Information Design Using Card Sorting Page II

Preparing the survey

The front page

The general style of the survey should be kept light and friendly. This is an opportunity to show the users that you have their interests at heart. The front page introduction can be used very effectively to deliver this message.

Some information should be provided on how to complete the survey. More importantly, however, the introduction can "market" your project. By highlighting the participation of the actual users in the design of the system, you can go a long way towards reducing or eliminating the normal resistance to change. (This should be the first of many times that you involve the users.)

As a side note, if you have access to someone with desktop publishing skills (either within your team or elsewhere), you may wish to consider "prettying up" this document. If a friendly but professional style can be presented, it is likely to be received more warmly by the readers.

Usage survey

In preparing for the session, you have identified a list of potential content. While the aim of the session is to structure these topics, it is also useful to know how frequently the information is being used at present.

This can be obtained via the use of a simple survey which asks the users to rank how frequently they use each item.

When you provide the survey to your participants, it should be emphasised that you are interested in how frequently the information is used, regardless of the source. The participants should therefore consider all ways of obtaining the information, such as:

• paper manuals
• online
• colleagues, supervisors or managers
• personal notes, resources, etc.
• computer systems

Regarding the format of the survey itself, we have found a simple tabular layout works quite well:

How frequently do you use these topics (1 = very frequently; 5 = never)
1.	Internal phone directory	1 • 2 • 3 • 4 • 5
2.	Staff leave allocations	1 • 2 • 3 • 4 • 5
3.	Events	1 • 2 • 3 • 4 • 5
4.	Tender information	1 • 2 • 3 • 4 • 5
5.	Sales histories	1 • 2 • 3 • 4 • 5
6.	...	1 • 2 • 3 • 4 • 5

We've chosen a very simple scoring system (from 1 to 5), without specifying exactly what is meant by each number. While this may lead to some variation between user's responses, the survey is only really designed to draw out the 1's and 5's.

Note the number against each topic: the use of this will become apparent later, when you document the outcome of the session.

Finally, leave some blank lines at the end of the list for the participants to add any topics they believe have been missed.

General feedback

A small survey at the end of the list of topics provides the opportunity to obtain further information about the users and their needs. While it is important to keep this section brief, it is nonetheless possible to ask questions such as:

• What is your job role?
• How long have you acted in this role?
• What are the main methods you use to find information?
• What are the most common pieces of information that you use?
• What other information would you like to have online?
• Do you have any other feedback or suggestions?

The first two questions will help you to correlate usage with levels of experience. The second two questions provide a double-check on their answers to the survey, while the last two questions are a common courtesy.

Creating the cards

Each topic on your list should be written on a index card (sometimes called a filing card). Make sure that you also write the topic number in a corner (keep it small, so as not to distract the users from the actual topic titles).

Regarding the right cards to use, use whatever is on hand, as long as it is a reasonable size. If you need to purchase some cards specifically for this exercise, we would recommend 76x127mm cards (or thereabouts). These are large enough to be easy to write on and handle, but small enough so that you can fit a fair few on a table.

It is also possible to do a card sorting exercise with Post-It notes, but is our experience that cards are a bit more durable, which makes them easier to handle.

Note that you will probably need at least a hundred cards for each session, if not more. (This is once you include the additional cards written up during the session.) So buy in bulk, to avoid running out at a vital moment.

As a final note, it should not be necessary to highlight the importance of writing clearly and readability on the card. Please, no last minute scribbling!

Selecting the participants

The attendees at the card sorting session must be the actual end-users of the system you are building. This cannot be emphasised too strongly.

There is no point in doing card sorting with a group of managers: they are not the end users. (Unless of course the system is designed specifically for management use.)

Where there are a number of distinct groups of users, hold one session for each. Each session must only have a single group of users, to ensure that the structure is representative of their specific needs. This allows you to compare the results from the different user groups and identify the differences and similarities in their requirements.

A card sorting session must have at least four participants. Any less than this and there is insufficient discussion and disagreement. Without this healthy debate, it is hard not only to come up with meaningful results, but also to finish the session at all.

More than eight users, and the group becomes more difficult to manage. If nothing else, it is hard to find a table large enough so they can all fit around it. Ideally, all the participants can reach the cards on the table, although this is not always necessary (usually one or more users will "take charge" of placing and moving the cards).

Preparing for the session

You will need to gather a number of items, in preparation for the card sorting session:

• the cards, and enough survey sheets for all the participants

• blank cards

  These will be required for both for adding missed topics and for writing the heading cards (see later for a description of heading cards).

• a thick marker pen (for writing on the heading cards)

• a number of rubber bands (for keeping the cards in order at the end of the session).

• A Session Participation Form

  This is filled in by the users at the session (you may not know who they are going to be in advance), giving their names and job roles.

• A notepad and pen, for recording the participant's verbal feedback.

You should also make sure you understand what is meant by each of the cards (you will almost definitely be asked), or ensure that you have someone sitting in on the session who knows the subject matter well.

Running the session

Introducing the session

Before starting the session itself, it is important to introduce yourself and to provide some information about how the session will work.

Walk the users through the card sorting process, and show to them how each stage works. It is important to highlight to the users that they should organise the cards in a way that works for them. Emphasise that they are to forget everything they know about how the information is currently organised - this is their chance to structure the content to meet their needs.

This is also an ideal opportunity to discuss the project as a whole. Assume that the users have been told nothing about the project (this is commonly the case), and start by discussing the overall goals and aims. Fill the users in on what will be delivered by the project, along with a realistic outline of the project timetable.

The card sorting session is a not-to-be-missed chance to 'market' the project. The more enthusiastic the users are, the lower the resistance to change when the system is implemented.

Filling in the survey

After you have introduced the session, hand around the survey to the participants. Explain how to correctly fill in the survey, and then give them enough time to comfortably complete it.

Be prepared to answer questions about the meaning of the various topics that you have identified. If you don't have the knowledge, ensure that a subject-matter expert is present at the session who can field these questions.

Also ask the users to write down any topics they think have been missed from the list. (Some empty rows should be added to the end of the survey for this purpose.)

Having the users fill in the survey at the beginning of the session may seem like a waste of time. There are several important benefits, however:

• It familiarises the users with the list of topics, which will come in handy when they are sorting the cards.
• It allows the users to ask questions about the survey, which helps to eliminate sources of confusion.
• It ensures that 100% of the surveys are completed and returned to you.

While the users are filling in the survey, take the opportunity to sit back and relax: all the hard work in a card sorting session is done by the users.

Go to page: 1  2  3 

Author
James Robertson is the managing director of Step Two Designs, an online development company based in Sydney, Australia. James specialises in XML development, information management and systems design.