Between-Subject Experiments

Experiments Surveys & forms

Between-subject designs require different groups of participants to see different stimuli, trial sequences, or instructions. Testable handles this with the **`subjectGroup`** column.

How it works

Each row in the trial file has a subjectGroup value. When a participant starts an experiment they are automatically assigned to a group. Testable then only runs the rows that match their group.

• Rows with no subjectGroup value are shown to all participants.
• Rows with a value are shown only to participants assigned to that group.

Basic two-group design

Set subjectGroup to a number (the group size) on each condition-specific row:

Row	subjectGroup	description
Instructions	(blank)	Shown to everyone
Condition A trials	1	Shown to Group 1 only
Condition B trials	2	Shown to Group 2 only
Debrief	(blank)	Shown to everyone

The number indicates the relative size of each group. A value of 1 in all groups means equal allocation. Use different numbers for unequal allocation (e.g. 1 and 2 gives a 1:2 ratio).

Note

If you have more than two groups simply add more unique numbers. Each unique number creates a separate group.

Multi-level (factorial) groups

Combine multiple grouping dimensions with a colon-separated label:

subjectGroup = A:1
subjectGroup = A:2
subjectGroup = B:1
subjectGroup = B:2

This creates four groups: A:1, A:2, B:1, B:2. Each participant is assigned to exactly one cell. The number after the colon is the relative size; the letter before it is the group name.

Rows with a partial match (e.g. A) without a sub-group level are not used for multi-level designs, every distinct A:B combination is a separate group.

Even (balanced) allocation

By default Testable assigns participants sequentially to groups. To ensure perfectly balanced assignments across parallel sessions, enable Balanced Group Assignment in Setup → Participants & Data. This forces strict alternation so that groups remain as even as possible even when multiple participants start simultaneously.

Viewing group assignments

Each result file includes a column identifying the participant’s group assignment, so you can filter results by group during analysis.

Next steps

• Randomisation: randomise trial order within a group
• Participants & Data: enable balanced group assignment