CHAPTER 6
TESTING MY COLLABORATIVE WORLD
Having built my world, I wanted to run a pilot test which would test
marbles world against my stated objectives. The pilot tests would be somewhat
informal, yet give me some experience with running tests should a more formal
experiment become warranted. I would ask the participants various questions
about their interest in the world, get feedback about the technical aspects of
the world itself, and ask about their desire to return to participate again. At
the same time, I also wanted to obtain data from the participants that would
hopefully support or refute a simple hypothesis. I considered a few possible pilot
test goals such as comparing performance between anonymous participants and
participants who met each other beforehand or comparing collaborative behaviors
of participants before and after participating in the world. The architecture
of the world, being so easily extended, opened a long list of feasible pilot
tests I could run. The world could be modified with minimal effort to run many
different pilot tests.
I decided to run a pilot test which would grapple with a question
related to one I had read often in magazine articles about Virtual Communities:
“what makes people remain in and return to a Web site?” Many articles propose
that people like to visit Web sites where they feel like they are part of a
community. The community can be as simple as those who drink the same cola, or
as life enhancing as people who are fighting big government against a perceived
wrong. The idea that people come to a virtual place where they feel comfortable
with the other people who are there parallels a real world
phenomenon
that keeps people coming back to their church or university. I wanted to take
the comfort idea in a slightly different direction and consider whether people
remain in and return to Web sites because they have a say in what takes place
at that Web site. Howard Rheingold’s Electronic Minds Web site was an example
of a text-based Web site where people visited and drove the conversation
themselves. I visited Electronic Minds and opened a discussion on whether the
analogy of an Information Superhighway is a good one for explaining the Web to
the masses. I certainly was drawn to Electronic Minds because I had a say in
what happened there. I quickly perceived myself as being a part of a community
with the others visiting.
In terms of internet based 3D environments such as my marbles world,
allowing for audience participation is a newer frontier. I imagine a sandbox
world where a participant can connect via the Web, bring his or her own visual
objects into the world, and share an experience with others using the objects
each participant brings. Or, I can imagine a gallery world where participants
bring their own works of art into the world and work together to present them
in the most attractive and functional manner. Gallery world’s gallery could
even be built from scratch by a participant base given appropriate tools. Or, I
an imagine a virtual golf world where participants build a golf course before
playing golf.
These three imagined worlds can all be feasibly built using VRML, an
external authoring interface and a Java server. The worlds I suggest in the last paragraph each have different
levels of participation. A sandbox world could have virtually no pre-set
objective outside of sharing time with others. A gallery world might be quite a
bit more organized. A golf world might follow very specific rules for playing
golf. The questions that begs research are “how much freedom to Web
participants want?” and “are Web participants more apt to stay in and return to
Web sites where they have control over what takes place in the world?”
The first question is very interesting, but I believe depends too much
on individual preference to test with a limited subject group. The second
question is interesting as well, but I think may have too broad a scope for a
simple pilot test to provide meaningful results. The question I decided to
investigate is whether Web participants perform better in worlds where they
design the world and participate in the rules determination than those Web
participants who just show up, read the rules, and participate in a pre-built,
3D world.
Since marbles world provides a library of objects, a library of
effects, and a library of rules from which to define what takes place in the
world, there is a opportunity to set up a pilot test between two groups: one
group which together defines the rules, builds the board, and runs the
simulation, and another group which just runs the simulation the previous group
defines.
The hypothesis I offer is that the group that creates the world through
its participants will perform better given the rules they choose than the group
that just follows the rules and design set by the previous group. I am not sure
of the full implications of those expected results, but I believe they are
related to experiments run on a participant’s sense of immersion in 3D worlds.
Those experiments suggest that virtual participants feel more immersed in a
world when they are interacting with it based on their ideas and actions. If
the hypothesis holds true, Web designers would be supported in creating Web
sites with more flexibility. If the hypothesis is negated, Web designers may
well be justified in continuing to mass produce more static Web pages designed
to work well with even the first Web browsers and Web TV.
It seems to me that the Web becomes significantly less important if Web
participants are not craving participation with others. Broadcast TV can
deliver content to the masses effectively. The Web provides a possible direct
connection between any two computers (or people) in the world. With the Web,
there exists an opportunity to use technology in completely new ways. If people
can use a Web connection to create something new of value, proliferation of the
Web may be quite a wise investment. Since my hope through this project has been
to become effective at using a technology that provides new and useful ways of
collaborating electronically, this pilot test will also provide some initial
feedback as to whether people would use the technology given the chance.
As a last paragraph before outlining my pilot test, I want to mention
how communicating and sharing through a Web site could dramatically change the
way certain things happen in society. Today, many of the products and services
consumers purchase in the marketplace are first introduced through a product
development cycle which very often takes place within the walls of large
corporations. Consumers get the opportunity to provide feedback to product
development personnel through surveys and focus groups. Although the Web has
made it easier for consumers to provide feedback, flexible Web sites could
allow consumers to design their own products and then take them directly to a
product or service provider for realization. Similarly, many games are created
by game development companies which instead could be creatively developed and
tested within the confines of a flexible Web site by independent Web
participants.
Product and game development worlds can be built along the same design
philosophies I follow in building marbles world. The question is whether people
would use them given the chance.
Preparing
Marbles World for a Pilot Test
In order to prepare marbles world for the pilot tests I wanted to run,
I decided to remove some of the code which had provided flexibility. I figured
the results would be too hard to analyze if participants could choose from the
full rules set I had created and tested. How could I compare simulations run in
competitive mode and collaborative mode? I decided to limit the pilot tests to
collaborative mode since I was most interested in collaboration through the Web
anyway. I then decided to use a fixed set of objects, events and physics to cut
down on the learning curve for participants. By discussing the upcoming tests
with fellow colleagues, I came up with a fixed way to manage the slant queue. I
decided the world would load initially with the first two slants already
determined. I found that the group design activity was more interesting with
the two slants already in the queue. Lastly, I spent time fixing the server
code so that a specific list of questions would be asked to any group that was
determining the rules. I then was able to focus my code test plan more
specifically.
My
Pilot Test Plan
I will find twelve or sixteen subjects who are willing to spend a
couple of hours maximum in front of a computer monitor participating in a
virtual world. Three or four subjects will participate at a time. For each
group, I will give them written instructions to read and then spend time with
them answering all there questions except questions about group strategy. Each
group will participate in two simulations. The first time around, the group
will collaborate in a world designed by a previous group. The second time they
will choose the rules and physics for the world, design the world, and then run
the simulation in a manner similar to the first time. All participants will
work toward an objective of reaching a fixed number of goals in as little time
as possible. After each simulation, all participants will answer a post-simulation
questionnaire. I will then look at the differences between answers from design
round participants and non-design round participants.
All participants will be able to read about the world before they
participate and ask questions about any directions or procedures they do not
completely understand. But, the participants will have no previous physical
practice with the world. I will only require that participants have average or
better mouse skills. My world will be fixed in collaborative mode in order that
a group’s success will depend on how quickly the participants get at least one
marble (but any marble) to reach all the goals that appear in the world. The
world will keep track of how many frames it takes to finish the simulation, as
it also keeps track of the rules and design. Should the participants be unable
to reach all of the goals within 2000 frames, I will stop the simulation and
count the number of goals they were able to reach.
After their participation, I will ask each participant questions that
rate their experience in the world as to how much they enjoyed the experience,
how well they believe the technology performed, how likely they would return to
the world, and how immersed they felt in the experience. The simulation will be
considered more successful if participants enjoyed the experience, supported
the technology, felt likely to return, and felt immersed in the simulation. My
quantitative data will report on the performance of the groups. I will
introduce a scoring system to the world that will score points for getting to
goals in the shortest time possible. I will compare the cumulative scores
between the design simulations and the non-design simulations as a more
objective comparison of the two groups in action.
The
Post-Simulation Questionnaire
The questionnaire I will give to each participant after they
participate appears as follows:
Please
answer the following questions on a scale of 1 to 7, 1 meaning you strongly
disagree and 7 meaning you strongly agree:
1. I enjoyed participating in the simulation.
2. I became immersed in the simulation with
little awareness of other things going on around me.
3. I would like to participate again with the same group of
participants.
4. I would like to participate again with a new group of participants.
5. I thought the technology worked well.
6. The interface was natural to use.
7. I felt like I was treated as an equal in the simulation.
8. I learned something about collaboration during the simulation.
9. I had more control of what took place than I anticipated.
10. I would like to have more control of what happens in simulation.
Please
provide written comments related to the following:
11. Please describe your frame of mind when you started the simulation.
12. Please describe your frame of mind during the simulation.
13. Please describe your frame of mind right now.
14. Please list any frustrations you experienced while participating in
the simulation.
I ask
question 1 because I believe Web visitors must enjoy participating in a world
if they are to become part of its community for the long-term. To evaluate my
world, I must determine whether participants enjoyed the simulation. I ask
question 2 because I am interested whether my world was immersive or not.
Immersive worlds are more successful than non-immersive worlds in keeping a
community together and vibrant. I ask questions 3 and 4 since more successful
communities have members who want to return to visit often. I ask question 5 to
differentiate between disinterest from problems with the technology and the
disinterest from the underlying idea for the community. I believe the technical
problems can be easily overcome. Fundamental idea problems would be more
difficult to overcome. I ask question 6 to grade myself on my interface design.
I ask question 7 to confirm that a server is a fair and just facilitator of
actions. I ask question 8 hopeful that participants will learn about
collaboration during their participation. I believe they would learn even more
if they participated in a competitive
mode as well. I am not providing competitive rounds as part of the pilot
test. I ask questions 9 to get a sense of whether expectations from reading my
instructions coincide with the reality of the simulation. I ask question 10 to
see how participants feel about not having as much control as many video games
provide. I ask questions 11, 12, and 13 to get a sense of how participants’
thoughts change throughout the pilot test. I ask question 14 to get feedback
that could help me make the simulation better. I believe each question will
provide valid feedback. I now run the pilot test and review the results in
Chapter 7.