Biometric Data

What is an EEG?

Eeg explain

EEG signals on the surface of the scalp are very weak – they are usually measured in microvolts, which are millionths of a volt. Since the signals are small, it is very easy for them to be overwhelmed by sources of interference, such as poor electrode connections, movements, even excessive blinking. These interfering signals are called EEG artifacts. If a participants had too many artifacts, we eliminated their data from the analysis.

EEG Findings

With 50 successful participants, we gathered 250 million points of data. This group of participants consisted of the following:

  • 25 women, 24 men, 1 undisclosed
  • A range from age 13 to 65+
  • A range of household income from $0-20,000 to $150,000+
  • Ethnic identities including British Isles, South Asian, Eastern European, Southern European, French, Southeast Asian, African, Middle Eastern.
Icon 50
Icon hat
Icon wave
Icon movie

7 electrodes per person, 512 measurements per second/electrode, 24 minutes of video

The research staff of the Live Lab focused on analyzing the 5 bands monitored through EEG - delta, theta, alpha, beta, and gamma. Analysis of this data revealed that 2 neighbourhoods in particular generated significantly different reactions from the participants.

Eeg 1

1

Eeg 2

2

Eeg 3

3

Eeg 4

4

Eeg 5

5

Eeg 6

6

Eeg 7

7

Eeg 8

8

These 8 circular images are brain maps of the median response for each neighbourhood video, focused on visualizing the delta power band. Increased activity in the delta band is generally associated with relaxation. The data shows that as participants become more familiar and relaxed with the process of watching the videos, the brain map shifts from dominantly blue (left) to a dominantly yellow/red (far right).

A different way to view the data is shown in the five graphs.

Chart delta
Chart theta
Chart alpha
Chart beta
Chart gamma

Neighbourhoods 3 and 6 appear to deviate from the other neighborhoods for Theta, Delta and Alpha power. Such decreases in these band are associated with a sense of excitement or greater engagement, suggesting that participants found the videos of these neighborhoods more engaging.

Entering this process, Cobalt Connects and the Live Lab team were not certain if the research would yield any interesting results. While we cannot specifically determine exactly what factors caused the increased sense of engagement from this data, they suggest that the videos of different neighbourhoods did elicit unique responses on a physiological level. More in-depth research will now be conducted to further explore this further.

Neighbourhood Video Findings

While our EEG participants were simply watching the neighbouhood videos, the remainder of audience was using tablets to provide equally valuable data. As they watched 2-minute videos of each neighbhourhood, participants were instructed to move a slider on their tablet up or down as they felt the ‘vibrancy' of the scene changed; move it up for increased vibrancy and down for decreased vibrancy. The resulting chart is potentially the most telling data of the entire study.

Chart slider

This graph shows the median response from all participants for each of the 8 neighbourhoods. Each coloured line represents a neighbourhood over the duration of the video from left to right, with the orange line being the average of all neighbourhoods.

We have intentionally left off the names of the neighbourhoods as our interest isn't to determine which is the best or worst, but in understanding the factors which influence a participant's rating.

As you can see in each neighbourhood there are peaks and valleys where the audience rated something more or less vibrant. You can also see that there are neighbourhoods that remain above or below the average for the duration of the study, and others that dance from one side of the line to the other.

Over the past few months Cobalt staff have been evaluating the content of each video, connecting the participant ratings to the elements visible in the video for each neighbourhood. Essentially linking the factors revealed in the Taking Stock and Active Exploration phases with the Live Lab data.

Based on this analysis Cobalt has determined key factors that push a neighbourhood's score up or down by varying degrees, and which elements can potentially override the others. Similar to the EEG, this data is rather exploratory and further research is needed to truly zone in on these relationships. Cobalt is excited to move this research forward over the coming years in partnership with the Live Lab and McMaster University.

Below are a few examples of the relationships we've discovered.

UP FACTORS

Icon up
  • Increased pedestrian activity are associated with an increased rating from 10-40% depending on the pace, volume and variety of modes (i.e. walking, cycling, small groups, individuals)
  • The style of architecture are associated with an increased rating from 10-50%, with differing ratings for set-backs, period or heritage, amount of window, etc.
  • A range of natural elements are also associated with an increased rating to varying degrees

DOWN FACTORS

Icon up
  • Lack of natural elements, contributing to lack of colour, are associated with a decreased rating from 10-30%
  • Vacant or inactive storefronts are associated with a decreased rating from 20-30%, which is compounded as they become consecutive properties
  • Larger intersections and the associated increased traffic sound are associated with a decreased rating up to 31-40%

Over the coming months Cobalt will be delving into these relationships in greater detail and sharing our findings via our findings page and blog.