Skip to main content
Search form
  • 00:24

    My work is trying to understand how people experience the city.So I trained as an architect back in Athens, Greecea few years ago.And as an architect, you learn to designspaces of all kinds-- buildings, urban design, urban spaces.Progressively, through my studies and thenmy postgraduate studies, I was much more

  • 00:45

    interested to understand how people perceive, let's say,our designs.What's the perception, what's the feeling space is creating?So my research is trying to understandhow a pedestrian or a set of pedestrians experience the cityand how the environment affects them psychologically,cognitively, and in other ways.The one aspect of it is trying to understand

  • 01:07

    how pedestrians walk in the city and whatis their psychological and cognitive state behind them,what are their motives, what are the criteria they apply.This is very much tied with CASA's researchinto transport models, understandinghow transport works and how we can betterpredict where people go and how they navigate in the city.

  • 01:31

    So that's one aspect of it.One other aspect we were as much interestedis how we can use this sort of datato discuss how the city accommodatespeople, better or worse.We were doing some work with visually impaired peoplein collaboration with Guide Dogs for the Blind.And if you [INAUDIBLE] we're tryingto look, first of all, if we can tap

  • 01:52

    into the differences of the urban experience of someonewho is visually impaired and someonewho's not visually impaired.They were walking in a different way,you can see the environment, and youcan perceive the environment in different ways.So that's one level, is to ask is there a difference.The second level is, if there is a difference,can we go then with a map of where stress occurs

  • 02:16

    or where stress is mitigated and discuss itwith an urban planner, a team of urban planners,or the local authority and say, is theresomething we can do to improve this area,we know it's stressful for peopleand we can understand why.For example, it's a non-controlled junctionat a very busy part of the city.Can we do something to improve this?

  • 02:36

    So, it's both trying to understandbut also trying to advocate and act and provide datato people who need it, like planners and designers.The way we do this is by using various sets of methodsfrom different disciplines.So we use GPS tracking, for example,to track the movement of an individual

  • 02:59

    as they walk through the city.And we use techniques from neuroscience and psychology,like mobile brain imaging.We're using a technique called electroencephalogram, or EEG,in combination with psychophysiologicalmeasurements, like looking at heart rate and skinconductance, which are all indicators of the arousaland the emotional state of someone.

  • 03:21

    So this is an EEG cap.So basically, it has a number of cables, the red cables,you can see.These electrodes have a rim of copper,and it's a copper wire that goes back to a little center.This is the receiver and amplifier.

  • 03:41

    So this measures the electrical activity.Underneath each electrode amplifiesbecause the current is very faint,and then transmitted wirelessly to the tablet or the computerthat is recording.So this is why it's mobile.By combining their location and howthese indicators of how they might be feeling,the arousal, the excitement they are experiencing,

  • 04:03

    we're trying to understand how the part of the environmentplays in their psychological state.And then through the tasks we give them,we try to understand how the psychological influenceof the environment affects their decision-making processes,their behavior.The benefit, I think, is a question of validationand a question of understanding.

  • 04:27

    So we can use each method to validate the other.So we can say, for example, that people's self-reported data--what they remember or what they madeof an experience or a situation or a place-- is in lineor is not in line with the continuous dataof their emotional arousal, excitement, attention,

  • 04:49

    and so on.So this is one step.The other step is to say, OK, if they're different,why could they be different, why could itbe that we measured a high level of arousalbut the person didn't really recall the situationor didn't say it was as arousing as something else,and then try to understand what makes it different.

  • 05:10

    Because it's not purely a scientific question.In a way, we're not just trying to find if people can dosomething or cannot, we're trying to understand howthe environment elicits different experiences.So any kind of state that we can bring to the table is useful.

  • 05:31

    We're ready to go from this site.So each color is a different electrode.And what we see is the electrical activitymeasured at each location.There are a few steps of post-processingto clean up the noise.There are frequencies that are external like from the mainsor the sources of noise from your guide and movement.

  • 05:54

    For example, we have a signal from the muscles,so this is when you turn your head, when you blink,all of this appears in the signal.It's all electrical activity.When we're going into the field, whatwe do is we would give the route in some formto the participant, we tell them theyhave to walk from A to B and C and so on.

  • 06:16

    We try to make it as simple as we can for them,making sure they all go through the same pathor a controlled path, so we have to makesure we get the best data.And before they go, we have to give them a GPS devicebecause we want to know their location as best as possible.We give them an EEG headset, and that helps measure their brain

  • 06:37

    state.Then, depending on the experiment,you would have a wristband, again, to measurethe psychophysiological state.And then we launch them, we sort of say,you can now go to the location we suggested.And then we follow them behind, tryingto make sure they're safe, first of all,

  • 06:58

    and also trying to record any observable data thatcould be meaningful for us, when they stumble on something, whenthey are crossing a junction and things.We don't want to tell them beforehand,but we want to have that information later.So we use wireless devices to collect that informationin a synchronized way so to have correct timestamps that

  • 07:21

    would allow us to do some post-processing later.All right, so now we're ready to go.I will be walking behind you, like three to ten meters.Don't pay attention to me, as if you're walking alonein the day in the city.So these questions existed for many, many yearsin research, for example, the restorative qualities

  • 07:43

    of natural environments versus urban environmentshas been in discussion and debate and researchfor at least 30 to 40 years.What we could not do then is to actually testthese things outside of the lab, in the field.Now, being in the field introducesa level of uncertainty, a level of non-controllness.

  • 08:06

    We cannot control how much traffic there will be,you cannot control whether there will be an ambulance witha siren, or a cat or a dog.We don't have that level of control as you do in the lab.But at the same time, we can observehow behavior is actually enacted in the real world.And using the research that has been

  • 08:28

    done over the previous years in the lab, with classic settings,with a screen and projecting, for example, imagesand wire stimuli and recording the response of individualsto the stimuli, we can try and associate previous findingsto findings in the field to reduce the uncertaintyand understand what are the sources of the noise

  • 08:49

    we observed.What has changed is that there are nowaffordable devices like our phones have a very good GPSantenna, very good, sufficiently good for what we want.We have mobile EEG that didn't exist five years ago, almost.We have a wristband that can measure your skin conductanceand heart rate very easily and very accurately

  • 09:11

    and give us timestamps to coordinate all of these datatogether at a later stage.So all these things didn't exist ten years ago,or they were very bulky, very heavy, very experimental.Another critical part apart from the participant's safetyis what I know what's happening to her.So if she stumbles, I could take a record with a timestamp.

  • 09:32

    So I can then go in the date and say,this is when she was very stressed,but it's because she fell down or she bumped into a friendor she got lost.So depending on what happens to the participant,we can always keep track of thingswe don't want to tell them to record because it wouldinterfere with their walking around,

  • 09:54

    but we need to know it later.So the main question when we are tryingto understand someone's experience of a situation,like this interview, for example,is what will you remember 30 minutes laterand what will you remember an hour later.And there are various ways of tryingto probe your memory about this situation, this event.

  • 10:15

    One way is asking directly, what doyou remember, how much detail do you remember,what moments can you recollect.And then what we will see most probablyis the moments that stand out, what you remember,what grabbed your attention.What it's harder to understand ishow this drifted and changed from moment to moment, which

  • 10:37

    were the critical points when perhaps your attention wasdrifting away, when people's attention driftsfrom the environment to their own thoughtsand then again back to the environmentbecause they have to cross a junction.These are the moments we are trying to tap into.So to tap into these aspects of the experience,

  • 10:58

    we're using continuous measurement, arousal,heart rate, brain activity that can alsobe used as an indicator of attentionand the balance of the emotion as well.So by tying these together in a continuous stream of data,we can look when there is a peak, when there is a drop,

  • 11:20

    how often do they care, do they care as often between people,is it the same in different groups of people.So we have this information.At the same time, after the experiment,we ask them, what do you remember.What's your evaluation of the situation, which place doyou think was more stressful?And then we compare the two and we're

  • 11:41

    interested-- will there be a discrepancy between the twokinds of measurement.Is it perhaps that we're tapping into a different cognitiveprocess?Or you have a high level evaluation of the situationbut then you may not experience the situation as intensely,for example.So by having the two, we can both compare but also

  • 12:03

    try to understand if it's a different experience we'relooking at, different aspects of the experience,then maybe, both sources of data are as interesting and usefulfor our analysis and whatever we want to do with the data later.So, after the experiment, we haveto gather all the data from the different applicationsand different sources, different devices.

  • 12:25

    We bring them all together.And then we have to combine them in various ways,and we use the timestamps for thisso we can synchronize across devices.And then we can start looking, let's say, at wheredifferent signals occurred.

  • 12:45

    What was the problems of signals in a particular location?Or we can even analyze behaviors.So, for example, if you have 10 people, 30 people,walking in a particular route, youcan look when did the researcher whowas following behind observe, let's say,more looking around or pauses or feeling lost.

  • 13:05

    Depending on what the experiment is looking at in particular,we can start visualizing that and tryto see if there are any hot spots wherewe have a lot of many occurrences of the same eventat the same location, which would suggest there's somethinghappening there which either has to dowith the experiment or an external factor thatneeds to be considered.

  • 13:27

    To do that, we use a combination of different softwares.So, we can use MATLAB, which is standardfor analyzing EEG signals.Once we have an output from MATLABor we use other outputs from the devices,we-- at least I am using R, which is an open source

  • 13:51

    statistical and geospatial analysispackage to combine data, visualize the data,manage in general.One way of making others understandis visualizing the data we have.And one of the benefits of this kind of method

  • 14:12

    is that because "we follow," in quotes, people as theywalk in the city.We can visualize their trace, and wecan project in the trace other dimensionsof the experience, their emotional state, their heartrate.And this is something that's very easy to communicate.

  • 14:32

    So we can produce maps and say, this is a map of one person,this is a map of 10 people, this isa map of the differences between two groups, for example.And we can see, for example, hot spots, or the differences.We can use colors and everything thatis used in photography and visualizationto display someone's experience or the difference

  • 14:54

    between the experience of the two people.That makes it that much more easy to communicate.It is not that it is just more easy to graspif it's a map or a representation of someone'swalk, it's also that it's easier to understand and get slightlyinto someone else's shoes so you can see a diagram like a linechart of how people's emotional state changes

  • 15:14

    from moment to moment.But it's easier to understand the same information projectedon a map and imagining that, OK, they walked and they crosseda junction and you can see the difference,and then they are at a busy street,and then you can see a different line width.Something as simple as a line widthcan communicate something complicated

  • 15:35

    like an emotional state.So again, using mixed methods even at the output is helpful.And then we are also experimentingwhen we can with more new media approaches.You can make a little artwork to exhibit and communicatethe results with non-scientists in a wider community.And it's a question of empathy in a way,

  • 15:58

    so you can see the data and their numbers,you can see visualization.It's more engaging, more easy to understand.Then the more easy to understand and grasp,then the better you communicate what the output is.


Panos Mavros describes how his research measures the perception of urban spaces. He uses electroencephalography and GPS tracking to determine how pedestrians react to different places, as well as how their memory of a place relates to their in-the-moment experience.

Looks like you do not have access to this content.

Mixed Methods: Exploring the Role of Emotions in Urban Behavior

Panos Mavros describes how his research measures the perception of urban spaces. He uses electroencephalography and GPS tracking to determine how pedestrians react to different places, as well as how their memory of a place relates to their in-the-moment experience.

Copy and paste the following HTML into your website