Knowledge Base

Project Analysis

Setup

The Project Analysis section contains the following extra options:

1. Independent Variables - The independent variables Age and Gender are present by default. Choose Project > Independent Variable > Add Independent Variable to add more independent variables, like whether the participants saw the commercial before, or their native language.

   The Independent Variables panel displays variables such as Gender, Age, Experience with commercial, and Nationality, each represented with a checkbox indicator.

   Double-click Independent Variables under a participant name to score them. Age and Gender can be estimated by FaceReader, or entered manually.

2. Stimuli and Event Markers – Define stimuli or event markers (Project menu) to mark episodes of interest. Stimuli have a fixed duration and can be linked to a video or image to show to the test participants.

   The Stimuli panel shows example stimuli such as "Twins eating" and "Frying insects," each assigned a number for ordering.

Stimulus Presentation Tool

Use the stimulus presentation tool to automatically show stimuli to the test participants and synchronize them with the analyses.

Important: In a two computer setup, synchronize the computer clock times with a time server. See Synchronize computers with a network time protocol in the FaceReader Help for how to do so.

On the Computer with FaceReader

1. Choose File > Settings > Data Export and under External Communication (API and Stimulus Presentation Tool) select the Enable External Control checkbox.
2. To add tests, open the Tests tab in the bottom pane of the Project Explorer. Then click on the Add test button.
3. Select the camera and the stimuli to show to the test participants.

   Optionally, choose to let participants enter their own name, age and gender, perform gaze calibration, randomize the presented stimuli, set the stimulus display size and the length of the pause between the stimuli.

On the Test Participant Computer

1. Start the Stimulus Presentation Tool and follow the instructions to connect with the FaceReader computer.

   The Connect to FaceReader dialog requires the following fields:

   • FaceReader ip/host – e.g., localhost
   • FaceReader port – e.g., 9090
   • Temporary storage path – e.g., C:\Users\Public\Documents\Noldus\Stimulus Presentation Tool

   Click Connect to proceed.

2. Select a test and click Start.
3. Fill in the participant details and click Start again, or let the participant do this. The test and analysis start.

   The Welcome screen prompts the participant to enter their name, gender, and age before clicking Start.

Analyze

Participants are automatically added when you use the stimulus presentation tool. Also, analyses are automatically carried out. Watch the Model quality bar when the test runs. The bar should cross the dashed line. If this is not the case, improve lighting or reposition your camera.

The Model quality bar displays a color-coded indicator ranging from red through yellow to green, showing the quality level of the facial detection model during analysis.

Optionally, score event markers during the analysis.

Output

Apart from the general FaceReader output options (page 15), the Project Analysis section has the following output:

• Quick Layouts — This will open a number of charts to give you a first impression of the data.
• Pie chart, Bar chart, Box plot, Line Chart or Circumplex model — Circumplex Model is only available if you scored stimuli in your project.

  Example: A bar chart titled "Twins eating (N = 5)" displays mean emotion scores with baseline correction using "Frying insects" as reference, with Temporal Aggregation: Mean and Participant Aggregation: Mean. Emotions shown include Neutral, Happy, Sad, Angry, Surprised, Scared, Disgusted, and Contempt.

• Heatmap — Only available if you used the Stimulus Presentation Tool and tracked the test participant's gaze. The latter is possible with the Advanced Research Module.
• Table — To calculate statistics on the analyzed facial expressions and other parameters of all participants, participant groups or single participants.

  Example table for "Twins eating (N = 5)" with Baseline Correction: None, Temporal Aggregation: Mean, Participant Aggregation: Mean shows the following mean values across all participants: Neutral 0.3521, Happy 0.5033, Sad 0.0387, Angry 0.0320, Surprised 0.0261, Scared 0.0051, Disgusted 0.0055.

• Stimulus — To add your stimulus videos to the visualization. You can play back your stimulus videos together (synchronized) with a Line chart with the facial expressions and other parameters.

  Example: The "Twins eating" stimulus shows a video of two babies eating at a table, with a synchronized line chart below showing emotion data over time.

• Participant — To add your participant face video to the visualization. You can play back your participant videos together (synchronized) with the stimulus video and a Line chart with the facial expressions and other parameters.

  Example: "Participant 3 - Hans - Analysis 1" shows a face video of the participant synchronized with stimulus markers for "Twins eating" and "Frying insects."

• Compare — To compare two or more data sets in a chart or table and carry out a t-test to test for significant differences.

  Example comparison table displays mean and standard deviation values for Sad, Angry, Surprised, and Scared emotions across two data sets. Highlighted cells indicate statistically significant differences between groups.

You can select what data to include in the analysis:

• Select what parameters to visualize/tabulate.
• Select a stimulus or event marker to visualize/tabulate the data during this stimulus/event marker.
• Select whether you want to do a baseline correction and visualize/tabulate absolute or relative values of your parameters.
• Select how to aggregate data over time.
• Select how to aggregate data over test participants.
• Select which test participants to include in your analysis.

The Bar Chart configuration dialog allows you to:

1. Select which data you want to use — available emotion parameters include: Angry, Contempt, Disgusted, Happy, Neutral, Sad, Scared, and Surprised.
2. (Optional) Select the advanced options — including Temporal Aggregation (e.g., Mean) and Participant Aggregation (e.g., Mean).

Source: FaceReader 10 Quick Start Guide, Noldus Information Technology