Homework 1: Contextual Inquiry and Design

For 17-770/46-863: Introduction to Human Computer Interaction for Technology Executives

Assigned Aug 31; Due Sep 7

(see how to submit homeworks on main homeworks page)

Here are the Device Choices made so far.

The first step in creating any user interface (or actually, any product at all), is to know the user, and in particular, to know the tasks and requirements for the product. One way that HCI professionals have found useful for getting a better understanding of users is through a process called Contextual Inquiry and Design, which you will be performing in this assignment on the Device you selected in Homework 0. Here is the list of assignments of people to devices.  Let me know if there are changes.

Normally, for contextual inquiry, you would just go and watch users performing their tasks in their normal course of business or home life. However, that is not feasible for the devices we selected. Therefore, we have to use as close a substitute as possible:

PART A: The first part of this assignment is to describe one or more representative tasks that users would perform with your device, that will exercise the parts of the device you want to create. You should come up with enough tasks that an expert doing all of them will take about 2 minutes to perform. You must describe the tasks by the function and results and not by the steps to perform them. Therefore, the task descriptions should work no matter what the user interface looks like (since you haven't designed a user interface yet!). For example, you might say "the clock should end up displaying the current time", rather than "push the HR and MIN buttons to set the time".

• The deliverables for PART A are the description of your device (what kind of device and what parts of the interface to it), and
• The script of exactly what you will say to instruct the users on which tasks they should achieve (remember, the instructions should tell them what to do, not how).

PART B: Next, you will perform a Contextual Inquiry on a user trying to perform your task(s) on an existing real device. Ideally, you would test multiple people with a variety of skill levels, but since we do not have time for that, you should pick one user who is not experienced with the device. Preferably, this should be a person who is not at all familiar with type of device, and will have a lot of trouble using it. If you cannot find anyone who has trouble with the tasks, then you need to (1) find a user who is less skilled, (2) switch devices to sometime more complicated, or (3) invent harder tasks to perform and revise PART A. In general, elderly people tend to be less skilled with modern devices, so finding someone over 50 with little technical experience, might be appropriate. It would not be surprising if your inexperienced user takes a long time to perform your tasks. If it looks like it will take more than an hour, then you probably want to provide some help.

You might want to videotape your sessions. The university has some video-cameras that it loans to students, but you have to plan ahead and get a CFA lending card. Here are the details: http://www.cmu.edu/computing/clusters/equip_lending.html. Let me know if you find other ways to borrow cameras.

In performing the Contextual Inquiry, you should read your script from PART A, and record all the user's actions, preferably with a videotape.

• The deliverables for PART B will be a brief description of the user (note: not their name, but instead their experience and demographic properties), and
• A summary of what the user did and said, and what you did and said. If at some points you have to help the users, because they cannot figure out what to do, that must be included in your transcript. It is not necessary to write down every word that the user says, just what is interesting and useful. Be sure to write down all actions on the device, whether correct or wrong.

PART C: Now construct Contextual Design models from the transcript. This must include an artifact model, which might be an annotated photograph or photographs of the actual device (or just drawings) showing the various buttons and screens that the user used. Annotate this with any breakdowns caused by the physical interface (e.g., if the user could not understand a label). You must also make a sequence model showing the sequence of steps that the user made. This will be close to your transcript, but the sequence model will not include what you told the user, and it will include specific breakdowns relevant to the steps. You must also make a flow model showing the flow of information between the user and various parts of the interface. Include any breakdowns relevant to the information flow. Finally, create a cultural model showing any influences on why and how the user chose to use the device. The other kind of model is a physical model, which you are welcome to make if it seems useful.

• The deliverables for PART C are the four (or five) models, with their annotations. You can draw the models using any drawing tool you like, such as Visio, Illustrator or even PowerPoint. Or, if you are very neat, you can even just draw them by hand on paper (however, we will deduct points if it is not legible). If you draw them by hand, it is OK to submit on paper in class, or else you can scan it (if the scan is legible).