Dataset view (interactive dashboard)

Click “Analysis” in the main menu to access the Dataset viewer. This page in an “interactive dashboard” that presents all aspects of the measurements and results obtained from a subset of the participants while viewing a particular Stim.

Before you continue, ensure you are familiar with the concepts below:

View a Dataset

To view a dataset:

  1. click “Analysis” in the main menu to access the Dataset viewer.
  2. Then, at the top of the page, select the Project and Stim.
  3. Finally, select a Dataset using the name you provided when you created it.

Once a dataset has been selected, it will load. This may take a few moments as the results are highly detailed including many graphs.

You can export each graph in PNG and CSV format by clicking the icons next to the graph name.

Results are presented in a series of collapsible sections each dealing with a different aspect of the data. These sections are in order:

Dashboard Sections

E-Score

The E-Score is an overall summary of the performance of your Stim. It includes Emotion, Attention and Association components. See E-Score for more details. E-Scores are calculated by comparison to benchmark results.

Engagement

Engagement is the overall response to the Stim. It comes in two parts – Relative attention over time (visual engagement) and Emotional engagement over time. These charts are useful for detecting particular moments in a Stim that trigger stronger reactions.

Peak Emotions

This section includes two charts – the most significant emotional shifts and the times of peak expression of each emotion. Note that these are aggregate frequencies within the sampled population. For example, if all participants expressed an emotion at a particular time the frequency would be 1 (100%). If no participants expressed an emotion, the frequency would be 0 (0%). Aggregate measurements reduce the inherent uncertainty in emotion recognition by considering all results. Shifts represent changes in the frequency of responses in the population.

Note that when exploring emotional responses, it is important to remember that the action coding units tend to represent a range of possible emotional reactions. “Negative” emotions are not necessarily bad.

Emotions over Time

Next we look in more detail at how each emotion action unit varied over time. We show both absolute frequency and shifts in expression frequency within the population. Shifts tend to be triggered by Stim content.

Attention

This section explores the eye-tracking results in more detail. We include both a heatmap video showing aggregate attention over time, and statistics for the Areas-Of-Interest (AOIs) you defined earlier.

If Heatmaps weren’t included in the Dataset this section will be collapsed.

Association

This section presents survey (explicit) and implicit association test results. In addition to changes in association before and after stims, we look at rational vs emotional responses and the difference between explicit stated opinion and the equivalent association strength measured.

Meta-Data

The raw data that was used to build the dashboard can be found in this section. Advanced users only.

Composite Video

To download and share highlights of the interactive dashboard with colleagues or clients, you can generate and download a video version of the page.