
Factorial Designs
A factorial design tests two or more factors at once by crossing their levels, so that every combination becomes a condition. In Labvanced, you set up the factors and levels visually and the conditions are built for you, with no code needed. Seeing the design take shape also makes it well suited to teaching.
Table of Contents
Build Your Design From Factors and Levels
You have the option to define the factors and levels of your design, and Labvanced builds the table of conditions from them, updating it whenever you add or change a factor. Each condition has its own trials, so a design is ready to run as soon as the factors are in place.

Trials_desktop task in the study for more.A ready-to-run template in which stimulus type and target position are set up as factors, producing the conditions shown above.
Randomization and Balancing
You can make a factor fixed, so every participant sees every level, or randomized, so variation such as the position of the correct answer is drawn from for each trial without adding trials. Randomization settings let you control trial and condition order and apply balancing, including nested balancing, so conditions are replicated equally. A preview of the trial sequence shows the result before you collect data, and order can also be randomized across blocks and sessions.
| If your design needs... | Approach in Labvanced |
|---|---|
| Every participant to experience every condition | Fixed factors |
| Stimulus position or style varied without adding conditions | Randomization factors |
| Equal numbers of trials per condition | Balancing options in the randomization settings |
| Different participants to receive different condition orders | Between-participant balancing and groups |
| Blocks or sessions presented in varying order | Randomization separators in the Study Design tab |
For the full approach to randomization and balancing in Labvanced, see Randomization and Balancing.
Examples of Data You Can Collect
A factorial design can use anything a Labvanced study records, and these are common examples. Every trial records the level of each factor, so each measure carries the condition it came from and your exported data can be split by condition directly in your analysis software. The same factor variables can also drive events, such as showing different feedback when a factor takes a particular level.
| Data | What it captures |
|---|---|
| Reaction times and accuracy | Recorded for every trial and tagged with its factor levels, so each cell of the design can be compared. |
| Physiological signals | Gaze, facial emotion, and heart rate recorded together from the participant's webcam, so each condition can be compared on physiological signals too. See Remote Physiological Measurement. |
| Questionnaire ratings | Ratings collected after or between conditions, such as a rating of each stimulus. See Questionnaires. |
| Cursor behavior | Cursor position and movement within each condition. See Mouse Tracking. |
| Gaze | Where participants look in each condition, measured with webcam eye tracking. See Webcam Eye Tracking. |
| Spoken responses | Voice answers recorded per trial. See Audio Recordings. |
Remote Physiological Measurement
Gaze, emotion detection, and heart rate detection from one webcam.
Questionnaires
Rating scales and item responses for each condition.
Mouse Tracking
Cursor position and movement during a task.
Questions about your study design or platform fit?
Teaching Factorial Designs
Students can use Labvanced to gain hands-on experience building experimental paradigms, factorial designs among them, because the visual way factors and levels turn into conditions is easy to follow. A student can watch the table of conditions change as a level or a factor is added, which makes factorial logic concrete before they run a study of their own. See Labvanced for the Classroom for how courses use the platform.
Popular Research Applications
These are popular applications of factorial designs across experimental psychology:
- Cognitive psychology: Designs that cross stimulus properties, such as category and position, test their separate and combined effects. Interference paradigms treat congruency as a factor with congruent and incongruent levels, as in the Stroop task and the Flanker task. See Cognitive Psychology.
- Marketing research: Designs that vary product attributes across conditions show which attributes matter on their own and in combination. Attributes can be varied within participants or, through groups, between them. See Marketing Psychology.
- Social psychology: A manipulation can be crossed with a second manipulation or with the group a participant belongs to, to test whether an effect depends on the situation. Groups handle the between-participant factor, while the trial structure handles the rest. See Social Psychology.
- Economics: Choice experiments can vary factors such as stakes and framing across conditions to estimate their separate effects and their interaction. Randomization and balancing keep the conditions comparable. See Economics.

Cognitive Psychology
Stimulus properties crossed in one design.

Marketing Psychology
Product attributes varied across conditions.

Social Psychology
Manipulations crossed with group.

Economics
Stakes and framing across conditions.
Choose the Experimental Design That Fits Your Research
Whether you want to compare different groups of participants, measure changes within the same participants, or follow them over time, Labvanced gives you the flexibility to build your study around your research needs.
- Within-subjects design: Have every participant experience all experimental conditions. For example, participants might complete both an easy and a difficult version of a task, allowing you to compare their performance.
- Between-subjects design: Divide participants into separate groups, with each group experiencing a different condition. For example, one group receives positive instructions while another receives neutral instructions.
- Mixed design: Combine both approaches in one study. For example, participants are assigned to different instruction groups, but everyone completes both the easy and difficult versions of a task.
- Longitudinal design: Follow the same participants across several sessions over days, weeks, or months. For example, participants complete the same task at the start of a term and again at the end, allowing you to compare conditions or performance over time.
How does this work in Labvanced? Define your experimental factors and levels to create within-subjects conditions, use groups to organize your between-subjects conditions, and use sessions to spread a study over time. This makes it possible to create flexible experimental designs within a single study. See Single Session Studies for how groups assign participants and Longitudinal Study Design for scheduling sessions.
Setting It Up in Labvanced
Factors and levels are set up in the task editor with no code required. For the step-by-step, see the Trial System guide and the Randomization Settings reference.
Frequently Asked Questions
How does Labvanced create the conditions of a factorial design?
Can I run a mixed design with both between-subjects and within-subjects factors?
Is the factor level recorded for each trial?
How many factors should I add to a design?
Further Reading
Trial System
Trial groups, factors, levels, and the conditions table.
Randomization and Balancing
Examples of randomization and balancing across stimuli, positions, blocks, and sessions.
Labvanced for the Classroom
Use Labvanced in teaching so students build experiments and collect data themselves.
Code-Free Experiment Builder
Build studies visually in the task editor, with no programming required.
Single Session Studies
The classic one-sitting format, with or without groups.
Longitudinal Study Design
Follow the same participants across days, weeks, or years.
Questions about your study design or platform fit?