UBC Computer Science - Fall 2003

Research Methods in Computer Science: CPSC 590

HCI Lectures: Dr. Joanna McGrenere

HCI Assignment #1



Date Posted: Friday, September 26th

Date Due: Monday, October 6th, in class


1. With respect to the McGrenere and Moore paper [MM]:

a) Which of the research strategies documented by McGrath best describe(s) the strategy used by MM? Justify your answer.

b) Which of the measures documented by McGrath  best describe the measures used by MM? Justify you answer.

c) Comment on the internal, construct, and external validity of MM's study.

d) McGrath claims that no one study can achieve generalizability, precision, and realism.

(i) Which one of these three desirable features was least achieved in  MM's study? Justify your answer.

(ii) Briefly describe one possible study design that would investigate the same phenomenon/behaviour as MM's study but that would maximize that one feature. One paragraph is all that is needed.


2. With respect to the McGuffin and Balakrishnan paper [MB]:

Answer the same questions as above, replacing MM with MB.


3. This question is an exercise in experimental design.

Joe is a graphical designer and an expert computer user. He has been using modern graphical user interfaces for many years and has also casually observed others working on computers. He is frustrated by many things, but one thing in particular is the time it takes to select items from the menu system. For example, while working on graphical image he finds that he is constantly having to re-position the mouse from the canvas area of the screen up to the menubar at the top of the screen, pop up the appropriate menu, and then select the desired menu item. While this may not seem cumbersome, it is cumbersome to Joe who performs menu selections  many many times per day.

Joe finds that context menus (those that pop up with the right mouse button while positioned on the canvas) are only a minor improvement. While context menus do save Joe from having to re-position the mouse, there are only a few menu items available in these menus and Joe still needs to attend to all the options in the menu to make his selection. This diverts attention from his primary task which is to produce a graphical image.

Joe believes he has a solution to his menu woes, namely, gesture menus. The following attributes describe a gesture menu:

Figures 1, 2, and 3 illustrate how gesture menus work.


Figure 1: In menu mode the user selects the Groceries submenu and then Fruit and Veg menu item.


Figure 2: The left part of this figure shows what would appear on the screen when Clipboard C is selected in menu mode. The right part of this figure shows what would appear on the screen for the same selection in gesture mode.

Figure 3: A sample 2-level menu hierarchy and the relative set of gestures.

Joe suspects that for menu items that he selects sufficiently frequently, he will be able to learn the gesture over time and thus perform the menu selection without having to use the visible menu. For those items that he selects relatively infrequently, he suspects that using gesture menus in menu mode will still be faster than using a traditional linear menu in a menubar. But Joe is not sure how novice users would manage with gesture menus.

Joe is not an expert in HCI or experimental design and needs your help to test whether or not his suspicions are correct. Your job is to design an experiment that will compare gesture menus to regular linear menus. In your design, take care to address the following issues:

In addition, please comment on the following with respect to your design: