Logic and randomization

Create a logic rule

Available operators: equals, not equals, includes, does not include, greater than, greater or equal, less than, less or equal, contains text, is answered and is not answered. The “Show” and “Hide” actions can target a question, a page or a section.

1. 1Open the “Logic” tab in the builder and click “Add Rule”.
2. 2Optionally describe what the rule does in “Rule name”, for example Screen out under 18.
3. 3In the “If” part, choose whether the rule runs when “all conditions” or “any condition” are met.
4. 4Use “Add condition” to set the question, operator and value. Add a “Sub-group” for nested conditions.
5. 5In the “Then” part, use “Add action” to choose the action: “Jump to page”, “Jump to end”, “Disqualify”, “Show” or “Hide”.
6. 6Use the “On” / “Off” switch to disable a rule without deleting it.

Rule check

“Rule check” lists possible problems in your rules. Make sure no findings remain before you publish the survey. It catches:

• Rules with no conditions or no actions.
• Conditions and actions that point at a deleted field or target.
• Conditions that look at a question on a page after the target (forward reference).
• Contradicting conditions for the same field.
• Rules that jump to different targets on the same condition: the rule that comes first in the list wins.

Flow map and path test

1. 1The “Flow Map” on the “Flow” tab shows the survey as Welcome, Page, Thank you and Disqualify nodes.
2. 2Dragging a connection from one node to another creates a jump rule. Finish the rule's condition on the “Logic” tab.
3. 3In “Path test”, pick answers and see the path a respondent would take, including skipped and hidden fields.
4. 4Use “Answer at random” to try different possibilities quickly.

Randomization modes

The “Randomization” tab lists everything randomized in the survey in one place. You can make the same setting in the item's own panel. Scopes: page elements, block fields, option order, and matrix rows and columns. The “Mode” options for each scope:

• None: items are shown in the order you wrote them.
• Shuffle: every respondent sees a random order.
• Subset: a random selection of the number of items you set is shown, in random order.
• Rotate: the order is kept and the starting point shifts. It approaches balance with large samples but is not a fully balanced design.
• Reverse: about half of respondents see the reversed order; used to measure scale-order effects.

Seed, anchors and blocks

The order shown to each respondent is stored with their response. The “Sample respondents” preview shows the orders three different respondents would see. The “Conflicts” list warns about problems such as logic rules that depend on a field some respondents never see.

• With “Seed” set to “Per response”, a new order is drawn for each response. With “Per participant”, the same participant sees the same order when they return; this requires marking a Hidden Field with “This field identifies the participant”. Without a marked field, the order is drawn per response.
• “Anchors”: options you anchor (for example Other or None) stay out of the shuffle and are always shown, even in subset mode.
• “Block”: used to shuffle the questions inside it or show only some of them. Set how many questions each respondent gets with “How many to show?” in the block settings.