Note: the ongoing "official" testing plan document can be found here: http://groups.drupal.org/node/7929
As announced here http://drupal.org/node/204667, Drupal will undergo formal lab usability testing this coming February. This project builds on the already impressive work of the usability group and elsewhere. We feel honored to play a part in this ongoing effort.
The testing process begins with the identification of high-level goals; target audiences; common user tasks for these audiences; and a set of debriefing questions for the evaluators that we recruit for testing (users who test the product are called “evaluators”). It’s critical to the success of this project that we scope the goals, target audiences and tasks appropriately.
The U Libraries testing team has drafted a goal statement, target audience list, task list and proposed a variety of debriefing questions as a preliminary measure towards moving this process forward. We now need the Drupal community to help set priorities; propose and refine tasks; suggest debriefing questions; and to help out with various other tasks.
Please take a look at the following rough draft and offer (specific) feedback towards the identification of goals, target audiences, tasks and debriefing questions.
Thanks!
As Drupal has yet to undergo any formal usability testing, our primary objective is to identify audiences and craft scenarios that will enable us to get a baseline understanding of Drupal’s usability for common tasks. As a result overall testing priority will be given to Content Contributors and Site Maintainers but some tasks may be reserved for the Admin role where administrative tasks would not require a great deal of prior training or experience (ex installation process ;) ).
Scenario Overview and Usability Goals
Scenario Overview
This round of testing will focus on the needs of users who maintain relatively "standard" corporate (or departmental) websites. This is the general context in which we identify users and their needs. The audiences that follow represent users who would play a role in this process.
[Placeholder Usability Goals]
@COMMENT: I just pulled these out of you-know-where to get the ball rolling. /libsys
- New users with minimal web design/development experience should be able to perform basic content management tasks with moderate direction. Such tasks include adding pages to the site, adding links to pages within site menus, placing pages within the site hierarchy and commenting on content.
- New users with moderate to advanced web design/development experience should be able to perform critical system maintenance operations. Such task might include administering users (adding them, adjusting access privileges, etc), creating data entry forms for content contributors (see definition below) to manage dynamic content, building hierarchical sites, adjusting the look and feel of sites, locating and managing versioned content, etc.
Audiences:
- Content Contributor
- We envision the content contributor role to be that of an authenticated user with permission to create content such as pages, menu items, and comments.
- Site Maintainer (editor/subadmin)
- We envision the site maintainer as an authenticated user with access to manage user permissions and content, including Blocks, Taxonomy, Menus, Comments, and all Nodes.
- Admin
- We envision the Admin as an authenticated user with access to all site settings, responsible for installation and configuration.
Content Contributor Scenarios
Summary
Our goal here is to create a scenario that reasonably approximates the common tasks performed by low-level users. These tasks should not require an intimate knowledge of Drupal. Ideally, users who are familiar with any web CMS, blog or wiki platform will be comfortable with these tasks.
The contributor tasks involve adding a page, creating a link in the menu system to that page, finding an old page in the system and reverting it to a prior version and updating their username and password.
Task Scenarios
- You’ve been asked to create a new page in your site’s “about” section that describes your organization’s facilities. Create a page containing the following text: “Our facilities are located in Minneapolis.”
- Now you need to create another site section.
- You realize that you made a mistake on the “About Our Facilities” page, please go back and correct the text to read: “Our facilities are located in both Minneapolis and St. Paul.”
- Your supervisor is not happy with changes you made to a page you updated some time ago. She wants you to change it back to the original. Please find this page and undo your change (set the page to a previous version).
Contributor Debriefing
- Go back and edit the page: what does”Split Summary at Cursor” mean to you?
- Click this button – what do you think it does?
- Explain if they don’t get it and ask them what they think.
- Scroll to the Bottom of the page. Without clicking, what do you think these links (Revision INfo, Authoring Info, Publishing Options) do?
- Click on each (expand fieldset), what do you think they do now?
- What does each option in “Publishing Options” mean?
Debriefing questions could include getting user feedback on:
- Node Add/Edit features (Authoring Information, Revision Information, Publishing Options) – what do they mean?
- Expected Features that were lacking
- Overall Impressions
Open Questions:
- Do we test the taxonomy system for these users? Which type of taxonomy do we test? (ex freetag? both?)
Site Maintainer (“subadmin”, editor, etc)
Summary
The goal of these tests is to create a scenario that reasonably approximates the common tasks performed by a site maintainer. These tasks will require some familiarity with web content management systems in general, if not Drupal specifically. We will endeavor to capture testers’ initial impressions of various aspects of Drupal, and then follow up with such instruction as is necessary to complete the tasks. Tasks will include enabling modules, setting up new users, adding content blocks, establishing menus and vocabulary configuration. Possible secondary tasks may include theme and site information configuration.
Site Maintainer Task Scenarios
[SCENARIO OVERVIEW] You are the webmaster for a campus department that has chosen Drupal to manage its new website. Your systems administrator downloaded Drupal and performed the preliminary installation steps. You must complete the installation and begin to build your new site.
- When Drupal is first installed, it creates an account that has full access to all system features, regardless of how the site is configured. This account is similar to the “root” user account in an operating system. For a variety of reasons, it is considered a “best practice” to avoid operating and configuring Drupal while logged on as this user. Rather, it is better to create Drupal “roles” (administrative groups of users) and to then set specific permissions for those roles. This will be your first priority.
- You are now logged on as this initial administrative user. So, your first task is to (a) create an “administrator” role (b) set the permissions such that this role has access to all enabled features© create a new user and (d) put this user into the administrator role.
- Log out and log back in as this new user.
- We’ll need another role for your next task, so go back to the user administration area and add one more role, the “faculty” role. Don’t set permissions for this role just yet.
- Your faculty want to have the ability to create and update biographies of themselves on your website. Biography pages should always contain the faculty member’s name, a picture of the faculty member and some biographical content.
- How would you get started?
- First, you’ll need to create the content entry form that provides specific web form fields for the collection of name, biography and picture data.
- Next, you’ll need to make sure that both “authenticated” and “anonymous” users (logged on users and visitors) to your website can access (view) the content that your faculty members create.
- You’ll also need to make sure that faculty members can add and edit their own biography pages. [leaving out how this is done here to see if they make the connection between users and roles based on the preceding task]
- One of the requirements of your site is to divide it up into various sections (“About the Department,” “Student Resources,” etc.). You can get started on setting this up by creating an “About” section and then allowing faculty to put their biography pages into this site section. To do this, you’ll take advantage of Drupal’s “menu” and “block” system features. Menus will allow you to create a hierarchical list of links, which will give you “site sections” and blocks are containers that hold the menu content [ick]
- [NEED INPUT HERE] This task may need to be rethought as configuring the menu system to work in conjunction with content forms and blocks to produce site sections is non-trivial. It will have to be reworded at a minimum.
- You now realize that you need an easy way to allow faculty to choose areas of research specialization on their biography pages. This can be accomplished using Drupal’s “taxonomy” system, which allows you to categorize your content using defined lists of terms. Go to the taxonomy page and make a list of research areas from which faculty can select their own.
- You need a list that contains the following areas: “Astrophysics,” “Rocket Science,” and “Brain Surgery.”
- [We should probably leave out any mention of associating the vocabulary with the bio content type – this tests to see if users make the connection between content type and taxonomy]
- You need a list that contains the following areas: “Astrophysics,” “Rocket Science,” and “Brain Surgery.”
Debriefing Questions
- What features did you expect but not find within these tasks, if anything?
- What do you least understand about what you saw? – What do you find the most confusing about this system?
Open Questions:
- We have yet to test these scenarios and do not have a sense for time to complete them. Testers (“evaluators” in usability parlance) will have about 1.5 hours each.
Comments and Questions
- Test Site Configuration Suggestions will also be Solicited
- Depending on the target audience, we are suggesting varying levels of pre-configuration w/different installations. This will make the tests more meaningful.
- An installation profile might be helpful promote scenario reuse and benchmarking
- We are particularly interested in the idea of trial run test feedback, in having Drupalers run through the scenarios with their own users and providing us with feedback
- We will have a D6 demo site up and running sometime in the next week or so
A few notes about the testing program:
- We will have 8 users for the entire testing program an industry standard number, more or less.
- Each evaluator will have 1.5 hours to complete the aloted tasks.
- We currently have split the testing plan 50/50 between the “Content Contributor” and “Site Maintainer” target audiences.
Key dates:
- January 14th – Kickoff Meeting
- This meeting will include a first pass review of tasks, audiences, etc.
- February 11th – Scenario/Prototype Review
- In depth look at scenarios and prototype walk through
- Feb 26th-27th – Testing
- Eight Evaluators will test the product
- Feb 28th – Issues Analysis
- A review of testing outcomes

Comments
Wow
This feels like a great opportunity. Maybe Drupal Usability folks should concentrate on this for a while... In Order not to scatter efforts too much in endless commenting I have opened a Googledoc for brainstorming and structuring our thoughts. Whoever wants to participate, send me a mail and I'll invite you as editor.
Life is a process
Life is a journey, not a destination
This is great!
And well timed with the Season of Usability programme.
I see there is no set of task scenarios and questions for 'site admin' role. I also think there is a simpler role missing, something like 'visitor', a user that reads searches comments and signs up on a stock drupal website. I hope to find time to go through this in detail, move it into a wiki page (Or google docs?) and start building it up, adding to it and improving it.
Can you provide more details on how many hours of testing the lab is available for, how many tests and evaluators will be available, how will they be sourced, and what experience and/or prerequisites they have?
Bevan/
Hey bevan, eigentor, It's
Hey bevan, eigentor,
It's good to see folks are getting interested in this project, we certainly welcome the help!
Regarding details the evaluator split and so on, that information is (buried) up above:
Some high priority Admin tasks (as defined above) may be rolled into the Site Maintainer tasks. Selecting and configuring a different site theme might, for example, fall into that category. These scenarios would target site-wide settings and operations that the community designate as being of very high value. The resource constraint of 8 evaluators over a period of two days forces us to establish these priorities.
A wiki or gdocs page is probably a good idea to record and refine scenarios and debriefing questions. I'll set up a wiki that spells out more details of the testing plan as well.
Wikified
I've moved the document into Usability Testing Plan where I'm currently reformatting for wiki markdown, tidying up a little and adding detail. Please don't edit the document on g.d.o until I've finished (about two hours from now should be good) as I will over-write your changes.
Cody and Chad, can you agree and confirm that http://groups.drupal.org/node/7929 is the primary copy of this document (where all changes should be merged) for the time-being, and notify the original author/s?
Thanks,
Bevan/
Bevan/
Testing details
Thanks for your encouraging enthusiasm, Thomas and Bevan. We're very excited about this process.
Regarding the details of testing, as posted above, we plan to have eight subjects test Drupal, for roughly an hour each (plus time after each session to capture important information). Perhaps Chad can comment here, but I don't recall if we'd discussed how many subjects we would assign to each role. We will likely draw from the greater University of Minnesota community for our subjects, seeking people who are familiar with basic web content creation (Content Contributor role) or with web CMSs (Site Maintainer role). We will likely exclude subjects who have prior experience with Drupal.
The "Admin" user tasks are up in the air at this point. We had considered including a clean install of Drupal as a task for an admin-type user, but after some discussion, we felt that we might do more good focusing on post-install tasks. Our rationale was threefold: 1) For each instance of Drupal, installation is a task that takes place at most just a few times (ideally...), but content contributor and subadmin tasks will recur daily, so testing those tasks will allow us to do the most good for the greatest number of users. 2) There is already very good work being done on the usability of the installation process, i.e. the FactoryJoe list. 3) We would face significant technical hurdles in establishing an environment within the usability testing lab for a Drupal install. We don't know, for instance, if it would even be possible to provide SSH/TTY access out of the lab to a server.
As to "Visitor" tasks, this is an interesting idea. We had not considered using this role in our testing. My only question about the viability of testing end-user experience would be whether we could create a site for testing that reasonably approximated the incredibly heterogeneous end-user face of Drupal sites. This is to say, are there aspects of end-user interfaces to Drupal-powered sites that are common enough to merit testing?
Of course, the reason we've brought all of this to the developer community is because we want your input. Keeping in mind that we have limited test slots, and wish to replicate our tests a few times each in order to establish some semblance of significance in our findings, we remain open to any and all suggestions.
Thanks for your feedback, keep it coming!
Regards,
Cody Hanson
University of Minnesota Libraries
"This is to say, are there
"This is to say, are there aspects of end-user interfaces to Drupal-powered sites that are common enough to merit testing?"
Well maybe certain tasks like adding a comment or using the forums and so on. But it doesn't seem like the best place to test at this stage. A lot of the data will, indeed, be site specific. And some will will converge heavily with general content adding and editing. Perhaps testing the registration process would be a good idea, but there are probably more pressing areas to test.
-
www.alanpritt.com
I just saw on
I just saw on http://groups.drupal.org/node/7905 that the dates are Feb 26-28. That's just after the CivicActions offsite, and right before DrupalCon Boston. I'll be idling in the US between those two events and would love to take part in the tests as an observer. Is this possible?
Bevan/
Bevan - I'm going to email
Bevan - I'm going to email you off list to see if we can work this out.
Done.
I've just finished an extensive pass on the document (so feel free to edit it now. Most of the changes were formatting, layout and text simplification -- all for easier reading. I added some @comments and @todos to make future work on it and planning easier. When I finished I realized I should have just used the wiki page you created the other day; http://groups.drupal.org/node/7905
I copied the remaining info there into the new wiki page and deprecated the old one. Sorry. :(
Some of the major points for discussion;
Also, when / where / how will the Jan 14 and Feb 11 meetings/reviews take place?
Bevan/
"Will the evaluator have
Finally, great start, thanks to you all for jumping in and rolling up your sleeves!
I moved this section out of
I moved this section out of the document as it makes more sense to discuss these issues here;
Discussion
Bevan/
Targeting partly on known weak spots
While it is sure important on not wanting too much and rather working out single tasks better: There are some areas in Drupal that are known - or supposed? to be hard to use.
e.g.
This might empower us to have statistical data on these. For some of these things make me cry out loud from time to time...
Life is a process
Life is a journey, not a destination
images and enabling modules
I would avoid testing core's insertion of inline images at this time. The current usability of this task is currently so bad, that I doubt we will get much useful data from this. As soon as we start making obvious improvements to the image insertion UI these test results will become meaningless. However, testing a contrib module such as IMCE (with its more mature UI) would provide more useful data; demonstrating whether we should implement some of these design decisions or avoid them. This would be particularly pertinent since there is a drive to make improvements in this area for Drupal 7.
Another area I have a personal interest in improving is the process of enabling a new module. I suggest we get the user to enable the search module as this is a module that beginners frequently enable and end up making support requests about (usually in regards to permissions). The eye tracking data would be particularly useful here as it can answer some specific questions... Do they read the status messages? Do they move down and check the status of the module after they enable it? How much of the header text do they read? What do they do after the module is enabled? Do they look for the help file for clues? ... Data here would without doubt have direct practical benefits.
-
www.alanpritt.com
You make very good points
You make very good points here.
I agree about testing image-handling; at this point in time it would be a waste of this wonderful resource the UMN is offering. On the note of IMCE and it's UI. There is an old patch that made the UI usable. I think it's too old to be worth updating now, but meh; http://drupal.org/node/139387
Testing module enabling would be great. There is a lot of discussion debate and action as to what is an improvement and what's not.
Another page I'm interested in adn unsure how to improve is the blocks configuration page. I don't think dragndrop table row is an improvement. I think having a table there is wrong to start with.
Bevan/
The long plan
When one thinks about what to test, one thing gets clear: there are much more areas to test then time and resources available. This leads me to the following:
Maybe we can look into our BIG LIST to do some grouping.
Life is a process
Life is a journey, not a destination
re: The long plan
Absolutely agreed. There is far more to test than we can possibly address in this session.
I think that one of our goals in this round of testing should be to standardize such that these tests can be replicated and repeated. I really think of this round as a benchmarking round, establishing a baseline from which we can improve.
Wording
This lab usability testing is great news!
A think that struck me when I first read the tasks scenarios was the way they were formulated. I realize it may be intended, but it seems to me some words may be confusing to newcomers.
E.g: "You’ve been asked to create a new page in your site’s “about” section". What's a section? It's a term you won't even see in Drupal's admin. So testers may start looking for this particular word, like a "sections" menu item or a "create new section" link they won't find anywhere.
What is the “standard” way to create sections anyway, with a menu structure, with Book module? Taxonomies? On some aspects Drupal is very different compared to other "usual" CMSes, and sometimes there is not a unique way to achieve even mundane tasks. In these cases prior knowledge may not help.
I'll comment on specific topics on the wiki page.
All good points. Generally,
All good points.
Generally, you try to construct scenarios that map more to the user's mental model than to the system. By not using the exact verbiage contained in the system interface you get more of a sense of how users would really interact with the product and avoid walking them through the test. "Site Sections" is a very common concept for organizational or departmental websites, which is why it made into the draft.
That being said, it may be too convoluted of a task given the lack of a clear path towards achieving "sections" in Drupal. This has less to do with terminology in the testing process and more with the various methods for achieving the result. It could be that we need to settle on a slightly less ambitious scenario for testing the menu system.
Thanks for the input, please wiki away!
It totally makes sense. I
It totally makes sense. I just wanted to be sure it was intended, so the wording wouldn't get in the way during the tests.
Actually "sections" are explained in the Site Maintainer scenario (task 3), but not in the Content Contributor one. Which seems good as you can then first test how some testers understand how the sections idea translates in Drupal, then how others deal with it when they have more detailed instructions.
couple edits
In response to a comment in the wiki, if nobody else will then I'll be happy to present this at Drupalcon. I'd probably need help, but I just wanted to confirm that it is a major goal to get these results into developer hands quickly. I don't think it will be a problem to get space in the schedule. So, I removed the comment and question mark.
/Note: I'm very new to doing studies like this, so if my edits are bad feel free to undo them. It's a wiki after all./
Site Maintainer tasks:
The discussion of UID1 being a "root" user and the idea of it being a best practice to not use that user feels both overly complex (what if they only use Windows or Mac and don't understand "root") and somewhat invalid (if you create an admin role with all permissions you are effectively UID=1). I changed this to simplify it and still provide motivation for taking the next steps without this extra information.
I shortened the description of task 2 as well so that we ask them to create the role without saying where to go. My (perhaps naive) feeling is that these should be relatively short and only include as much information as necessary, right? That way we test the usability of the system rather than the usability of our instructions, right?
For maintainer task 2.1. in the biography section we are pushing people towards a "nodes as profiles" type of solution, right? We will then need to provide an image handling solution (like imagefield or image attach) in order for them to be able to achieve the requirement for photos. Without cck and perhaps bio/nodeprofile/user_node modules this task is pretty impossible. Are we going to pre-install one of those? Profile module is an easier solution to this right now. Dries seems to stay opposed to users as nodes...
For site maintainer task 4 - we may want to remove the guidance to "use taxonomy" depending on the results of discussion about profile vs. users as nodes.
Finally, is our goal to test Drupal core as it is, or test out some contrib and some new ideas? I.e. I'd like to have an install profile that creates an about page, some taxonomies, menu items, and blocks right out of the box which I believe will make these tasks enormously easier. Should we make these sessions about testing proposed ideas or just about "what we currently have" ?
Thanks everyone for pitching in to help out!
--
Knaddisons Denver Life | mmm Chipotle Log | The Big Spanish Tour
knaddison blog | Morris Animal Foundation
Looking good....nice work
Excellent! Dries and I chatted about a presentation but did not publicly announce it as such. There will be a presentation at Drupalcon though; we haven't worked out the particulars yet. Cody and I will both participate, and I was thinking that we might have a panel discussion style presentation with the folks (such as yourself) who worked on the project...and who were interested in participating. Thank you for bumping that thread, I'm glad that you're interested.
Thanks, your changes look like an improvement to me.
Yes, that's the idea.
No, this is unrelated to profiles etc per se. It was simply a plausible scenario to test the creation of content types and related tasks.
I think the answer is that it's a little bit of both. On the one hand, some level of pre-configuration will be necessary to produce a meaningful testing scenario for content contributors. On the other, we need to be conservative in the configuration process to produce meaningfully broad impact. It's one of the principle challenges of doing usability testing on such a flexible OSS project.
Great.
Great.
If it's unrelated to user profiles I think we should make it something that won't be easily confused for profiles. How about a content type for "Student Resource" ? Are we trying to test cck with this? I think it would be valuable to test the core cck components such as adding fields for a number, text, etc. As I'm sure you know cck is very popular and is headed for core, so if we can make sure it is usable then that seems valuable even if it is "just a contrib". If so I can improve that task.
About the pre-configuration - let's see how far I get on the "newbie-Drupal" installation profile and then we can evaluate that more specifically.
knaddison blog | Morris Animal Foundation
Indeed, CCK May as Well Be Considered Core Now
Yep, we meant to test CCK for the reasons you've outlined. I'm assuming, for now, that Drupal 6 CCK will be ready for testing by February.
The task of adding fields to the biography pages was meant to test the CCK UI. Please do feel free to edit this scenario (Maintainer Task #2)to a "Student Resource" model to avoid confusion with profiles. Thanks!
great - done
Ok, edited to build a "student resource" concept.
I also switched steps 3 and 4. I think that partitioning a site into sections is a less common task and more complex in my opinion than dealing with taxonomies. I also think that the knowledge gained in building taxonomies might allow the user to more easily determine a way to build sections of the site. Is the idea here to build this using views? Nodequeue? Maybe we should leave that part off and think about some other steps.
Also, for the content contributor scenarios what about some "make a forum post" and "follow up to forum post" kinds of things? That would help test some of the most common things done in Drupal - adding nodes, commenting on nodes - and also indirectly tests out the forum summary page.
knaddison blog | Morris Animal Foundation
ab-fab, here is some follow-up
I don't personally have enough context to say whether using taxonomy system vs. dividing sites up into section is more common or not in most Drupal implementations; I suspect that a large percentage of Drupal sites do just as you say.
In our scenario of the departmental site, however, we wanted to test how users might accomplish the rather more traditional idea of putting pages into 'sections' of a site. I am very confident that 'site sections' in the context of departmental sites, small corporate sites, etc. is a very commonly needed task, maybe the most common.
I don't see this as a case of either/or though. Drupal's ability to dynamically generate content via simple taxonomy calls, views lists, etc points the way to the future for many organizations who are not quite their yet with their old Contribute or Dreamweaver-based sites. OTOT, I think there will, at least in the near term, be a need for most folks to have some 'site section' functionality. Drupal seems to have slightly better support for the former, and there seem to be a variety of approaches to accomplish the latter, which is probably why this task is hard to write.
That being said, I'd personally favor trying to figure something out in terms of the 'site sections' idea, just to see if the webmaster/developer evaluators could figure it out from our very basic directions - maybe we'll learn something about their expectations, needs, etc. On a related note, you'll notice that I added a couple of comments to the wiki page.
Re: contributor scenarios. Feel free to write something up for comments if you like - we do have an adding a node task.
Pfew!
site sections - past, present, future
Historically I think people most commonly created site sections with taxonomy and a single content type. When necessary, they would write custom sql to build a listing of nodes or use a contrib module that provided it's own content type and it's own listing system (even module).
Now I think most "sections" are created with a combination of those techniques AND content types+views and/or nodqueue.
We need (and there is at least discussion for) providing views in core with some really simple UI bits (i.e. single checkbox to "provide listing for this content type at example.com/TYPE" on the admin/content/types/TYPE page). That would make this task dead easy and should be simple enough to do in Drupal7.
So, given all that context, what do we test? :)
I don't know - I'll think about it some more...
--
Knaddisons Denver Life | mmm Chipotle Log | The Big Spanish Tour
knaddison blog | Morris Animal Foundation
I like the idea of having at
I like the idea of having at least one task (and that seems like it) which seems simple but is fairly complex in terms of how Drupal concepts interact. A lot of people try to do 'sections' when they set up a site, especially when new to setting sites up and CMSes in general. Dropping users in at the deep end like this is what actually happens, but we don't normally have the benefits of eye-tracking and the rest to watch that process - by the time you get feedback they've given up or got past that stage.
I'm just catching up with this discussion, but I'd be interested in perhaps testing one group with some minimal direction (not instructions though), and one group with none - and seeing how each gets on. I don't know how this fits in with the methodology though.
And I'd do sections with Panels2 (+ views + custom panes) now, fwiw.
I'm pretty sure we're
I'm pretty sure we're testing with drupal6, so it might not be possible to test many contrib modules unless they are upgraded and stable-ish by then.
Bevan/
What about book pages?
I have always thought the 'traditional' way of creating sections and navigation was using the book module. At least that's the way I used it at first. I remember that when I first started using Drupal I was looking for some way to set up a hierarchical site like the ones I was used to and it seemed like the most familiar way to do it was to create a book and set up my site sections as book pages. And that got me both hierarchy and navigation.
And even if you don't use the book module for site organization, there are many sites that use the book module to create, well, books. Especially educational institutions.
Maybe I missed it, but are we looking at that, or should we?
v. interesting. I didn't
v. interesting. I didn't know people still used book module! :)
Whatever happened to it's modern counterpart outline.module?
Bevan/
The 'about' section of my
The 'about' section of my site is done with book.module - although that started with the 4.5.x version and I have no idea how I'd do it now. It's also much nicer in D6, in case you hadn't seen yet.
Yes - This is how I personally have done it in the past
...and it probably makes sense to take book.module/outline into consideration. Thanks!
Good to hear CCK will be
Good to hear CCK will be included, only a small proportion of sites run only with core, so it makes for a more realistic environment imo. Also, it can only help to have CCK reviewed in this way before it's made its way in.
+1 for 'sections'
IMHO libsys is right: it is better to address the user more in general website terms that they know from their experience. Drupal-specific terms come in later. And don't we want to know how Drupal performs in "everyday" Content handling?
To me this whole usability thing is trying to see Drupal as unbiased as possible, instead of looking at it through "fan's eyes" (what I normally do, no doubt ;) ).
When sitting at the side of people who walk around Drupals back-end or content creation for the first times, this is sometimes quite an eye-opener. They are looking for orientation - for things that appear familiar to them.
I still remember my first steps in Photoshop or Quark Express. O goodie... The whole concept of "pixels" and "selections" in Photoshop kept me totally confused. I just did not get it. In Quark it was the very basic thing of two principal tools - one for a box and one for the content inside it. This keeps you busy for weeks. With every application there are those things that are very basic but proprietary and have to be understood, before you can work fluently with the tool.
It is different for People who know Taxonomy or have already worked extensively with Wordpress, for they somehow know it. But those won't be the ones that will run into trouble using drupal anyway.
Hopefully looking forward to this test being only an initial step, why not keeping tasks as general and basic as possible.
Life is a process
Life is a journey, not a destination
<nudges way into conversation>
Hi folks- I've just read through this post. I haven't been active on it before, so let me introduce myself- I'm Ron, I work with Bevan at CivicActions and will be assisting him in his Season of Usability tasks. At CA I'm one of two IA's (information architects) my background is in usability and IA and I would like to be of service here.
I'm going to read Bevan's http://groups.drupal.org/node/7929 now and would like to add my comments, etc. to it. I also want to integrate seamlessly into this without it being too disruptive so if you have any general comments for me please let me know...
Also happy to chat about it as well (rakanowicz on skype).....
thanks!
Ron
Common Industry Format for Usability Testing--
I would also recommend using this format for developing the test plan and reporting on the results. I've used it before, and while it is a lot of documentation it is also very complete.
See http://zing.ncsl.nist.gov/iusr/documents/cifv1.1b.htm
Just looking at Bevan's wiki at node 7929 I see things that still need to be defined:
what are the usability goals of drupal? easy enough for anybody to use? content published in less than 5 clicks? etc. etc.
are there any baseline user profiles (admin, editors, etc.)?
the task scenarios that we create should reflect the user profiles- e.g. a web editor for a mid-sized company who is responsible for posting content. What types of tasks does this role typically carry out?
what are the metrics around measuring these scenarios? strictly time to complete the task (and if so, how long is too long)? subjective scales (e.g. on a scale of 1 to 5....)? comparing usage to other types of CMS sites?
will eye-tracking capabilities be available?
That's a great resource Ron.
That's a great resource Ron. Thanks for that. There's a lot of work that could go into preparing for these tests and that could help us less-experienced folk do it quicker and more efficiently.
Chad and Cody will be able to answer your questions with details, but there is some information available from http://1help.umn.edu/usability/lab.html
Bevan/
Great Resource Indeed
I'm posting this link (http://zing.ncsl.nist.gov/iusr/documents/cifv1.1b.htm) to our testing resources thread....
Thanks~
Sorry for the delay in response, your questions are actually quite timely. Thanks for jumping in; it's great to see IA folks getting involved.
Yes, this does indeed need to be defined; there will be some activity on this over the next couple of days. Stay tuned.
Nothing more than what you see. Personas might be nice.
Yes, this is where we hope this is heading.
I think we normally use a scale of 1-5, ranking the severity of the given problem; it's a subjective measure but derived through a deliberative process with usability staff.
Oh yes indeed.
Clarification: UMN testing is not my project
Several times now I have seen my name together with 'UMN' as though these tests were my project. I want to clarify that this project is UMN's baby, and more specifically Chad (libsys) and Cody's initiative -- they are the ones that should be attributed for it.
I am merely involved in this project in the following ways;
Thanks Chad, Cody and UMN for providing this opportunity and making this possible!
Bevan/
Thanks to whoever changed
Thanks to whoever changed the author of node/7929 to libsys, I think that will help a lot.
Bevan/
I have decided to work
I have decided to work primarily on node add and edit forms for the SoU project. Next time I do a pass on the testing plan I'll review it to see if it will give me useful data or not.
Also, on that document there is a "Kickoff Meeting: A first pass review of tasks, audiences, this document, etc." on 14 Jan, which is today/tomorrow. What exactly is that? Is there any way we can help?
Bevan/
Tightening focus post-kickoff
Hello all,
As noted earlier, we had our first meeting yesterday with the staff of the usability lab. Their recommendation was that we narrow the focus of our testing to a single user type, rather than splitting our testing sessions between content contributors and site maintainers. They made a good case, and we're inclined to agree.
To that end, I'm going to be making edits to the wiki to reflect a focus on the site maintainer role for this round of testing. I'll preserve the work we've done so far on the content contributor role and tasks in hopes that perhaps a later round of testing at UMN or elsewhere might pick up that thread.
I happened to do some
I happened to do some software usability testing last week as a participant, with a one hour time slot. Looking at the current task for the site maintainer, it seems to me that it's very unlikely they'll get through the whole thing in 90 minutes (which I assume includes the setup and debriefing time as well).
Tasks 1-3 are quite compact, and permissions, content types, fields and taxonomy are a good group of functions to test in tandem - they're usually a necessary part of setting up any new content type. So this seems like a nicely self contained session in itself.
However I think the sections/book module task will be difficult to complete within the same 90 minutes, and it also deals with quite different aspects of Drupal. Using the book module would mean node/add and node/edit forms quite a lot, alongside drag and drop, and blocks (more drag and drop there in a different interface). These are what's missing from 1-3, so it'd be a shame for them to either get missed by lack of time or discarded. node/add node/edit also covers some of what we're missing from content contributor, and although the first task could involve 'testing' the content type, it's not really embedded into that task.
So therefore, I think it'd be good to discuss splitting these into two distinct tasks (content type / sections), if that gives enough testing per-task to be valid, then I'd much prefer it to diluting the detail applied to each stage. If we found in initial testing that the time-frame was over-generous for the two tasks, they could then be fleshed out further (input formats, taxonomy weights, content type display settings, re-ordering/reparenting books etc.)
I think this is where we're heading
We teased out the menu/section portion from the more cck-centric task in the last go-around. Where it make sense in terms of the "narrative" of the scenarios, we could also put lower priority towards the end. If users have time to complete them, great. If not, no big deal.
It's cool that you've gotten a little testing under your belt.
Drupalcon session proposal up
Greetings all,
I've submitted a proposal for a Drupalcon session for a report on the usability testing: http://boston2008.drupalcon.org/session/report-formal-usability-testing-...
Re-publish to front page
Hi, we need to update this post, make it more punchy for above the break and try to get it out this weekend before the D6 announcement.
http://drupal.org/node/204667
Let me know if you are interested in helping.
Kieran
DON'T MISS EARTH'S LARGEST GATHERING OF DRUPAL PROFESSIONALS!
Drupalcon Boston 2008 · March 3-6, 2008
Learn more at http://boston2008.drupalcon.org
Affordable sponsorship packages available
Late to the party?
If I get all the kinks to my schedule worked out, I should be in Minneapolis from the afternoon of Feb 25 to the afternoon of Feb 26. Are any volunteers needed to help out on-site at UMN? Schedule is still iffy (I'll be driving from a wekend trip in southern Missouri), but having something to do besides a meet-up would give me more incentive to getting there.
-Bryan
Bryan Ruby
socPub
Revised scenarios up
I've posted revised text for the scenarios in the testing plan. The revised text retains the core tasks we've worked on thus far, but casts them in what we hope are more user-centric terms which avoid Drupal language. The idea is to aim at a broad task that will seem familiar to our evaluators, rather than breaking it into steps which to closely reflect Drupal's structure. These revised tasks will likely be more challenging for evaluators, but we can always provide assistance toward completion, and the problems they run in to will be instructive.
getting close - last minute tasks?
Hey Chad and Cody - I'm getting excited for the testing as the start date approaches. I feel like we've had great community help at the beginning that tapered off more recently while you guys really brought some focus to the documents in the last weeks. Is there any last push of effort that you'd like from us? Any areas to really review prior to testing on Tuesday?
Thanks!
--
Open Prediction Markets | Drupal Dashboard
knaddison blog | Morris Animal Foundation
same here!
I'll have a bit of time tomorrow, then at least some on-flight reading if nothing else.
Re: Results....
Here you go