53. Test plans
53
in-session testing
user depended selection of features
questionnaires, video recordings
test and stop protocols
pilot, sessions, longitudinal
HMI, interaction
HMI:
create college and plans
Please: reflections, ideas etc.targeted conferences journals whatever….
No concepts user test. Another point of attention I will use the same “52” slides. I might skip some stuff ad-hoc
Create: I am a Phd on interactive play involved in a project on designing an interactive play environment both for children and this is about such an environment for people with profound intellectual and multiple disabilities. (PIMD)
[also see http://www.allaboutux.org/ for useful methods, descriptions etc.]
[proces verhaal wat ze niet vaak genoeg kunnen horen, interesante doelgroep, en laten zien dat je keuzes moet maken]
Presentation:
lot of information in a very low-level
general design and research process
User group:
Multiple limitation that strengthen each other so no compensational possibilities
No verbal communiactions
Spastic or parralized motor control
Very heavily handicapped your average down syndrome person.
R&D
entertainment for abled
ironically less entertainment for disabled.
A lot of them are under stimulated.
improve qol
Efficacy, autonomy, perhaps social-relatedness
Research can be hands on
Some normal practice tools we can use as inspiration for new technology,
I will go through them and explain how I applied them or why they can’t be used as normally because of the user group
Often time consuming so we would use focus group of the users,
However our user group can’t communicate verbally
So instead, we talked to several caregivers, managment,
[From http://www.mirtetak.nl/uploads/Loes%20en%20Ward.jpg]
Some normal practice tools we can use as inspiration for new technology,
I will go through them
Entertainment use
Interaction : no
Diaglogue is missing
Did provide inside
[screenshot from http://www.barryemons.nl/c/Snoezelen/6/]
Not enough stimulation, too relaxing, even the care giver would occasionaly fall asleep
Very limited amount of research,
SID project by Larssen et al.
[TODO perhaps movie]
Obesity -> exergaming
An ability perspective,
Participate in society
Research approach, goals, plans and evaluation
efficacy, autonomy, social-relatedness
A function is a specific or discrete action that is necessary to achieve a given object. (Blanchard&Fabrycky) [presentation productfunctieanalyse Bonnema 2005]
Action: verb + noun
hold camera sensor
protect camera sensor etc..
take pictures
Search for what is important
Core function, subfunctions, and supporting functionality
And search for alternative ways to come to that function
One might also look at societal functionality, which is related somewhat to the earlier mentioned trends in care taking
As there is little in existing ideas and products I looked at recent technology developments and ways to interact, sense and activate
[skip >3m] One quite nice piece of information was a siggraph lecture/workshop from Beckhaus and Kruijf, a little outdated 2004, but still quite accurate.
Send questions about preferences, current situations, iritations, product ideas etc.
Large group quick analyses, often low level of detail
This user group cant write and even indicating preferences is hard
Instead I looked at people who looked at a set of our user group
stories of care takers and experiences with products.
Platforms, websites, designers and books.
User profiles (explain user scenarios): family, preferences, wants and needs, targets with therapy and care taking , habits , abilties etc.
Contacted a master student and phd-student (one of the few working on it) and bought a book on designing
So I did not yet implement this, but one could do this.
[pictures : front page multisensory rooms, flaghouse website, villans website and larrybissett.com]
A little bit of history repeating,
path dependecy might give new directions,
Or pattern recognition
[picture taken from another lecturer on history based design]
Another way to look at trends is to generalize.
For instance using TRIZ or the simpler Systematic Inventive Thinking.
TRIZ, a study based on a large number of patents and rrecognizing patterns across disciplines, i stole this sheet from valery who gave an inspiring set of courses. I hope he doesn’t mind
We are not that creative ;)
[slide from valery who gave an inspiring set of courses.]
Another example showing the cross context point
[slide from valery who gave an inspiring set of courses. I hope he doesn’t mind ]
Or a mobile phone
[slide from valery who gave an inspiring set of courses. I hope he doesn’t mind ]
Being part of the user group to really understand what is happening, in this case for instance to be part of the therapists/ care takers. Perhaps even mimiciking disabilities might help.
From cultural and social anthropology
Don’t have permission everywhere to take pictures, should ask etc. but you have to ask the gardians. So this again brings an additional step in the process. Luckily most of the guardians will say yes as they also want the project to succeed and agree with the care takers in most cases.
[alvimann , morguefiles]
We did more a traditional field study, just observing and casually interviewing the care takers. Limited amount of working together with user group but close observation were very useful
For instance, self-regulatory behavior (ticks) … even we do that … [explain over stimulation under stimulation]
It showed a certain sense of rhythm and we now think they are quite capable in that.
For instance rolling on the floor, we came up with an interactive matras that reacts on the side one lies on although the waterbed is quite similar already.
[pictures of vita-care tms, dichterbij housing homeltjeshof 2x, dichterbij brochure,
I won’t say requirements, because they are boring right? But unfortunately we do need them.
Incontinent and drooling water resistant
Gloves used against self-harming, biters should be resistant to that so not swallable and be too big or bitable.
[not the actual gloves used, often washing gloves suffice)
Many are epileptic
Moodboards, compare to existing products.
Some say childish primary colours
others say they are adults
Normal controls wont work, no fine motoric skills
[picture motomed]
Many in wheelchair
Not autonomous
Lifting hoistdistyraced by surroundings
I won’t say requirements, because they are boring right? But unfortunately we do need them.
Some normal measures to look at when testing a designed object. There are some that give insight mainly to help to improve (more subjective measures) and some that generate absolute numbers (objectve).
Both can be useful in different cases. If you want to proof a medicine works a qualitative approach is preferred, if you want know how you can improve a certain interface a qualitative approach might be preferred although testing both in a quantitative way works as well.
To complicate things some measures I mention as qualitative can be quantified. At the other hand some quantitative data might be used in qualitative way or are subjective in nature (e.g. a questionnaire on feelings or preferences).
Again we can only apply those on the care takers
Sometimes write down what is happening directly, other times record, and or use one-way mirrors to let the user feel comfortable. There all sorts of ways to facilitate observations using so called annotation schemes to categorize and to get some quantitative data.
You have to be trained quite well to recognize and interpret behavior. Some might be seen by us but many things will only be picked up by the care takers that know the person for quite some time. So video recordings are in order. Again this also requres consent from guardians wich have to be prerpared. Furthermore it might be good to have a medical ethical committee approving the research with the designed product to begin with
[also look at pre post conditions observations]
Think of heart rate, eeg, fmri, temprature, skin conductance (sweat), saturation (level of oxygen in blood).
A problem is that many are oversensitive so “wiring them up” might influence the test too much.
Think of heuristics, expert observations and annotation, expert interviews, expert ordering of different versions of the product, watching experts in the field and interview them etc.
[heuristics from mit course slide, the “field” observations picture the rupps, agweb.com]
Either real user actions (tracking with accelerometers, eytrackers or kinects etc) or using the programs system actions and input.
Some complications occur, they are in a wheelchair, this is not recognised that well. The open-endedness complicates the logging, did they intend to do a certain action?
Strange facial expressions and deformations might diminish applicability
Focus of attention is short and at the same time calmness is necessary as they might respond slowly
[eyetracking picture from mirametrix website]
In our user group differences are enormous. At the same time it seems there are not enough users to test to generalize results. Comparing usage of a blind person and a deaf person of a visual installation seems useless to test the installation.
Furthermore in-subject testing is complicated because they have quite severe mood swings.
[pic between subjects http://cbrownley.wordpress.com/2011/05/01/difference-in-means/]
[mona lisa http://sunnylol.com/different-versions-of-mona-lisa]
Lab or confined space testing.. However for this group a confined space might be neccesary as they are collocated and might be affected in unexpected ways..
One change or various options…
The number of tests that can be done is limited due to the amount of people, travel time, schedules (sleeping time etc), influences of medication. At the same time the variety in users might request a variety of solutions, thus we probably need several options in order to test multiple users.
[link to rite method, short term iteration steps]
[http://windows-7-theme.en.softonic.com/ , http://cartelthemes.com/windows-wallpaper-themes-19610-hd-wallpapers-background.html ,
Because of the mood swigs first-use testing is kind of limited. In the SID-project, with the waterbed shown earlier, it was mentioned that some users only after 5 times recognised the interaction.
[image of wii (gadget effect) http://www.xgn.nl/nieuws/50076/nintendo-stopt-met-productie-wii-in-japan/]
Due to these earlier mentioned limitations, we have to make a user selection, only some can test the product. At the other hand we want to have an inclusive design, it is exactly because these people are forgotten so often in interaction design that we design for them. Now we seem to be forced to again exclude a sub-group.
[http://www.british-sign.co.uk/british-sign-language/how-to-sign/deaf/]
http://www.designblind.co.uk/
Taking into account the mood swings, the differences between people it seems applicable to alter design choices during the test
As some user might be easily over stimulated while others are under stimulated and won’t even notice the product of it is not multi modal
we need to address this before testing certain choices
As some user might be easily over stimulated while others are under stimulated and won’t even notice the product of it is not multi modal
we need to address this before testing certain choices
As some user might be easily over stimulated while others are under stimulated and won’t even notice the product of it is not multi modal
we need to address this before testing certain choices
We need pilots to test the actual system, does it respond as planned, do all the technical stuff work, does the principal work on a non-user group,
We need sessions to test whether the users understand the product and make rapid changes for the next test session
We need longitudinal tests to show the product could be used and helps to attain the predifined goals.
BUT….
Although there are a lot of complicating factors it is not all bad..
Happy movies of SID-project
And it is new so anything is quite cool anyway
Example inspiration movies of SID project:
http://vimeo.com/groups/114237
Movies from the project
http://sid.desiign.org./
Show concept, tech push
A ball
And I end with this panda, because pandas have a very imperfect design, but I really like this picture!