Skip to main content

How to Create and Organize Your Extraction Questions

  • Updated

Each extraction question becomes a column in the extraction table. The Column Title is the column header shown in the table and in your exported CSV. The Question field is the full prompt collaborators read and answer for each article.

Create your questions before your team starts extracting — changing the form structure after extraction has begun may require collaborators to redo work.

How to create a question

  1. Go to the Data extraction tab.
  2. Click + Add Question on the right side of the screen.
  3. Enter a short Column Title — for example: Study design, Sample size, Primary outcome.
  4. Enter the full Question prompt — for example: What is the study design?, How many participants were included?
  5. Select an Answer type (see table below).
  6. Toggle Required Question if this field must be completed before an article can be marked as done.
  7. Click the Group under dropdown. Type to search existing groups or type a new name and select Create "..." to create one on the fly.
  8. Click Add Question to save.
Important A group must be selected or created before you can save a question — the Add Question button stays inactive until the Group under field is filled. If no groups exist yet, type a name in the dropdown and select Create "..." to create your first group.

Answer types

Answer type What collaborators enter Best for
Free Text Any text Variables where answers are too varied to predetermine — key findings, study conclusions, or specific data quoted from the article
Numbers A numeric value only Quantitative data — sample size, mean age, follow-up duration, effect size
Specified One or more options from a predefined list Fixed-answer variables — Yes/No, study design types, risk ratings. Prevents unmatched answers across collaborators.
Best practice Use Specified answer types wherever possible. In collaborative reviews, Free-text answers for the same variable can produce unmatched answers across collaborators — for example "RCT", "Randomized controlled trial", and "randomised trial" are three separate values that won't align automatically. Predefined Specified options eliminate this inconsistency at the source.
Note When collaborators give different answers to the same question, Rayyan flags these as unmatched answers. Resolving them works similarly to resolving screening conflicts — collaborators compare responses, discuss, and align on the correct answer before export. See How to Use Blind Mode in Data Extraction for the full reconciliation workflow.

How to organize questions into groups

Groups are named sections that organize your questions into logical categories, making the extraction form easier to navigate. Common groups include Study Characteristics, Population, Intervention, Outcomes, and Results. You can create groups on the fly when adding questions, or plan them in advance before building your form.

To move a question to a different group, open the question, update the Group under field, and click Save changes. Closing the panel without saving discards your changes.

How to reorder columns

Drag and drop columns in the extraction table to reorder them. This changes how questions appear during extraction — but does not affect column order in the exported CSV. The CSV always reflects the order questions were created, not the order they appear in the table. If CSV column order matters for your analysis, plan your question creation order before you start.

Note Column names in the exported CSV are prefixed with a number reflecting creation order — for example, 671-STUDY DESIGN, 672-SAMPLE SIZE. This numbering is assigned automatically by Rayyan and cannot be changed.

Continue to: How to Manage Articles and Questions in Data Extraction

Related articles

Comments

0 comments

Article is closed for comments.