G.C. Millar Preliminary Oral Examination Presentation

Maps & Minds: Geospatial Analytics for Studying Human Cognition

Garrett C. Millar

Preliminary Oral Examination

November 4th, 2020

Helena Mitasova
Committee Chair
Geospatial Analytics

Ross Meentemeyer
Committee Member
Geospatial Analytics

Laura Tateosian
Committee Member
Geospatial Analytics

Aaron Hipp
Committee Member
Geospatial Analytics

Hi everyone, thank you very much for being here, I really couldnt be more excited about presenting to you all today. With that said, lets go ahead and jump right in
This presentation (as Helena mentioned) serves as a dissertation proposal for a doctorate in the field of Geospatial Analytics
The proposed title: Maps & Minds: Geospatial Analytics for Studying Human Cognition
by me: :D
To start off, I thought it best to bring it all back a bit, and give a brief background on my overall research career.
As you all know, I started my research career in cognitive science (or psychology).
I specifically started out in Human Factors & Applied Cognition (or Human Computer Interaction)
The lab I was involved with researched how computers can help people learn better.
I stuck with this research for 2 years as an undergraduate, and then continued it for 2 years as a graduate student.
NEXT STEP ⇒

Background & Motivation

In those 4 years, I became very well-versed in 3 Cog Sci topics in particular :
1. emotion, NEXT STEP ⇒
2. learning, NEXT STEP ⇒
3. AND human–computer interaction. NEXT STEP ⇒
I ended up finding the field of Geospatial Analytics and from which I ended up becoming a Geospatial Scientist,
As a Geospatial Scientist, I saw extremely exciting opportunities to continue researching these Cognitive Science topics of emotion, learning and Human Computer Interaction,, but approach them from a Geospatial Science Perspective instead
The resulting research is what has formed my dissertation I am proposing today
In this dissertation, I
1. NEXT STEP ⇒ combine my experience with emotion research with geo-statistics and -modeling techniques to explore interactions between the environment and human emotions / behavior ( this formed my CH 1);
2. NEXT STEP ⇒ I then leverage my background in learning research with theories of embodied cognition to facilitate more natural ways to learn about landscapes (which makes up CH 2);
3. NEXT STEP ⇒ and for my last Chapter, Chapter 3, I propose using Human Computer Interaction principles to evaluate and improve geospatial technologies for better decision-making (CH 3)
we'll be discussing each of these in more detail today, so LETS GO AHEAD GET STARTED

Maps & Minds: Geospatial Analytics for Studying Human Cognition

Completed Work

Chapter 1 Space-time Analytics of Human Physiology for Urban Planning

Chapter 2 Tangible Landscape: A Hands-on Method for Teaching Terrain Analysis

Ongoing & Proposed Work

Chapter 3 Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

just real quick as a formal outline of my dissertation:
For my completed work: NEXT STEP ⇒
- there is chapter 1, titled: NEXT STEP ⇒ ,
  - Space-time Analytics of Human Physiology for Urban Planning
- and there's chapter 2 NEXT STEP ⇒ :
  - Tangible Landscape: A Hands-on Method for Teaching Terrain Analysis
and for my ongoing and proposed work, : NEXT STEP ⇒
- we have chapter 3: NEXT STEP ⇒
  - Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

so let's jump into chapter one
NEXT STEP ⇒

Space-time Analytics of Human Physiology for Urban Planning

Chapter 1 Space-time Analytics of Human Physiology for Urban Planning

Background

Research Objective: Use mobile sensing to improve the understanding of how people experience their environment.
How?
- Combine location and wearable emotion data to inform urban design decisions.
- Use visual stimuli derived from viewsheds to provide more accurate environment interaction metrics.
- Develop web-based mapping system to explore complex human physiology data across urban landscapes.

Millar, G. C., Mitas, O., Boode, W., Hoeke, L., de Kruijf, J., Petrasova, A., & Mitasova, H. (2020). Space-time Analytics of Human Physiology for Urban Planning. Computers, Environment and Urban Systems, 85, 101554. https://doi.org/10.1016/j.compenvurbsys.2020.101554

So, this research's main objective was to: NEXT STEP ⇒
Use mobile sensing to improve our understanding of how people experience their environment.
We did this by NEXT STEP ⇒ using location and wearable emotion data collected at a high spatiotemporal resolution to calculate viewsheds of what cyclists saw along a cycle highway.
NEXT STEP ⇒ These viewsheds were then used to link cycling experiences to visual stimuli that were actually present at specific locations along particpants ride.
NEXT STEP ⇒ A dynamic web-based mapping system was also developed to help us and researchers alike better explore complex moment-by-moment human physiology data across urban landscapes.
NEXT STEP ⇒ mention newly published CEUS article!!! -
i also just wanted to quickly mention this research was recently published in the elsevier journal of computers environment and urban systems

Data & Methods

CHIPS: Cycle Highway Project

Study purpose: Adapt existing cycling infrastructure for increasingly
heavy e-bike use.
Participants: 12
Cycle track: 18 km cycle highway
Type of bike: e-bike

My research on this topic stems from a much larger project, the CHIPS project which stands for "Cycle Highways Innovation for smarter People Transport and Spatial Planning" spanning Belgium, Germany, the Netherlands and the UK.
this project involves NEXT STEP ⇒ adapting cycling infrastructure to better account for increasingly heavy e-bike use.
think how much faster an e-cyclist is in comparison to your standard manual bike.
the signs (directions) are too small,
AND cycle paths cant properly account for the presence of both regular and e-bike use (e-bikes passing by fast)
To begin exploring how cycle highways can be adapted for increasing e-bike use, researchers at Breda University of Applied Sciences (in NL_ ran a small human-subject study
in this study : they collected data on NEXT STEP ⇒ 12 participants, NEXT STEP ⇒ who were asked to cycle down an 18km stretch of cycling highway between Tilburg and Waalwijk in NL NEXT STEP ⇒ and these participants were using an e-bike.

Data & Methods

Human Movement & Physiology

Data & Methods

Environmental Location

Data & Methods

Dynamic Visualization

Geospatial Application E-Motion - Uses cyclists' physiological data to see how emotions can be affected by our environment.

Our first attempt at combining human and environmental data resulted in this dynamic viz tool being shown right now
This application allows you to dive into the detail of spatiotemporal physiological data
So, just to quickly demonstrate . . .
- On the left, we can toggle back and forth between different participants's collected data,
- We can coordinate the data seen on the map, with the skin conductance chart in the top-right corner.
- by clicking on one of the points, we'll receive additional GPS location, but also recieve an automatically displayed street view image (1st person)
  - With this, it is easier to understand the specific environmental features someone likely interacted with at that particilar location
Overall, by allowing us to explore the location and emotion data on a moment-by-moment basis, this tool helped inform where we should direct further investigative efforts
For instance, it helped us realize that the cyclists line of sight needed to be considered in our statistical analyses

Analysis

Environment Interaction Metrics

Speed
Distance to Roads
Direction
Direction Change (Turning)
Viewable Land Cover Area:
- Developed
- Natural
- Recreation
- Water
- Business
- Agriculture
- Forest

Analysis

Viewsheds

Results

Descriptive Statistics	Multi-level Statistical Modeling
	☝︎ — More emotional in large, viewable natural areas. ☟ — Less emotional in large, viewable developed areas.

DESCRIPTIVE STATISTICS :
- For our results NEXT STEP ⇒ we first wanted to provide summary statistics:
  - To do so, we used a NEXT STEP ⇒ hexagonal binning process to summarize and graphically represent summary statistics of the measured data
  - in the graphic shown now, figure A is showing: Phasic means; and figure B shows: Phasic Standard deviations
  - These spatially distributed summary metrics show prevailing lower values of phasic skin conductance at the start / end of route in more urban areas, while higher phasic skin conductance values are seen midway along the route in more rural settings
MIXED-EFFECT LINER MODELS:
- Then, to empirically examine and determine environmental influences on cyclists' emotional experience, we used a MIXED-EFFECT LINEAR MODEL
- Our multi-level model demonstrated two main findings:
  - First, NEXT STEP ⇒ participants became more emotional in large viewable areas managed for natural resource uses (these were less-developed areas, containing more, Recreational, Agricultural, & Forest Land Cover in View. )
  - Oppositely, particpants NEXT STEP ⇒ became relatively less emotional in large viewable human-built areas (extents of developed and business areas in view).
  - we also found that the extent participants were turning was negatively related to their emotional excitement
  - In other words, the more participants were turning, the less excited they were.
Lastly, I just wanted to quickly note the most IMPORTANT ENVIRO INTERACTION METRICS / PREDICTORS used in our model:
- Developed in View −6.229⁎⁎⁎
  Recreation in View 3.921⁎⁎⁎
  Agriculture in View 2.531⁎
  Forest in View 20.361⁎⁎⁎
  Speed −47.916⁎⁎⁎
  Distance to Road 21.767⁎⁎⁎
  Direction Change −28.494⁎⁎⁎

Importance & Contributions

Presents novel approach for linking biosensing with location data to examine how the visible environment affects peoples' experiences.
Presents a first-of-its-kind dynamic visualization tool.
Reproducibility and applicability of methods and visualization tool.

So, the main purpose of this research was NEXT STEP ⇒ to study relationships between spatial-emotional data and features/stimuli that people actually saw in their surrounding environment.
Similar research has been attempted in the past, however, the approaches these studies took were too general ...
they only correlated people's physiologcial data with environmental features that POTENTIALLY resided in the participants' general vicinity
Whereas our methods instead take sub-moment, sub-meter spatiotemporal physiologcial data , and empirically link it to what people actually saw in their environment
Second, and in my opinion, the most important and BIGGEST CONTRIBUTION of this work is NEXT STEP ⇒ the Dynamic Viz Tool:
- Nobody else has a visualization tool that allows researchers to look at physiologcial data on a second by second basis, and tie it to geographical, surrounding environmental information.
And the last contribution to note concerns the NEXT STEP ⇒ reproducibililty and applicabilty of the developed methods
In fact, we initially developed these methods and dynamic viz tool for two separate projects; both projects having collected location and emotion data (cycling & museum ➡︎ both had emotion / location data).
Because of this, our methodlogical framework and mapping tool are not only highly reproducibile, but can be used by a wide variety of researchers,
As such, we assert these methods are valuable for any researcher wanting to map people's emotions and experiences.

Chapter 1 Space-time Analytics of Human Physiology for Urban Planning

Publications

Millar, G. C., Mitas O., Boode W., Hoeke L., de Kruijff J., Mitasova H. (2020). Space-time Analytics of Human Physiology for Urban Planning. Computers, Environment and Urban Systems, 85, 101554.
https://doi.org/10.1016/j.compenvurbsys.2020.101554

Mitas, O., Mitasova, H., Millar, G. C., Boode, W., Neveu, V., Hover, M., van den Eijnden, F. and Bastiaansen, M., (2020). More is Not Better: The Emotional Dynamics of an Excellent Experience. Journal of Hospitality & Tourism Research, 1096348020957075.

Software & Applications

Geospatial Application E-Motion - Uses cyclists' physiological data to see how stress and emotions can be affected by our environment.
Geospatial Application Emotion Musuem - Uses a 3D model of a museum and geotagged emotion data to explore which exhibits in the Vincent Van Gogh Centre visitors find the most exciting.

Tangible landscape: A hands-on method for teaching terrain analysis

Chapter 2 Tangible Landscape: A Hands-on Method for Teaching Terrain Analysis

Background

Research Objective: Test the effectiveness of a hands-on method for teaching spatial concepts using Tangible Landscape.
How?
- Test students’ acquisition & transfer of knowledge.
- Examine students’ ratings of the system’s usability & user experience.

Millar, G. C., Tabrizian, P., Petrasova, A., Petras, V., Harmon, B., Mitasova, H., & Meetenmeyer, R. K. (2018). Tangible landscape: A hands-on method for teaching terrain analysis. In Proceedings of the 2018 chi conference on human factors incomputing systems (pp. 380:1–380:12)., New York, NY, USA: ACM.
[Winner of the Honorable Mention for Best Paper Award]. https://doi.org/10.1145/3173574.3173954.

Methods

Interaction, Feedback, & Example Solutions

Methods

Evaluation Materials

Topographic Map Assessment (TMA) Assessed students’ acquisition & transfer of spatial skills.
Tangible Lesson Assessments: Landforms Measured student’s knowledge specific to content in the landforms lesson.
Tangible Lesson Assessments: Cut & Fill Measured student’s knowledge specific to content in the cut and fill lesson.

the Topographic Map Assessment NEXT STEP ⇒ ,
2 Tangible Lesson Assessments:
- one for NEXT STEP ⇒ the landforms lesson,
and the other for NEXT STEP ⇒ Cut & Fill lesson
a User Experience Survey was also adminstered.

And I will describe each in detail....
So, we administered the Topographical Map Assessment (developed by Newcombe & colleagues) to measure students' overall acquisition & transfer of spatial skills, given to participants in the 1st session & 2 weeks after the last session
- And what I mean by "Spatial Skills" here is that students are able to understand how elevation is encoded on topographic maps & how 3D terrain shape is represented on maps)
2 Tangible Lesson assessments were also adminstered before and after each tangible lesson (minus Water Flow; TMA had hydrology questions..), which measured student’s spatial knowledge specific to the content being taught in each tangible lesson.
- So, the landforms assessment tested them on geomorphology concepts introduced to them with Tangible Landscape, and the cut & fill assessment was structured the same way, just for eath moving / grading concepts instead.

Results & Importance

Tangible Landscape supports both improved user experience as well as marginal, task-specific knowledge building.
Several implications for the design and implementation of tangible teaching methods for learning about Landscape Architecture (and other topics).
Acknowledgements
- Software design & development: Drs. Anna Petrasova, Vaclav Petras, Payam Tabrizian, and Brendan Harmon
- Study implementation: Carla Delcambre of the Landscape Architecture department and her Grading and Drainage course at North Carolina State University
- Our findings showed overall, that NEXT STEP ⇒ participants performed significantly better on the Cut & Fill (earth moving) assessment after having completed the analogous task with Tangible Landscape.
- And this can potentially be explained by the ability to directly feel, grasp, and manipulate the various tangible materials as well as the projected visual feedback which helped them better understand the effects of their changes.
- Also, results from the User Experience Survey highlight how Tangible Landscape allows users to try, see & feel, and directly experience multiple variations of a given solution. NEXT STEP ⇒
- ** Specifically, action is reversible with Tangible Landscape, and this encourages users to explore without risk of consequence.

Chapter 2 Tangible landscape: A hands-on method for teaching terrain analysis

Publications

Millar, G. C. Tabrizian, P., Petrasova, A., Petras, V., Harmon, B., Mitasova, H., & Meetenmeyer, R. K. (2018). Tangible landscape: A hands-on method for teaching terrain analysis. In Proceedings of the 2018 chi conference on human factors incomputing systems (pp. 380:1–380:12)., New York, NY, USA: ACM. [Winner of the Honorable Mention for Best Paper Award]. https://doi.org/10.1145/3173574.3173954. Millar, G. C. Tabrizian, P., Petrasova, A., Petras, V., Harmon, B., Mitasova, H., & Meetenmeyer, R. K. (2018). Increasing Underrepresented High School Students' STEM Career Awareness and Interest: An Informal Geospatial Science Program. American Geophysical Union, Fall Meeting 2018, December 12, 2018. ED12A-05.

Related Work

Millar, G. C. Tabrizian, P., Petrasova, A., Petras, V., Harmon, B., Mitasova, H., & Meetenmeyer, R. K. (2018). Hands-on Methods for Teaching Landscape Form and Processes. In US-IALE Spring Meeting Abstracts. Petrasova, A., Tabrizian, P., Harmon, B. A., Petras, V. Millar, G. C. Mitasova, H., Meentemeyer, R. K. (2017). Learning topography with Tangible Landscape games. In AGU Fall Meeting Abstracts.

Software & Applications

Geospatial Application Tangible Landscape Activities (GAPS) - These activities lead students through an introduction to geospatial computation simulation and modeling using a geosptial tangible interface-Tangible Landscape. Geospatial Application Wake STEM TL: Exploring Forest Fires - Introduces students to computational science and applications of its concepts for real-world spatial phenomenon.

Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

Chapter 3 Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

Background

Research Objective: Investigate and discover how web-based SDSS can better support collaborative ecological decision-making.
How?
- Conduct HCI experiment to investigate information acquisition patterns of decision-making groups.
- Collect user metrics to investigate and standardize constructs of system suitability for web collaboration contexts.

is to discover and investigate how web-based SDSS can better support collaborative ecological decision making.
Basically seeing how we can facilitate online / remote collaboration as people are trying to understand and manage the spread of plant disease.
And so to do this I'm proposing a NEXT STEP ⇒ Human Computer Interaction experiment that will specifically investigate the information acquisition patterns of decision making groups as they try to solve specific disease management problems together.
And while they do that, I'm proposing NEXT STEP ⇒ to collect user metrics similar to what was shown in TL study (UX survey, etc) to investigate and standardize constructs of system suitability for web collaboration contexts in the problem space of managing plant disease spread virtually.

Methods

Experimental Design

This study will adopt a 2x2 within-subjects experimental design, in which every participant will experience every possible problem space; aka no one will only work by themselves, everyone will experience both collaboration conditions, and attempt to solve both disease scenarios.
task complexity (disease scenario) and collaboration mode (split between people working individually and working in groups) - are designated as the IVs, and for DVs, Human Computer Interaction metrics (user metrics) will be continuously collected as participants complete the tasks:
- HCI / User metrics: how long it takes them to think of a solution, how many times they had to try a solution until it was correct.
This is will be done in efforts to flesh out possible effects collaboration can have on people's ability to understand and properly manage spreading various cases of biological invasions (differing disease characteristics), how to manage them, and how their decisions and actions can influence can influence their predicted spread.

Methods

Setting & Apparatus

Pest or Pathogen Spread Web Platform (PoPS)

In the proposed research, users must collaboratively use PoPS which is the Pest or Pathogen Spread Web Platform, an interactive online geospatial user interface system (also can be known as an SDSS), to effectively reduce predicted pest and pathogen spread while minimizing treatment costs during a management scenario
PoPS is uniquely suited for examining the potential impact of management strategies on spread, allowing users to compare a forecast without management to forecasts based on “what-if” scenarios that treat or remove hosts. (SIMILAR TO TL'S EXPLORATION])
and this is actually very similar to what i mentioned about tangible landscape; how it allows users to explore without the risk of consequence; the similar concept can be thought of here

Methods

Disease Scenarios

Methods

Disease Scenarios

Task difficulty defined by three general disease characteristics:
1. Dispersal ability
2. Detection ability
3. Landscape heterogeneity

This task difficulty is defined on three general disease characteristics:
1. NEXT STEP ⇒ dispersal ability
2. NEXT STEP ⇒ detection ability
3. NEXT STEP ⇒ landscape heterogeneity
Each of these characteristics are to be split into two categories, representative of low and high difficulty.
For instance, low difficulty will involve a fictitious pathogen (NOT a real-world pathogen) capable of short-range dispersal (dispersal ability, is not asymptomatic and thus easy todetect (detection ability), and operate within and across a homogeneous landscape (landscape heterogeneity).
Oppositely, the high difficulty disease scenario will entail a fictitious pathogen capable of long-range dispersal (dispersal ability), is asymptomatic and thus difficult to detect (detection ability), and operate within and across a heterogeneous landscape (landscape heterogeneity)

Data & Methods

Evaluation

System Usabilty Scale	Semi-structured Interview

once participants complete these scenarios will then be asked to fill out a user experience survey NEXT STEP ⇒ very similar to the one that was used in the tangible landscape study
this is specifically an adapted version of the system usability scale (or SUS)
and it will be used to evaluate overall usability and user experience of the online platform pops (or online spatial decision support system)
The primary constructs the user experience survey assesses are effectiveness, efficiency, and satisfaction.
These constructs are general classes of usability that really should be incorporated in any attempted measures of usability.
After having completed the scenarios, NEXT STEP ⇒ post-study discussion groups will be held after completing the scenarios
These discussions will follow a semi-structured interview format to offer a more natural and relaxed environment for participants to explain or state what issues they might have run into or what they really liked about this system
and in my opinion this approach really allows for more natural and more open and honest answers instead of having to fill out a more rigid questionnaire

Data & Methods

Proposed Analysis

Prediction: Participants in collaboration group will engage in more effective and accurate decision making.
- Analyses:
  - Paired Samples T-test examining differences between collaboration and individual groups on overall disease management task performance.
Prediction: Higher usability and user experience ratings with higher disease management task performance.
- Analyses:

so i'll conduct statistical analyses to investigate overall constructs of the participants ability to solve these problems
this research predicts that NEXT STEP ⇒ in the collaborative setting, participants will engage in more effective decision making since they can discuss with their team members about what the best possible solution might be and kind of help each other resolve any possible uncertainty in the proposed solutions
to carry this analyses out i'll NEXT STEP ⇒ just run a really straightforward paired samples t-test comparing the differences between the collaboration group and the individual group on their overall performance on the disease management scenarios and tasks
- NOTE IF NEEDED: PAIRED SAMPLES T-TEST MEASURES SAME PARTICIPANTS AT MULTIPLE TIME POINTS; ALLOWING PT'S TO ACT AS THEIR OWN CONTROL
then for usability and user experience the only thing that i'm really able to predict with this is to say that NEXT STEP ⇒ usability and user experience will be rated higher given higher performance on these disease management scenarios
my predictions for this are limited because there really isn't a lot of literature out on if people perceive a system better or a like a system more if they're allowed to work with others on it
and this is what i'm really actually very interested in seeing in the results; seeing how their ratings of the system's usability and effectiveness to manage the spread of plant disease changes between working with others or working alone.
i will start this analysis NEXT STEP ⇒ with a pearson product moment correlation
if this correlation comes back significant i will then use a regression which will use collaboration type as a predictor variable and then ratings of usability and user experience as the outcome.

Importance & Contributions

Identification of user experience and usability constructs of SDSS use within the overarching problem space of environmental problem solving.
First to use HCI principles to improve SDSS.
General usability framework that can be applied and used for multiple use cases.

So this study's overall purpose and importance is really to NEXT STEP ⇒ identify any potential issues in spatial decision support system design specifically for environmental problem solving
this will be done in the hopes that we can help policy makers and land managers access and understand spatial data to assist their overall management of of plant disease
i do think it's very important to note here that over the years a lot of spatial decision support systems have been developed, but they really always seem to be designed and developed for very specific and individual use cases
There isn't anything that gets close to an overall general framework or set of usability standards that can be applied across multiple use cases for multiple user groups.
which, more or less, is what i'm shooting to address with this research: that is, to be NEXT STEP ⇒ one of the first to empirically use HCI principles to improve SDSS.
But also, to NEXT STEP ⇒ hopefully identify user experience and usability constructs of SDSS use within the overarching problem space of environmental problem solving

Chapter 3 Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

Related Work

Gaydos, D., Jones, C., Jones, S., Millar, G. C., Petras, V., Petrasova, A., Mitasova H., Meentemeyer, R. (under review). Evaluating Online and Tangible Interfaces for Engaging Stakeholders in Forecasting and Control of Biological Invasions. Ecological Applications.

Related Software & Applications

Geospatial Application Pandemic Dashboard - Flexible and dynamic web-based geospatial analytic visualization tool using globally available data, which can be quickly deployed for any pest or pathogen species, and potentially multiple species at once. This forecasting framework and visualization system and its user interface can help phytosanitary agencies anticipate what invasive species are entering the US, helping us move towards anticipating new problems before they arise, as opposed to a reactionary approach.

Timeline

Fall 2020	Spring 2021
Conduct APHIS user study Finish by December - January Begin drafting Chapter 3 publication	Finalize Chapter 3 publication Submit written dissertation Dissertation defense presentation

Presentation Link: gcmillar.github.io/presentations/prelim_presentation

An Actually-Fun-to-Look-at Online CV: gcmillar.github.io

Extra Slides

Other work Increasing Underrepresented High School Students’ STEM Career Awareness and Interest: An Informal Geospatial Science Program

Research Objective: Develop activities with Tangible Landscape and other related curricula lessons to increase underrepresented high school students' spatial thinking and interest in GIS (STEM)
Lead evaluative procedures for:
- Improving student competence in science
- Nurturing student enthusiasm for science
- Interesting students in research or other science-related careers

I also wanted to briefly go over some of my work that won't necessarily lead to any dissertation chapters, but was the very first project I was on as a student here at the center: GAPS - Geospatial Applications for Problem Solving
And is another way I've explored how we can learn about the world. This time more focused on how high school students can learn about it, and more importantly be excited about doing so!!
So, GAPS is a free Saturday program for high school students interested in using the latest cutting-edge technologies in the geospatial sciences to solve real-world problems, and in learning about all the exciting benefits of a career in the geospatial sciences.
And my main responsibilities included:
1. developing and running activities with Tangible Landscape and other related curricula lessons to increase underrepresented high school students' spatial thinking and interest in GIS (STEM) and...
2. Lead the evaluative efforts on the program as a whole, basically to see whether or not the program was an effective way to get students excited about geospatial science and science in general
3. I did this by focusing on the three overall goals you see on the slide, (which will revisit here in a moment):
  1. Improving student competence in science
  2. Nurturing student enthusiasm for science
  3. Interesting students in research or other science-related careers

Developed Activities

Water Flow		Trail Planning
Landforms		Channeling
Cut & Fill		Ponding

See Developed TL Activity Website: gcmillar.github.io/Tangible-Landscape-Activities-GAPS/index.html

I just wanted to quickly let you get an idea of what these TL activities really consisted of:
1. Water Flow: Students are asked to assume the role of a landscape contractor as they try to find appropriate villa construction sites across a mountainous territory.
2. Trail Planning: Students are assigned to plan a low-intensity hiking route in the Smoky Mountains meant for seniors and children.
3. Landforms: Students use given sand to create the required landforms and label them with the colored markers.
4. Channeling: Students are tasked with reshaping the topography with the overall goal of diverting all the simulated water from the “construction site” to the “stormwater drain” located downhill.
5. Cut & Fill: students must modify the given landscape using cut and fill projection (basic and advanced)
6. Ponding: Students are given a limited amount of sand to build a damn on a stream to impound the maximum volume of water possible.

Results & Contributions

Goal 1: Improving student competence in science

Average 3.39 (out of 4) on being able to explain background material during project presentations

Goal 2: Nurturing student enthusiasm for science

56% of participants indicated an increased interest in learning science

Goal 3: Interesting students in research or other science-related careers

80% of parents stated their child gained skills to use in a STEM career

Acknowledgements

Program facilitation: Drs. Eric Money, Kyle Bunds, Helena Mitasova

Reference

Millar, G. C., Money, E.S., Bunds, K.S., Mitasova, H., & Meetenmeyer, R. K. (2018). Increasing Underrepresented High School Students' STEM Career Awareness and Interest: An Informal Geospatial Science Program American Geophysical Union, Fall Meeting 2018, December 12, 2018.

User Experience Survey

(Ras et al., 2012)

Examined how students perceived and interacted with Tangible Landscape, & how they collaborated to solve a problem
Constructs:

Performance expectancy
Pragmatic quality:

physical objects (wooden carving tools, physical landscape model)
visual objects (projection, digital feedback)

Effort expectancy
User experience

Results

Knowledge Building: Tangible Lessons

Individual Scores

Mean Scores

Here we see the results from comparing students' pre- and post-test scores on the Landforms & Cut-Fill Lesson Assessments.
On the left, is the distribution of participants' individual scores (out of 100%), and on the right is the overall mean scores.
After running paired t-tests for each lesson analysis, no significant differences for the Landforms Assessment were shown,
HOWEVER, the analysis did reveal significant differences for the Cut & Fill Assessment, which means participants scored significantly higher on the Cut & Fill assessment after completing the Cut & Fill tangible lesson.

Cut & Fill: (t(2.73), p = .016).
Pre (M = 53.25, SD = 19.27))
Post (M = 59.97, SD = 16.14))

Results

Knowledge Building: TMA

Individual Scores

Mean Scores

Results

User Experience

All items rated above the neutral value of 4 (out of 7)
Most advantageous aspects of Tangible Landscape?

ability to explore various solutions for the given problems (e.g., water flow, landforms, cut and fill)
physical objects allowed students to change parameters (e.g., location of solution points) very quickly
projected visual feedback helped them better understand the effects of changing those parameters

	Tangible Landscape: A Hands-on Method for Teaching Terrain Analysis. 2018 ACM Conference on Human Factors in Computing Systems (CHI), Montreal, Canada, April 2018			Hands-on Methods for Teaching Landscape Form and Processes. US Regional Association of the International Association for Landscape Ecology (USIALE), Chicago, Il, April 8-12 2018
	Increasing Underrepresented High School Students' STEM Career Awareness and Interest: An Informal Geospatial Science Program American Geophysical Union, Fall Meeting 2018, December 12 2018			Mapping the Emotional Dimension: Measuring Human Behavior Across Space & Time to Inform Tourism & Leisure Management. Visiting Scholar Residency at Experience Measurement Lab; Breda University of Applied Sciences, Breda, Netherlands.
	People & the Environment: Geo-analytical Guidelines & Software Tools for Measuring Interaction & Experience with the Built Environment. Visiting Scholar Talk at Harvard School of Public Health; , Harvard University, September 2019, Boston, MA			People & the Environment: Geo-analytical Guidelines for Measuring Environmental Interaction. Visiting Scholar Talk at Harvard University Center for Geographic Analysis, Harvard University, September 2019, Boston, MA.

Maps & Minds: Geospatial Analytics for Studying Human Cognition

Garrett C. Millar

Preliminary Oral Examination

November 4th, 2020

Background & Motivation

Maps & Minds: Geospatial Analytics for Studying Human Cognition

Space-time Analytics of Human Physiology for Urban Planning

Chapter 1 Space-time Analytics of Human Physiology for Urban Planning

Background

Data & Methods

CHIPS: Cycle Highway Project

Data & Methods

Human Movement & Physiology

Data & Methods

Environmental Location

Data & Methods

Dynamic Visualization

Analysis

Environment Interaction Metrics

Analysis

Viewsheds

Results

Importance & Contributions

Chapter 1 Space-time Analytics of Human Physiology for Urban Planning

Publications

Software & Applications

Tangible landscape: A hands-on method for teaching terrain analysis

Chapter 2 Tangible Landscape: A Hands-on Method for Teaching Terrain Analysis

Background

Methods

Interaction, Feedback, & Example Solutions

Evaluation Materials

Results & Importance

Chapter 2 Tangible landscape: A hands-on method for teaching terrain analysis

Publications

Related Work

Software & Applications

Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

Chapter 3 Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

Background

Methods

Experimental Design

Methods

Setting & Apparatus

Pest or Pathogen Spread Web Platform (PoPS)

Methods

Disease Scenarios

Methods

Disease Scenarios

Data & Methods

Evaluation

Data & Methods

Proposed Analysis

Importance & Contributions

Chapter 3 Usability Evaluation of Web-Based SDSS for Collaborative Management of Biological Invasions

Related Work

Related Software & Applications

Related Presentations

Timeline

Extra Slides

Other work Increasing Underrepresented High School Students’ STEM Career Awareness and Interest: An Informal Geospatial Science Program

Developed Activities

Results & Contributions

User Experience Survey (Ras et al., 2012)

Results

Knowledge Building: Tangible Lessons

Results

Knowledge Building: TMA

Results

User Experience

User Experience Survey

(Ras et al., 2012)