The purpose of this set of exercises is to help you practice using R. I will provide a set of data that could realistically have come from user studies. Your job will be to transform the data into an appropriate form so that you can run the right inferential tests to help you make decisions about the data. Once you have done this, you will also practice writing sentences that clearly articulate the procedure that was undertaken.
Much of this is grunt work, but see the links from the previous worksheet to find out how to format the data properly as needed.
You can work in small groups or as a class to complete these tasks. (If it is easier, perhaps use the data only to one decimal point.)
In 2002, OS X introduced a taskbar that defaults to an interaction where the icon actually expands in size. The idea is that the expanded size allows for the icons to be clicked faster.
You decide to test out this theory by having participants click on alternating targets. The independent variable is whether the targets are expanding (as in OS X), or staying the same size. You have them complete this alternating task 50 times each before you have them do the other condition. You decide to use a between-subjects design, and recruit 32 participants to complete this task. Using the data below, determine whether the expanding targets actually results in faster performance.
| Participant | Static Target | Expanding Target |
|---|---|---|
| 1 | 100.89876 | |
| 2 | 98.27522 | |
| 3 | 102.77773 | |
| 4 | 100.42039 | |
| 5 | 99.42488 | |
| 6 | 103.92898 | |
| 7 | 101.21053 | |
| 8 | 95.69733 | |
| 9 | 104.07195 | |
| 10 | 105.83715 | |
| 11 | 100.85767 | |
| 12 | 103.08823 | |
| 13 | 96.72939 | |
| 14 | 99.24429 | |
| 15 | 98.54360 | |
| 16 | 104.32236 | |
| 17 | 92.59521 | |
| 18 | 100.05848 | |
| 19 | 99.49896 | |
| 20 | 94.88110 | |
| 21 | 97.50323 | |
| 22 | 101.74019 | |
| 23 | 96.24544 | |
| 24 | 95.25246 | |
| 25 | 96.92449 | |
| 26 | 95.00141 | |
| 27 | 95.55145 | |
| 28 | 97.15211 | |
| 29 | 93.57841 | |
| 30 | 97.09549 | |
| 31 | 90.91585 | |
| 32 | 93.84789 |
Ref: McGuffin, M & Balakrishnan, R. (2002). Acquisition of expanding targets). In Proceedings of CHI 2002, 57-64.
EdgeWrite is a character-based input technique for mobile devices. It was first implemented on a PalmPilot (which uses a stylus pen) where a physical template with a square-cutout was used to restrict movement of a stylus. It provides a different kind of "graffiti", but allows people to "throw" the stylus in to the various corners. This allows text entry to happen quite quickly. (See a video here.)
To evaluate their solution, a study is conducted to compare EdgeWrite versus standard Graffiti text entry. The researchers recruited 10 participants, and this was a within-subjects design (participants completed all tasks with each condition). The conditions were: EdgeWrite, and Graffiti. The researchers designed 3 separate sentences that needed to be completed, and each of these three sentences were written in both conditions. The collected data includes the words-per-minute as measured for each task, as well as the number of errors that were committed for each task/condition pair.
EdgeWrite Data
| P | wpm1 | err1 | wpm2 | err2 | wpm3 | err3 |
|---|---|---|---|---|---|---|
| 1 | 9.94 | 2 | 6.36 | 2 | 6.97 | 2 |
| 2 | 6.97 | 2 | 6.24 | 2 | 8.29 | 3 |
| 3 | 6.81 | 2 | 6.83 | 3 | 7.23 | 3 |
| 4 | 3.66 | 2 | 5.77 | 1 | 7.05 | 3 |
| 5 | 6.43 | 2 | 7.37 | 2 | 4.38 | 3 |
| 6 | 4.29 | 3 | 5.51 | 3 | 5.90 | 1 |
| 7 | 4.18 | 4 | 4.90 | 2 | 6.94 | 0 |
| 8 | 3.28 | 1 | 6.87 | 1 | 7.83 | 2 |
| 9 | 6.97 | 1 | 8.40 | 2 | 7.22 | 1 |
| 10 | 6.31 | 1 | 5.49 | 1 | 6.08 | 1 |
Graffiti Data
| P | wpm1 | err1 | wpm2 | err2 | wpm3 | err3 |
|---|---|---|---|---|---|---|
| 1 | 9.99 | 4 | 6.85 | 4 | 6.32 | 4 |
| 2 | 6.60 | 4 | 8.26 | 5 | 9.27 | 2 |
| 3 | 4.43 | 4 | 8.17 | 5 | 9.02 | 3 |
| 4 | 8.22 | 4 | 8.50 | 4 | 7.40 | 4 |
| 5 | 7.49 | 3 | 9.04 | 4 | 3.43 | 4 |
| 6 | 5.87 | 4 | 7.33 | 4 | 5.15 | 4 |
| 7 | 5.87 | 3 | 9.68 | 5 | 7.24 | 5 |
| 8 | 8.66 | 3 | 7.85 | 2 | 6.50 | 2 |
| 9 | 6.23 | 4 | 7.41 | 2 | 8.82 | 5 |
| 10 | 7.39 | 3 | 5.66 | 4 | 8.06 | 4 |
Ref: Wobbrock, J.O., Myers, B.A. and Kembel, J.A. (2003). EdgeWrite: A stylus-based text entry method designed for high accuracy and stability of motion. In Proceedings of UIST 2003, 61-70.
The bubble cursor is a target acquisition technique based on area cursors. The bubble cursor improves upon area cursors by dynamically resizing its activation area depending on the proximity of surrounding targets, such that only one target is selectable at any time. Empirical studies show that the bubble cursor significantly outperforms the point cursor, and that bubble cursor performance can be accurately modeled and predicted using Fitts' law. You can see a video of BubbleCursor, or try it out.
Since others have already demonstrated that BubbleCursor is faster, the question you are trying to answer is: do people like BubbleCursor (BC) more compared to a standard interface (S)? You allow your (10) participants to try out the interfaces, and then ask them two separate likert-style questions (scale of 1-7):
| Participant | SEE (BC) | SEE(S) | EASE(BC) | EASE(S) |
|---|---|---|---|---|
| 1 | 3 | 7 | 6 | 5 |
| 2 | 4 | 5 | 4 | 4 |
| 3 | 3 | 8 | 6 | 5 |
| 4 | 5 | 6 | 4 | 5 |
| 5 | 2 | 6 | 5 | 6 |
| 6 | 4 | 6 | 7 | 6 |
| 7 | 4 | 5 | 6 | 5 |
| 8 | 5 | 5 | 5 | 6 |
| 9 | 3 | 4 | 4 | 5 |
| 10 | 3 | 4 | 4 | 6 |
Ref: Grossman, T. and Balakrishnan, R. (2005). The bubble cursor: enhancing target acquisition by dynamic resizing of the cursor's activation area. In Proceedings of CHI 2005, 281-290.