Setting Up The IA Framework
The emphasis for this week is on mounting the Information Architecture (IA) for the project. Heeding Mr.Reddy’s advice for an improvisation of various IA techniques to encompass an maximum coverage on user’s partiality, we decided to go for a hybrid of our own IA style: Improvising cardsorting + persona-focused, observation + common task analysis + storyboard. Why a person-focused cardsorting, you might have asked? Because our idea is a novelty, we really wanted it to zoom into our target users, hence person-focused cardsorting would make our IA more target-focused.
Our IA Approach
* Cardsorting
We did a preliminary analysis, brainstorming all possible attributes that could come out from our head, that would connect with our carparking theme, effectively coming up with a content inventory of approximately 40 odd items. We then filter the inventory list by striking off either duplicated or highly similar items. With that, the list was shortened further and eventually identified 30 main items for users to perform the cardsorting on.
Needless to say, cardsorting itself is a useful tool that is exceptional in identifying how users group data, as well as how they would make the data flow. As what was mentioned previously, we are going for an improvised approach in gather user groupings of data: Targeted persona card sorting. As such, we will be approaching users directed from our Advanced and Novice personas, but voiding Anti-Users since instinctively due to credibility issue; they wouldn’t be keen on our proposal anyway, hence their results might be highly biased and pre-disposed. With all the settings in placed, we arranged for 12 sessions of cardsorting, a mixture of online (generic, overview) and offline (targeted, focused) sessions.
For the offline (targeted) group of users, we have actually prepared the materials: physical sets of 30 cards and coloured stickers beforehand, for them to physically perform the cardsorting process. We begin the sessions by briefing them, highlighting our intentions of this test, introducing the tools (cards) and showing them a quick demo on how to do it: grouping by pasting colour tabs onto item cards. For this specific group, we also kept a close eye on their actions, observing their expressions and reactions when card sorting. We felt that it was necessary to study the context in which a task is being performed by observing and experiencing tasks as from that, we can identify which are the tasks that are really intuitive and which requires closer thoughts.
As we hope for more user feedbacks for identifying common groupings between user, we decided to branch out our IA to more people via the online group. We discovered a great tool, that’s free of charge, interactive and helpful in results consolidation into excel: the online cardsorting exercise via Optimalsort.com. We setup our cardsorting questionaire on OptimalSort.com, including some introduction and highlights that would guide them through on our exercise. They are then required to name classification categories, of how they would group the items by dragging and dropping into these categories. Through this online group, we would be hopeful of settling the disparities that could arise between diff point of views in classifying certain items, which would indirectly help in better establishing the majority’s common perspective.
* Common User Task Analysis
The online results were compiled similar to that of offline card sorting sessions. The OptimalSort results consolidation tool made life simpler for us, so what we eventually did was to use eye-ball analysis to identify common groupings between users. Out of which we discovered readily the following groupings:
a. Carpark Info
b. Booking Info
c. Member Details
d. Statistics
e. Viewing of live Info
f. Search
g. FAQ
With the above common groupings identified, it made life easier for us as we are able to analyze further and identify the common user tasks associated with them. Identifying common user tasks are essential because they would be what that what defines the system and associates the user with the system. As such, the 6 key user tasks, based on what we unearthed in the previous sections are:
- Search
- Log In/Out
- Registration
- View Info
- Website Booking
- SMS Booking
* Storyboarding
These 6 common user tasks would be pivotal in our next step: Storyboarding of information workflows. With common groupings and common user tasks in place, we analyzed all possible workflows, refined and derived a flow chart of information/procedural flow for each common user task. Each flow chart would be analogous to that of a procedural cycle. As such, we have to down key details such as process descriptors, user’s interaction, process input and output, so as to fully document the storyboard. The storyboards (an example shown below for booking) would be particularly vital in the next phase: prototyping, which we are really excited about.
Reflections
IA is probably one of the dullest phases that we had anticipated before its arrival. As mentioned in the previous post, being students in Computing, we rarely seek feedbacks from users, and it had somewhat affected us in our entrance into this IA phase. We (Alvin and I) often felt that IA is redundant, that we can also decide for the users, such as ‘Now these are the links, these are exactly what you have to do to reach this step’. We wanted to organize the content for the website or application directly, often having the urge to go straight into doing our way, but were often reminding ourselves to focus on users. Hence, we thought it would be a ‘formality’ procedure. But we were proven wrong. We were somewhat surprised by the outcomes of cardSorting. Some of the classifications are really what we never thought of, and it seems to be such a good remedy for a few of our dilemmas. What we would have done, is really different from what the users have provided us. It really brought us another perspective, which is really more ‘accurate’, since it would be really what the users one. This provides valuable input for understanding the natural categories people have for the content before we begin the design of the new structure.
One point that we constantly remind ourselves is to observe the users, and it’s a really useful technique for our cause. The notes that we took down while observing our offline users helped a great deal in our decisions for identify common user’s task. Furthermore, we realized, as programmers, sometimes our point of view is confusing for users. This is very true indeed, after observing and comparing how users group and how we will do it our way. We are also thankful that we sought the feedbacks from the online group of users, as it really aided us aplenty in concurring the disparities arising between the offline targeted users. We achieved greater insights into what the common groupings will be. However, we sincerely believe that for massive projects, we really need to conduct and consider the inputs from users (targeted):
Reflection quote of my own: Every users are unique; everyone have different ways of grouping, but most have something in common. Hence, even when it’s impossible to please all, we should at least try to please the common majority.
