Job Simulator- Perception neuron

Hi,

There is no documentation because we are developers and not scientists making a publication :)

For user testing we used 4 games/experiences, two of our own, The Lab from Steam and Job simulator. People played them in 2 sessions that lasted 30-45 min sessions. One session was  with Oculus Rift +Touch and other HTC Vive. Most would have wanted to continue playing. Then later people answered in a survey which included sections for things like nausea, fun factor, easy of use etc. which we then used to make minor changes to ours.

Oculus was not the only one doing VR research for the first generation consumer versions and there were others too like Valve and HTC that made the Vive.

Hi Js2K.
Thank you for your answer.

Yes that's true, you're not scientists making pubblications.
But without having none data, I can't tell you nothing about nothing
of your experiments.

In any case, the biggest respect fot your experimental developer
work.

Only thing I can tell you here is that:
-using too many experimental conditions, makes you (very, very) difficult to control all the involved -or that might be involved- variables in your experiments.
This may invalidate all the detected data, pratically making your
effort useless, forcing you to do -not always- correct software
or hadware changes.

In fact with 4 games experiences you can't perfectly control all the involved variables.

In this case, it should be better to design just 2 experimental conditions, within the same (game + peripheral) experimented both with Oculus and Htc-Vive.

For example a modified (by you) game, experimented in
a simple session:
1- with  (Oculus + a specific peripheral)
2- with (Htc + the same specific peripheral)
3- survey subministration + (structured video analysis)
4- succesive modification of software + hardware
.

When I've bought Oculus there was no Htc-Vive in the market
nor was possible to know if they would ever put it in the market.

In any case, if you would you like to share here knowledge related to Oculus development or Htc-Vive, I would be more than happy to participate
with my knowledge.

Thank you
for you attention.

Best regards

Yes we were only focused to have good experience in our type gameplay and only in our experiences so the testing was limited to exactly the features and devices we will support. Every tester played with both devices but the sessions were split and reversed.

- In session 1, group 1 used Rift to play section 1 of the game and group 2 used Vive to play section 1 of the experiences. Then survey was filled.
- In session 2, group 1 used Vive to play section 2 and group 2 used Rift. Then survey was filled.

Good morning Js2k.

Focusing just on your gameplay plus interaction with devices and interested features is good.

Don't forget that your hypothesis on the related human computer interaction has always to be very minimal and clear-cut.

The software and hardware study of human-computer interaction always have to discover the specific flaws of both hardware and software.
If your research is focused on confirm or validate your initial thoughts, or developings you will -Never- understand if your thoughts, hypothesis, developed hardwares and softwares brings to naturalness of brain- computer interaction.

When you did the experiment design you've splitted groups and sessions, here you did well. Avoiding the learning effect is a Very, Very important way to prevent bias and invalidating of data.

But then you did bad when you did the session 2.
In fact you correctly sent group 1 to use Vive -very well- but the group 1 should have had to do always section - 1 - to understand which flaws might afflict hardware, software and the interaction between the two.

The same should have had being done for the group two:
correctly you sent group 2  to experience the Rift.
But the group 2 in this case -to be compared with the group 1- in the same experience, should have had experience
the Oculus with the same session - 1 -.

As said you always have to hold in mind the concept of clear-cut and minimal hypothesis.

If you don't do minimal, clear-cut hypothesis and experiment design there's a high probability for you, to invalidate your data.
The consequence is that you then can't clearly understand if your software has -or not- some flaws.
You neither can understand if you did or not some progress, neither is understandable if your progress has positively hitted the software or if continues to have flaws. Another issue should be the possibility that you confound your software's flaws, with the hardware's flaws or vice versa.

So as said always keep in mind minimal, clear-cut hypothesis and experiment's design concept.

If you have some questions, -don't worry- you can ask me without problems.

Best regards.

Nicolò