[Sigia-l] Instructive Interaction vs. Contextual Help vs. Print Documentati on

[John O'Donovan]

Shep said:
> Anyone familiar with these concepts in production or pre-production?
> Web, client-server, free-standing - any format?
> Does this concept scale well?
> Anyone have any first-hand knowledge?
> Pros?
> Cons?

Hi Shep, I have generally found instructional help to be a good thing.

There are a couple of angles to instructional help and how you actually
integrate it. As an example take a look at this game I did some years ago:

Go to this page:
http://www.bbc.co.uk/radio1/dance/
Then scroll to the bottom and click the Fatboy Slim Mixing Game link.

When you open the game click on the how to play section and you are given a
walkthrough of how to play that both textually and visually describe how to
play the game. You also get an artificial mouse showing you were to click on
the very first page - we put this there because some people opened the game
and got confused, being unsure what to do next.

There were some complex issues to explain in the interface so we went to the
effort of creating a help section that showed people what the key tasks were
very clearly. They figured out the rest.

The upshot was that we did not get a single request asking how to play - it
hit the spot contextually for the audience.

2 caveats:
1) Please don't tell me we could have used a detect script there was a
reason at the time we didn't.
2) If you get gaps between samples I know - there didn't used to be but
there is now due to some timing issues in versions of the Flash player. The
game is quite old and was only designed to last for 6 weeks during a
competition. It's just there for fun...


Personally I think help works best if it combines three things:

1) Contextual help that tells you what everything does in brief summary as a
quick reference (ideally accessible through contextual help in the
interface, e.g windows F1 key)
2) Examples that show you how to do tasks and combinations of things to
reach a goal
3) The ability to combine the previous two into the interface and show you,
within context, how you actually do what you are trying to do. Lead you
through the interactions...

Wizards and other forms of contextual interface are becoming more prevalent
but they can tend to hit too low a common denominator. They get annoying
real fast.

When you look at very complex applications like 3D renderers the value in
printed material quite often comes after you know how to use the thing and
from the value added examples in tutorial books. Examples that show how to
achieve complex results. So the value builds up in three ways:

1) What does everything do and what does it all mean
2) How to use the interface to get things done
3) Showing advanced and complex examples

An underlying issue is also whether you need to explain any theories or
paradigms so that the interface makes sense. This means linking off to
deeper reference and theories to portray foundation concepts.

In the interface for most development environments you will find extensive
contextual help backed up with examples. You also find features like
auto-complete that bring the idea of contextual and reference help right
into the users tasks.

http://www.macdevcenter.com/pub/a/mac/2002/10/29/dev_osx.html?page=last

I'm not saying it's perfect but it saves remembering thousands of function
names, which for me frees up valuable brain cells to do something less
banal. Quite often help in these sort of tools is used to look up spelling
of function names, how many parameters, etc. so address this problem and you
have achieved a practical implementation of contextual help by offering
lookup lists.

Something to remember is that often people do not need to know everything
about an interface to reach their goal and this is often forgotten in help
files.

Cheers,

jod