual practice, particularly the discipline of articulating principles in your own words rather than borrowing them.
Forty hours of executive coaching paired with a peer cohort of senior Asia-Pacific leaders. The most direct influence on the Capability Inventory. The Korn Ferry methodology insists on distinguishing capability development from career advancement, the central question of the inventory.
The structured framework J&J used to define what a marketing professional should be able to do at each career stage, not what they should know. The competency-not-knowledge distinction is the spine of the entire model.
The advanced module of the Blanchard partnership program. Situational leadership, performance conversations, and feedback that creates change rather than defensiveness. Influenced the Friday Synthesis practice, particularly the requirement to write the boundary statement, where would this principle mislead me, not just what worked.
Basic management discipline: setting direction, structuring meetings, allocating time, building trust. Most managers never learn this formally. Most never need to, and that is precisely why mid-career plateaus happen so reliably.
Certification earned through an actual process improvement project at Janssen-Cilag Vietnam, not a classroom exercise. The DMAIC discipline (Define, Measure, Analyze, Improve, Control) is the structural ancestor of the Decision Hypothesis. The insistence on measurable predictions and post-implementation review is identical.
The full process excellence curriculum: Six Sigma analytical rigor, Lean's discipline of removing what does not create value, and Design Excellence's principle that quality must be designed in, not inspected in. Together, these taught the lesson behind the model: a practice you keep on a schedule is one you can look back on; talent without a record leaves nothing to look at.
Every practice in the Discerning Practice Model is the distilled, simplified, generalized version of something Dr. Nguyen first encountered in formal training and then applied across hundreds of real management decisions.
The Friday Synthesis is not the Ken Blanchard reflection model, but it would not exist without it. The Decision Hypothesis is not Six Sigma DMAIC, but it would not exist without it. The Operating Manual is not the Korn Ferry leadership philosophy exercise, but it shares its DNA. The Stretch Inventory carries the J&J Marketing Competency Models inside it.
What the model does is take the irreducible core of each, the smallest repeatable practice behind what the original program was trying to teach, and deliver it as a habit a person with a job and a class can actually keep on a Friday afternoon.
In one 2025 study, university students were randomly assigned to learn with or without ChatGPT. The group with the assistant wrote essays that scored higher and showed no greater knowledge gain. The authors named the pattern metacognitive laziness: students outsourced the planning, monitoring, and evaluation phases of learning to the tool rather than scaffolding their own thinking with it.
The Discerning Practice Model is built around the judgment only you can make: writing from memory, predicting before you ask, naming a principle and where it misleads you. Used around the practices, an assistant helps. Used in their place, it leaves a draft you can save, but not a record you have checked against what really happened.
Grades and an assistant's draft show what you produced. The record shows what you decided, what you expected, and what really happened.
The record is what you have checked against what really happened. A saved draft is not yet that record.
The Discerning Practice Model rests on six classical pillars of learning theory together with the self-regulated learning tradition, and on a body of recent empirical work that has reaffirmed their relevance. Both layers are listed below.
Four questions, in your own words, on paper or in a document of your own. One prompt arrives by email every Friday at 07:00 GMT+7. Fifteen minutes to start. Thirty when you have them. The practice is free and stays free.
Start the free weekly practice