A decision-making framework helps you make a choice that you can explain and revisit. It separates requirements that must be met, preferences that can be traded, and facts you still need to verify. A weighted score can organise your thinking, but it cannot make an uncertain offer certain or decide your values for you.
This guide uses a fictional career decision to show the whole process. The options and numbers are invented for the exercise; they are not job offers, salary benchmarks or predictions of career outcomes. You can copy the structure into a notebook or spreadsheet without buying a tool.
Write the decision as a question
Make the question specific enough to answer. “What should I do with my career?” leaves too many possibilities open. “Which of these three confirmed opportunities should I choose for the next six months?” identifies a manageable decision, provided the opportunities really are confirmed.
Record the decision owner, deadline and options. Include staying with the current arrangement when that is feasible. Note who will be affected and which conversations you need before deciding. Family responsibilities, accessibility needs, location and working hours can be material requirements; do not let a scoring exercise conceal them.
Then list the information you have and its source. A written term, a dated conversation and an assumption are different kinds of evidence. Record them separately. If a deadline is unclear, ask the person responsible for it rather than treating an online rumour as an extension.
Apply requirements before preferences
A requirement is a condition an option must satisfy for you to consider it. A preference improves an acceptable option. Define this distinction before looking at totals, because a strong preference should not compensate for failing a necessary condition.
In the fictional example, Mira is choosing an internship while completing a course. She sets three requirements: no clash with compulsory classes, a written description of duties, and a named person to whom she can raise a work question. These are Mira's requirements for this exercise, not conditions every internship must use.
She has three options: a campus service team, a remote documentation team and a local support team. Suppose all three provide the required written terms and compatible schedules. They can enter the comparison. If the remote option's class compatibility were still unconfirmed, it would remain “pending verification”; its attractive flexibility score would not make it eligible automatically.
Use a small requirements table with the condition, evidence, status and next action. “Unknown” is useful information. Keep it visible instead of converting it into a pass or a fail without support.
Choose criteria that describe your actual priorities
Mira now chooses three preferences: learning fit, schedule flexibility and mentorship availability. She defines a score from one to five for each, using the same meaning across all options.
For learning fit, a one means the described tasks have little connection to the skill she wants to practise; a five means most of the described tasks directly involve that skill. For flexibility, scores describe the confirmed ability to move hours around classes. For mentorship, scores describe a stated review arrangement, not the reviewer's reputation or a promise of rapid progress.
These are personal judgements based on information available to Mira. They are not validated assessment scales. She should write a short reason for every score and change it if the evidence changes. “The company sounds impressive” is not sufficient evidence for a five in mentorship.
Keep the list short enough to avoid counting the same preference twice. “Flexible hours” and “convenient schedule” may largely describe the same benefit. If both receive large weights, that benefit can dominate the result accidentally.
Work through a weighted comparison
Mira gives learning fit a weight of 0.50, flexibility 0.30 and mentorship 0.20. The weights sum to 1.00. She chooses them before computing the totals and records why learning currently matters most to her.
The scores below are listed in this order: learning, flexibility, mentorship. Campus, remote and local refer to the three teams described above.
| Option | Scores |
|---|---|
| Campus | 5 / 2 / 4 |
| Remote | 3 / 5 / 3 |
| Local | 4 / 3 / 5 |
The weighted total equals learning score multiplied by 0.50, plus flexibility score multiplied by 0.30, plus mentorship score multiplied by 0.20.
For the campus option, the total is 5 × 0.50 + 2 × 0.30 + 4 × 0.20 = 3.90. The remote option totals 3 × 0.50 + 5 × 0.30 + 3 × 0.20 = 3.60. The local option totals 4 × 0.50 + 3 × 0.30 + 5 × 0.20 = 3.90.
The first and third options tie. That is a reason to examine the trade-off, not add an arbitrary decimal to manufacture a winner. The campus option scores higher on learning; the local option scores higher on flexibility and mentorship. Mira can ask which difference is more consequential and whether the scores rest on equally strong evidence.
The totals are comparison aids on this invented scale. A 3.90 does not mean a 78% chance of success, and the 0.30 difference does not measure future income or satisfaction.
Check how sensitive the answer is
A useful comparison should show whether a small change in priorities changes the ranking. Suppose Mira now considers weights of 0.30 for learning, 0.50 for flexibility and 0.20 for mentorship, reflecting a possible heavier course schedule.
The campus total becomes 3.30, the remote total 4.00, and the local total 3.70. The remote option moves into first place. This tells Mira that flexibility has a meaningful effect on her choice; it does not establish which set of weights is correct.
Also test uncertain scores. Suppose the local team's mentorship score might be three rather than five because review availability has not been clarified. Under the original weights, its total would fall from 3.90 to 3.50. A short question about the review arrangement may therefore be more useful than another hour changing spreadsheet colours.
Record both scenarios. Do not quietly select whichever assumption supports the option you already prefer. If the decision repeatedly changes under plausible assumptions, identify what additional evidence could resolve the uncertainty and whether you can obtain it before the deadline.
Gather evidence without pretending to have it
For Mira, three practical questions are: what tasks will occupy a typical week, which hours are fixed, and when is feedback available? She can ask the relevant contact for an answer or example, while recognising that a description of expected work is not proof of how every future week will unfold.
Keep a dated note with the question, the answer, its source and what remains unclear. If the contact does not answer, record that fact. Do not turn silence into confirmation. If a trial task is offered, clarify its scope, permission and expectations before doing it; a trial is not automatically an employment offer.
For an important decision, consider whether another person should review your interpretation. They may notice an overlooked constraint or weak evidence. Their perspective can inform your choice without replacing your authority to decide.
Consider failure and reversibility
Ask what could make the chosen option unsuitable. In this exercise, fixed hours could expand, the assigned work could differ from the description, or feedback could be less available than expected. Write a response you can actually take, such as clarifying expectations early or arranging a review conversation.
Distinguish a reversible experiment from a commitment that is costly to undo. A short personal practice project may be easy to stop. Relocation, fees and contractual commitments need more careful consideration of the actual terms and your circumstances. A framework helps expose these issues; it does not interpret a contract or provide personalised financial advice.
Avoid assuming there will always be another opportunity. Identify what you would do if the preferred option became unavailable. A feasible fallback might be continuing the current course and seeking another placement; an imaginary fourth offer is not a fallback.
Write the decision record
Mira's record can be brief:
“I am comparing three eligible internship options for a six-month period. My original weights give the campus and local teams the same total. Flexibility changes the ranking materially. The local review arrangement remains unclear, so I will ask about it before choosing. I will not treat the provisional mentorship score as confirmed. If no sufficient answer arrives before my own decision deadline, I will decide using the information I actually have and record the remaining uncertainty.”
This is a draft record within the fictional exercise. It does not claim Mira contacted anyone, accepted an offer or achieved a result. After a real decision, replace future actions with what actually happened and retain the earlier reasoning.
Review the process after acting
Choose a review point appropriate to the decision. Check whether your original assumptions were accurate, what new information appeared and whether your response needs to change. Separate the quality of the reasoning from the outcome: a well-considered decision can encounter unexpected events, and a lucky outcome can follow poor reasoning.
Use the review to improve future criteria and questions. Do not retrospectively rewrite uncertain information as something you knew all along. A dated record makes that easier.
Frequently asked questions
Must every decision use weights?
No. A clear requirements check and a short explanation of the trade-offs may be enough. Use weights when comparing several acceptable options with distinct priorities, and explain the score definitions.
What should I do with missing information?
Mark it unknown, identify who could answer, and consider how much the answer could change the decision. If you must act before receiving it, record that limitation rather than inventing certainty.
Does the highest total have to win?
No. A score is one input. You may discover a missed requirement or an issue the model represents poorly. Revise the framework openly and explain why; do not secretly change the calculation to justify a preferred answer.
Further reading
NASA's decision-analysis guidance discusses mandatory and enhancing criteria and the importance of documenting uncertainty in engineering decisions. That is its original scope. Mira's career exercise and scoring values here are independently invented; they are not a NASA assessment or endorsement.
Related guides
For related practice, read the risk-assessment guide and use the competency-assessment exercise to keep capability evidence separate from preferences.