What's Currently Wrong with Our Tutorials
- Nov 01, 2012
- 3 Comments
This week I'm in the process of trying to trim down our tutorials into something that is more manageable and less over-whelming to our users. The basic assumption here is that our tutorials have too many steps. So we need remove some of the steps and only show steps that will help the user learn something new. Once I started into investigating this process I realized that there are many problems with our current system which may severely impact the design of a trimmed down tutorial. So I started taking a look at some of the problems I had during our previous testing and some possible ways to address those problems.
-
Users typically don't notice the post-it/tutorial directions. So they don't understand why they can't click on anything. I think exploring similar and some alternative designs may help here. I also think that a simple heuristic of three failed clicks (or similar) on the on-request stencils could cause the post-it to flash (or similar). However, some of the ideas for alternative designs may help with shorting the tutorials.
-
Users do not understand why they are doing a tutorial. They are surprised that they are now in a tutorial because there was no indication that a tutorial was going to happen. We need give users an expectation of what is going to happen and what is going to be expected of them. Maybe we should change the introduction of the tutorial to address this. However, I don't see this as a problem with a solution that may impact shorting the tutorials, so this will have to be addressed later.
-
Users do not know that can use the “Show Me How” feature to complete a step. This may be an artifact of how I conducted user testing and the directions I gave users. The alternative tutorial presentation styles may help address this issue, further a simple heuristic may be able to help us out here. Maybe we can give focus to the Show Me How, when the user isn't making progress. Parts of this may be useful for a simple “Auto-advance” feature to shorten the tutorial.
-
Users still zombie click through the stencils-based tutorials when they click “Show Me How”. This limits how users can learn new programming skills. Exploring new tutorial presentation styles may help here... but I fear we may loose part of the effect of the on-request stencils. I think maybe removing the “setting the interface in the right state” steps from the tutorial may really help here. I should be able to test this easily and it is probably a partial solution to shortening the tutorial regardless of the current mode of the tutorial. However, I think I need to figure a way to communicate this to the user.
-
Frequently, users do not read the things that would help them. This usually causes them to get stuck and not know what to do. I think we can do a better job of providing simple little affordances of where users should focus their attention. This is probably useful for shortening the tutorial since easy steps don't really need to be “auto-advanced” or described with words.
-
The tutorials are too long. Users loose interest as their patience and attention wain while the tutorials drags on forever. That's what I'm working on! ;)
-
Users have no real indication what their current progress in the tutorial is. I think this may help with patience knowing what's left to complete. But, honestly I don't think like a kid well enough to access this claim. So I'm not sure what to make of it now. I should probably think of simple ways to expose this during my testing just to get a handle on whether this is a problem and whether it would help anything.
-
All aspects of our interface need tutorial assistance. Currently the main interface is the only part of the tutorial system with the tutorial; the dialogs have no tutorial support. The stencils solution is very limiting. We need deeper tutorial support in the Croquet toolkit. This can wait for now.
-
The stencils may not be a sufficient mechanism to control the user. We probably need to build into Croquet a way for a widget to accept or reject input. I think we can do this reasonably with using point-cuts from aspect oriented programming. This can wait for now.
-
Users expect the play button to always work. This follows from the previous point. We need more control over the entire interface not just intercepting clicks at a superficial level. For now, I think I should just cut a hole in the stencil over the play button. Oh... but we have multi-modal play... this will be harder than I thought. Maybe have a play button on the stencil for now?
-
The tutorials do not support re-ordering of sub-steps or alternative paths to the same solution. This follows from above. The models and their views need a much more active roll in managing the tutorial. This will come from re-doing the tutorial generation process to be more flexible during tutorial execution. Too much work for now and it doesn't really impact shortening the tutorials too much.
-
Users enter the tutorial from the remix process. They often realize when they are almost done with or done with the tutorial they got the remix wrong. This is obviously not a problem with the tutorial. Shannon and I started working on some decent solutions to this problem this summer. We felt that users really need to have some expectation of what happens in the remix process otherwise it just seems like they get a lot of seemingly meaningless dialogs that keep popping up.... preview it, now name and describe it, now re-cast it, now you have a tutorial. This process needs a fresh look.
-
Users don't use the tutorial outline combobox. I'm not sure whether they don't notice it or whether it has no use to them since they are trying to get through a series of sequential steps. Maybe the alternative presentation styles may address this. So I think I'll investigate this a bit because it may help with ideas for shortening the tutorial.
-
Users possibly loose track of their intended goal: remix an animation. I'm not sure about this one, but I think many users may have forgotten why they are doing this long process which ultimately results in a remixed animation. I'm not sure what to do about this, but it isn't important for the shortening the tutorial now. Further, if we shorten the tutorials this may not even be an issue.
Comments
jordana said:
<p>A progress bar or some other visualization of progression sounds like an awesome idea.</p>
neonninja3.14 asked a question:
neonninja3.14 asked a question:
So how do I work collision detection correctly? I'm trying to recreate pong, and rather than acting correctly by bouncing off of the paddles, the ball doesn't turn enough or turns too far on the z axis and won't hit the second paddle when it lines up with it on the x and y axes.
Log In or Sign Up to leave a comment.