# How to choose a research method

The method follows the question. Do not pick the method first.

A question about a change in a number points at a trial or a cohort. A question about what people say points at interviews or documents. A question about what has already been published points at a review. Name the method after the question, then say why the other methods do not fit.

## Steps

1. **Read the outcome.** A number, a text, or a map of prior studies.

2. **List the methods that can produce it.** One primary method. One you rejected.

3. **Write the reject reason.** Cost, access, or a mismatch with the question. Not taste.

4. **Search that method.** Add the method word to the search so you do not screen the wrong study type.

## ScholarIQ at this step

ScholarIQ finds published studies of that method. It does not run the trial or the interviews.

## How long does this take for one person?

Count the records first. The method follows the question. Do not pick the method first. The time follows the count, not the size of a team.

## What do I write down before I stop?

The rule, the date, and the count. A question about a change in a number points at a trial or a cohort. A question about what people say points at interviews or documents. A question about what has already been published points at a review. Name the method after the question, then say why the other methods do not fit.

## Do I need a second reviewer?

Not to start. ScholarIQ does not score a second reviewer. If the protocol asks for dual review, a second person sees the same record.

## What is the usual error?

A file with no reason. A question about a change in a number points at a trial or a cohort. A question about what people say points at interviews or documents. A question about what has already been published points at a review. Name the method after the question, then say why the other methods do not fit.

## How do I know this step is done?

The count matches the records. If a flow-diagram box would not match, the step is not done.

## What goes in the methods?

The rule, the date, and the count. A reader must be able to open the same record.

## Can I skip this step?

No, if you will call the work a systematic review. If you skip it, write what you did. Do not hide the gap.

## What if I cannot finish?

Stop. Write how far you got. Records you did not screen are not exclusions.

## What does a supervisor check?

The rule, a sample of decisions, and the passage.

## What is outside this step?

A guideline panel, a CRO, and a GRADE meeting. ScholarIQ finds published studies of that method. It does not run the trial or the interviews.