How to compare options before making a decision

To make a considered decision, define the question, separate essential conditions from preferences and compare available options using the same criteria. Record missing information separately. The outcome may be a choice, or a specific check you need before choosing.

Consider an illustrative example: you want to learn to make work presentations and are choosing between independent study and a group course. This is a way to organise a comparison, not a recommendation for any particular course. You can adapt it to a workflow or an approach to a project.

Define the result and essential conditions

Replace “Which option is better?” with “How can I prepare and deliver a trial presentation by the agreed date?” Establish who will assess the result and what feedback you need. If the output is still vague, start with the distinction between a goal, project and task.

List the constraints: available time, acceptable costs and the need to review your own work. An impossible timetable cannot be offset by an instructor's popularity. Attractive materials and being able to attend the sessions have different importance: one may be a preference, while the other is essential.

Set the decision deadline and identify who has authority to choose. If the choice affects colleagues, agree on the essential conditions first. Otherwise, you may discover later that everyone was solving a different problem.

If the choice requires agreeing on scope or timing with someone else, use the work-negotiation preparation guide. It helps turn an option into a specific proposal and distinguish it from a confirmed agreement.

Build a small table of facts

Complete the same fields for each option. “Unknown” is more useful than a confident guess because it identifies the next question. Do not add a criterion solely because your favourite option scores well on it.

The official multi-criteria analysis manual covers option appraisal and sensitivity to input judgments. Our example below is much simpler than a full model. The table makes the reasons behind a choice visible.

Test two options against the same example

Suppose independent study fits your timetable, but you have not found anyone to review the presentation. The group offers assignment feedback, but its mandatory meeting times are unknown. Neither is an established winner: each has an important gap.

• For independent study Ask a specific person whether they can provide feedback and when
• For the group Obtain the timetable and an explanation of which assignments receive feedback
• For both Check whether preparing the trial presentation fits the time available

If neither meets the essential conditions, change the scope, keep looking or agree on another deadline. You do not have to choose between two unsuitable offers simply because they were the first options on the list. A further option should answer the actual gap, rather than restart the entire search without a purpose.

Use scores only when their meaning is clear

For a simple choice, facts and a short explanation of the trade-off may be enough. If you use scores, define the scale and explain the priority of each criterion beforehand. Do not automatically turn an unknown value into an average score.

Check whether a different reasonable priority or a corrected uncertain fact changes the conclusion. If it does, record that dependency and obtain the missing information. Avoid counting the same advantage repeatedly under different labels. The manual discusses both sensitivity and double counting in detail.

Numbers do not turn preferences into objective facts or guarantee an outcome. To examine the influence of the first estimate and a favoured option, use the guide to judgment errors in decisions.

Record the decision and the next check

A short note could read: “I choose independent study because the group's timetable does not fit. A colleague has agreed to review a draft. Next, I will outline the presentation. If feedback becomes unavailable, I will reconsider the learning arrangement.” Fill this in using confirmed agreements.

Do not keep gathering information without a question to answer. If research keeps expanding, set a stopping condition with the information-overload guide. If circumstances change substantially, return to the process for revisiting plans.

A useful decision record does not promise a perfect choice. It lets you explain the reasoning, take the next action and identify which new information would actually justify choosing differently.