PGA Workshop 1 Notes - Plenary Sessions
September 15-16, 2014
(Note taker: Cynthia Jimes)
Day 1 - 9.15.14
SETTING THE STAGE
GPII – video clip: wizard-based, cloud based, stores profiles in cloud, so that users can go to any device and experiences same thing based on his/her user profile
Statement of problem:
GPII - Infrastructure:
makes it easy to find out what things will help a person to use digital interfaces
allows users to invoke the features they need, anywhere, any device
makes things easier and less expensive
Discovery tool work seeks to answer: how do we discover what it is people really need?
Themes and principles of our work:
cross context, any device, auto personalization
applies to both interfaces and content
focuses on what do we need to do to help user be able to use this. focus isn’t on the person, but what we need to do for the user to be able to access it
Demo of autopersonalization features for seniors, other users
Large range of preferences - so complex
users often unaware and unprepared to be explicit, so tools have to support user autonomy, and be smart engaging helpful etc.
only 11 percent of users ever modify their settings on their interfaces. so how do we build tools that help those folks that don’t, and how can tools reflect an environment of use? voting, etc.
Current needs and preference ideas:
inherit settings from a user’s “favorite” things
offer a few settings in use followed by an offer to store them for future use
infer from people’s performance, what their settings might be
discovery tool: friendly inventory of users needs
What has come up for the group--let’s pause--are there things you’ve seen in your settings that might be useful for the group at this point?
Phil: the primary area to provide accommodations is ergonomics, adjusting the work station. you’ve talked about screen interface and making that more accessible...and at UMN also accessibility focused on the web. But if you are in a wheelchair that doesn’t help. So having the type of controls so that the physical needs are also met when you come up to a workstation, in addition to perceptual needs. this is important too
Jutta: We can think outside box using that example Phil gave. So the kiosk...one of approaches we are looking at there is to use your personal device and a remote link to connect to the kiosk. so not change the physical manufactured kiosk per se, but will achieve an accessibility (physical needs being met) by helping user connect remotely. So it is good to think beyond the obvious when we work toward solutions in this workshop
Whitney: are you assuming an internet connection in general with gpii?
gregg: gpii can be installed and can have prefs on your key token so it is possible to have it work without having ever been connected to internet
Whitney: yes, polling places/voting is constrained environment, lots of physical constraints, ften in basements, etc., and everything has to be constrained in the room
Jim: yes, all four settings have unique characteristics, and we’ll see this from our breakout work. Constraints in senior center is least perhaps constrained, voting both most constrained, then learning is somewhere in middle.
Our starting context for discovery tools
with discovery tools: we are looking at what are the first things we need a user to know to be able to use a computer at all.
Some users can’t use a computer at all until basic prefs are set (if blind, non english speaker, switch input users, etc)
so discovery tools need to be very flexible to reach as many users as possible
Stakeholder response statements:
Phil - UMN (OER): In a uni environment, we have a full range of needs and disabilities. At same time students are sort of forced to use specific tools. Like a chat client. so then one student with a need may not be able to use that client, so then instructors go searching for others. inefficient, lots of work. so if we find a cloud based solution we may help reduce workload and this problem
Amy V. OASIS (Older citizens) - What i find interesting is that I am constantly hearing that seniors only want their questions answered, and not to sit in on a long training sessions. Need support when they need it and not through long workshops, etc. and pga/gpii type of solution can help solve for this
Shanee - NIST (voting) - The voting domain - serves everyone. lots of differences and preferences. so the end user for a voting system is everyone. at NIST we do standards, adn these are written for manuf. of voting systems to be purchased by election officials who train the poll workers on how to help voters. so there are lots of people involved in the voting process. Also, when we looked at accessibility, there is a lot going on there. there is security--can’t plug in your own device, then there is privacy--your ballot has to be your own, should do it without assistance; voting occurs every four years. Now looking at voting anywhere--so you can for example do it on ipad or library or wherever--on a personal device--then you go to the voting place and cast it. so separates ballot marking and ballot casting. but still need to be able to verify your marking and casting. how do you verify it if you’ve marked it on own device. we are looking at this.
Whitney, Center for Civic Design (voting): two things i want to add to shanee. voting is not just casting ballot. there is a lot to it. and that stuff is very important as well. elections are slow to change for all kinds of reason. we are in the process of doing a lot of rethinking about what it means to vote and be an engaged citizen in 21st c. so there are lots of opps for preferences outside the casting of the ballot (how do you know where to go? how do you decide what you will vote on? how are you engaged in the voting process? how can can interaction around voting decisions be simplified? etc). The process for voting from voting worker differs from the voter perspective. Voter perspective’s process is differeint--it’s about where do i vote, who am i voting for, etc.[she has a few readings from their reserach to share]
Scott CAST (older citizens): in my realm, input devices are where the barriers are. credential savings and password saving is also issue. if there is a way to integrate all credentials from all programs like skype etc gmail etc., that would be helpful
Emily: PhET (OER) - simulations in a classroom setting is complex. could have a range of ways to use them in a classroom. students working together, a teacher demonstrating only, etc. effective use of simulations requires that everything has to be fast and seamless because class period short. so seamlessness that gpii/pga seeks to answer will help with this. it can alos help with teachers teaching so many challenges simultaneously. so teachers might be facilitating collaboration, class discipline, new technology on top of that, etc.
Mari L. DLM (assessment) : Looking to get at real time solutions that will get students engaged. any one individual may not need a quick set of prefs but instead different goals for different contexts. and it is by subject for students often. so different pref settings for math than ela
[Stephen Bauer from NIDRR just joined]
[themes from above stakeholder responses:
Meeting contextual, varied needs of users (OER, older citizens, assessment)
the importance of real time, or targeted support for needs/preferences (older citizens, assessments)
large number of constraints of the setting, complexity of all that is involved in the setting (voting)
thinking outside of the box with the setting -- and the opportunities for preferences in areas/issues surrounding the setting or what is to be the perceived main task/goal (voting, OER) ]
CO DESIGN PROCESS
taking the one size fits one approach
integrated co design where we start with small successes, and move through rapid and then longer cycles of gathering requirements, then the analysis, then design, then implementation
our focus thus in the workshop is around defining hte challenge, sneaking up on the problem, and then gathering requirements. thus workshop is not about designing or solving the problem
Creating a plan for our co design process
Exercise: At each table, come up with an edge case that we need to consider in making the co design process work. create a growing picture of the individual and the context. (the context is us, working together, on co design)
Discussion of the goal of the exercise, in comparison to the goal of the workshop, and to the goal of our work over the next year, and how we are modeling a process for the ways we will work together over the next year by considering the boundary needs of all participants.
Themes from this exercise: helping people contribute effectively and efficiently through clearly defined tasks and timelines; flexibility in ways they can contribute; meeting needs of busy and varied schedules
SHARE OUTS FROM SESSIONS ON USE CASES:
Learners in a science class; Need to consider mechanics of learning vs. cognitive piece of learning, attitudes around learning;
Policy environment, student themselves. Assessment of learning now but there is a transition towards more self assessment, assessment for learning, formative assessment, which is more flexible. Variability across states and different test models that students may encounter across districts and states
Scope is part of the contextual piece here, because it is a larger process from knowing where to vote to deciding who to vote on to casting the vote. There are also several other issues--different leg by state, policies, election systems is currently evolving,e tc.; it’s a constrained setting in a way because it’s not all over map. it’s voting and not assessment; but that also means that we can’t change things because there are regulations
Very varied context.
Bob - broken arm, distracted, wearing glasses
Pat -13 years, moving and will transfer; feeling disengaged from formalized learning. likes bats. episodic mobility needs-fatigue increases over course of day. familiar with voice recognition, screen reader, alternative pen grip, adjustable track pad.
Minjoon, 28, family oriented, likes TV, korean singing, home help aid. vision loss, color blind, newly naturalized immigrant, remembers candidates but has problems readin gtheir names
Maud -86, retired teacher, likes to watch TV, vision impaired so some activities she likes hindered, uses magnifying glass to read, lives alone in home, widowed, familiy lives far away; no experience with technology. given an ipad though by her fam and they asked her to go to apple store to get it set up; knows she is missing out on contact and info with her family
reach 80% on quiz, collaborate with class, use LMI and access PhET simulation
He wants to pass the test, fit in more at school, wants to build a bat house at home; educators want to assess him to see where to place him in his new school
Get her ipad set up adn start using to interact with family; she wants to use facetime to talk to her family with
Needs and Prefs
Speech recognition, speech to text, organizational tools; ways to offer pereferences and changes on the fly based on needs of user at a given time
Learning content related to passion of bats, wants to take the assessment test at same time as rest of class; he needs to use AT more as day goes on (for fatigue)
color contrast, text size to be enlarged, ballot language or written language should be korean, switch to english as necessary, audio ballot that is read aloud
she likes to turn up volume on tv; large text; she might also like high contrst and text to speech; somethign to help with the tremor in her hand
Steps to accomplish goal
make sure his needs and prefs are transferred from school to school while maintaining privacy; good assessment to determine knowledge gaps; try out the ATs before taking the test to try it out
registers to vote, learn about election and when, who, where to vote, time, sign in with poll worker at voting place, show ID, decide what langauge, mark, cast
steps of using her IPad for facetime--listed these but didn’t share them
need solutions that are on the fly, changeable; what to encourage at what point given that sometimes user will be able to own the adaptations around preferences and other times need to be suggested based on inferences
Privacy issues, maintain challenge as part of learning process so it doesn’t get too easy; changing landscape in assessments
privacy issues, transmission of preferences, storage of your preferences; security-hinders own devices used; internet connectivity not allowed; different rules in each state; training poll workers on preferences and working with individual voters; how do you get the voter to get to know what to do; voting systems--budgets to buy new voting systems not there, systems are 20 years old now; what is approved by state, some states only have paper ballots, some mail in only etc., no api standards
older citizens might not want to admit/share that they have a specific need or preference, but once they see it they may want to use it; and may not be aware of what will help them [how do you help them know what they don’t know they need]; may use support from family or others--co browsing, so if the tool can recognize and be set up and prompted for supporters to help is important for this group.
Also it is about the way the settings are made, and not just which settings. so if elderly or vision impaired, i can say siri, i’m having trouble reading the text
[think about the ecosystems for each of the above settings, where can levers/policy levers be pushed, where do they constrain, where are the low hanging fruit, etc., and where are overlaps and distinctions in terms of the ecosystems. so privacy, differing policies and legislation are issues for some settings not others; issues around not wanting others to know that one needs AT is an issue for some settings and not others; lack of awareness of one’s own needs may be more of an issue for some (seniors) and less so for others (oer?)]
First discovery tools (shows diagram)
FD: It's about discovering what it is they need before we talk to them about what their preferences are. It is about how you get them from zero to get them to use things
If using AT, can do first capture, and have pre info on them – so FD is easier, done. This is First capture – for people already usig a computer
First thing we do is get first essentials – can they touch a keyboard, etc. etc. and how can we present information to them
Then first diglit: are there dig literacy barriers to determining perferences
Above can be done with professional or support as well as individual
Then you are ready for First explore, where you discover initial set of prerferences. The above bullets are first discovery
[discussion of where assistant comes in. goal is to get best out of assistance. Let’snot overemphasize role of individual or helper. The burden is on design to solve this.
well designed tools will facilitate a paired exchange as well as autonomous situation. So think about problems so we can design solutions for them]
GROUP DISCUSSION: COMMONALITIES AND DISTINCTIONS ACROSS SETTINGS
Considerations around: Where are you getting your data, where is it coming from
verification, confirming and surfacing that something isn’t working well (oer had this too)
practicing and allowing users to do practice (oer too)
Peer sharing with others who might have similar needs to you
interface requirements - tweakable - (elderly talked about this too)
Dashboard - metrics and feedback in your dashboard o your performance (also useful in OER)
different control panels for teacher and student (came up in OER too. so teacher had an interface for assigning stuff to students in oer)
privacy and security (voting too)
input data - data coming in as starting point--need to support varied stuff there (seniors had this too, but in a different way)
not everyone will be allowed to have ANY accommodation. some students are allowed to have one accommodation, and others are not
limited number of settings that this application area can handle (seniors, too)
text size limitations (unique) - so notion of constraints around personalization and settings.
Voting unique areas [check these]
[rich: did we address literacy and that some cannot read?]
what about overall, variety of types of systems, and the combinations of things/systems/ where people might need to have first discovery might be used (oer and elder is linked here). but in oerthe number of different apps and systems is a lot higher. voting - only one, assessment - a few, elder-a few. so there is a spectrum.
ways to get into process by selecting the context the user is in - eg video intro (voting had this too)
finding engaging content and meeting users where they are and personalizing with cats (assessment had this too)
gamification and ways to get feedback--part practice part reinforcement (elders too). also keeping the engaged
make the personalization process an interesting and engageing process, that is flexible, giving them as much personalization up front even in the personalization selectin process (seniors)
input and output with speech and audio. it gets at hardware. an input device or requirement for an input device before you make a preference selection that changes everything. how to let user know if you are using speech for input you will have headset so that audio output doesn’t interfere. do we mae an adjustment in the accomodation, or do we present the with best practices. inform the user of the effects of their choice. warn them what is going to happen (related to practice and tweak in elder ad to show/demo in voting)
technical question about how the preference choice is transported and whether it can be--so-transporting preferences from where they are being captured to the environment/application tool and back again (voting had this too). Loops/iteration comes in here too. but then there is knowledge from the applicaiton setting going back to discovery/preference tool to make that better. wanting the discovery process to have knowledge about the process and activities within the application setting.
so for learning - ways to enhance the learning experience based on providing data back to the discovery tools
what are we taking control of versus what are we providing options for; because can have a learner watch things over and over and do repetitive action, for reasons of cognitive etc, but the sstem might think you need better settings when in actuality you need to just watch it a few times. so maybe there is a setting “don’t bug me”
OER: a single user will very unlikely to be interacting with one system only. multiple systems.
a pretool with populated info from family member
needs to be quick and engaging and they need to see value as soon as possible (as with voting)
practice time and settings review
number of settings (maybe more limited like voting) and want to keep it somple and narrow the pre sets for this group too
demonstrate settings (like voting)
touch screen important
verify entries ability (voting)
collecting seed information through a seed tool
need instruction to use this tool in some setting if doesn’t have a helper, so maybe librarian helps them fill out the pre tool
relies on a casual/informal gatekeeper and support person (like voting
UNIQUE to elder: likelihood if they have assistance, it is NOT trained assistance
Additional considerations overall
with assessment and OER, less likely to encounter someone who hasn’t touched a computer, but with elder and voting you may
lots of scenarios where you are going in with no pre knowledge. with seniors, you are less likely to have pre knowledge, but usually some family member will.
NEXT STEPS - CLOSE
what are the next things we’d like to tackle in co-design - what are our candidates for next steps in co design, what are priorities?
rich: we need to identify 4-5 constraints in each environment and start here. so eg constrianed by assistants, limited internet access, issue with audio in library.
whitney: boxes and steps and lists and analysis--let’s not get all wound up in more of this. let’s start sketching and creating.
jutta: given the amount of commonality across settings, it seems we are settling on a toolkit of common functionality and features and then build out from that the features that are unique to the settings. so what core features and functions across that we need to design for? stakeholder round robin identified these
picking from examples for presets
provide an in-context way to to select preferences
ability to do pre tool/seeding with info by family member (for seniors)
more functional performance rather than question and selection (show options and inference/performance based inference); metrics based exercises that help preferences get selected
engaging animation or other ways to draw peiople in/ways to recruit or entice people to come into the setting/FD process
ease of use in making the preference selection
instant gratification and instant knowledge of what that looks like (trial and error)
showing the impact of your selections
HOW DO WE ENGAGE YOU IN CO DESIGN
need to make the process for design transparent so you can see what decisions have been made
takeaways from the meeting (round robin)
lots more constraints than I was aware of
commonalities more than expected
policy constraints esp in voting and assessment, how do we design with in that
one person can actually go through all of these phases/settings at some point of their lives, so how can these be linked. thinking about it like that
revisiting this all after 20+ years and opening the discussion up, and what this project can do for people; all of the possibilities
we’ve got a phenomenal team of people; strong opinions; making sure we harness all of this power is a burden and an opportunity at the same time
commonalities bodes well for sustainability
complexity of ecosystems and all the different ways users can enter in and get to the FD tool
this is more doable than I thought
application of this technology to PS education esp those institutions that don’t have a technology specialist in this area; we are talking not about changing content but delivery systems which is exciting