Knowledge Base

Data Selection

Aim

• Select what parameters to analyze
• Select what stimuli/event markers to analyze
• Select absolute or relative values for your parameters
• Select how to aggregate data over time and over test participants
• Select participants

Prerequisites

• Your license includes the Project Analysis Module.
• Your project contains video or camera analyses.

Procedure

1. Open the Project Analysis section (see Project analysis in short).
2. Select the chart or table of your choice. The Data selection window opens.

Select What Parameters To Analyze

The first section lists the parameters that will be included in your chart or table. Click select to see all the variables that are available and make a selection. The parameters in gray are the ones that are selected.

Notes

• It is not possible to select Valence or Arousal when you make a Pie Chart.
• You cannot select any parameters when you make a Circumplex Model because this chart always shows a combination of Valence and Arousal.

Select What Stimuli/Event Markers To Analyze

The Stimulus or Event Marker section lists the stimuli and event markers that were scored in the project. Click a stimulus or event marker to select it and to visualize/tabulate the data during this stimulus/event marker. If you do not make a selection the whole observation period is taken into account.

Advanced Options For Data Analysis

Aim

• To select whether you want to visualize/tabulate absolute or relative values of your parameters
• To select how to aggregate data over time.
• To select how to aggregate data over test participants.

Prerequisites

• Your license includes the Project Analysis Module.
• Your project contains video or camera analyses.

Procedure

1. Open the Project Analysis section (see Project analysis in short).
2. Select the chart or table of your choice. The Data selection window opens.
3. Click select next to Advanced Options to open the Advanced Options window and click the options you want to use:

Select Absolute Or Relative Values For Your Parameters

Select one of the options under Baseline Correction to visualize/tabulate either the absolute values or relative values of your parameters.

Baseline Correction is not available when you make a Pie Chart. You can only visualize absolute values as relative values can be negative.

• None (default) — With this option selected the absolute intensities of the selected data will be visualized/tabulated. The bar chart in the example below shows the mean absolute intensities of the facial expressions that the test participants showed during Commercial 1.
• All Other Measurements — To visualize/tabulate the selected data during a stimulus relative to all other time fragments. Use this option to calculate in which way the facial expressions, for example, during watching a certain commercial differ from the expressions while the participants were not watching it. The bar chart in the example below shows the mean intensities of the facial expressions of the test participants during Commercial 1 relative to the facial expressions that they showed during the rest of the test. The chart shows that the test participants were happier during Commercial 1 than during the rest of the test.
• Other Stimuli Or Event Markers — The selected data during a stimulus/event marker relative to the expression intensities during another stimulus/event marker. When you select this option, choose what stimuli or event markers to use as the baseline. With this option you can for example correct the facial expressions of test participants when they watched a certain commercial for the expressions while watching a neutral scene. The bar chart in the example below shows the mean intensities of the test participants during Commercial 1 relative to the facial expressions that they showed during a Neutral stimulus. The chart shows that the test participants were happier during Commercial 1 than during the Neutral stimulus.
• Interval Before Stimulus/Event Marker — The selected data during a Stimulus relative to the expressions just before the stimulus or event. With this option you can measure the effect of a stimulus on the participant's mood, by looking at the changes induced by the stimulus. Enter the duration of the interval before the stimulus or event marker in the Interval field. The bar chart in the example below shows the mean intensities of the facial expressions of the test participants during Commercial 1 relative to the facial expressions that they showed during the 5-second interval before Commercial 1. The chart shows that the test participants were happier during Commercial 1 than during the 5-second interval before Commercial 1.

Select How To Aggregate Data Over Time And Over Test Participants

Select one of the options under Temporal Aggregation to select how the values per test participant are aggregated over time. If you choose Activation Period FaceReader counts the percentage of frames which have an activation higher than the threshold. The default threshold is 0.2. That means that the value for 'Happy' is, for instance, 60% if the intensity of 'Happy' is 0.2 or higher in 60% of the frames.

Notes:

• Temporal aggregation is not available when you make a Circumplex Model or Line Chart because these chart types are based on temporal data.
• You can select multiple temporal aggregation methods when you make a table.

Select one of the options under Participant Aggregation to select how the values of the different participants are aggregated.

Note:

Participant Aggregation is not available when you make a Box Plot because a Box Plot by definition shows a number of participant aggregation values (Q1, Q2, Q3). See Box Plot for more information.

How Baseline Correction And Aggregation Is Performed

Suppose you have three participants with one analysis each and you scored the same Stimulus for all three participants. You selected All Other Measurements as Baseline Correction in the Advanced Options and Mean for both Temporal Aggregation and Participant Aggregation. As an example, calculation of the relative mean value of Happy during the Stimulus is done in the following way:

Participant 1 (P1) — (mean intensity Happy P1 within Stimulus interval) - (mean intensity Happy P1 outside Stimulus interval) = A

Participant 2 (P2) — (mean intensity Happy P2 within Stimulus interval) - (mean intensity Happy P2 outside Stimulus interval) = B

Participant 3 (P3) — (mean intensity Happy P3 within Stimulus interval) - (mean intensity Happy P2 outside Stimulus interval) = C

Relative Mean value of Happy = (A + B + C)/ 3

Source: FaceReader 10 Reference Manual (Help), Noldus Information Technology