Q1. IBM has recently designed a pointing device called the CAT, which IBM researchers claim to be superior to their competitors device, the DOG. According to their claim, users can rapidly perform a variety of pointing tasks such as high precision Computer-aided design editing (D=16, W=1), icon selection (D=16, W = 4), and menu selection (D=4, W = 1).

1. Fitts discovered that the pointing time using a hand held pointing device is a function of the distance (D) and the width (W) of the target, and determined a formula for the index of difficulty as iod = log₂(2D/ W), and for the time to point as t = C₁ + C₂* iod, where C₁ and C₂ are device constants. Use Fitts Law to estimate pointing times for both devices in all of the above tasks (C₁^CAT =0.1, C₂^CAT =0.2, C₁^DOG =0.5, C₂^DOG=0.1). (12)
2. Describe the assumptions of the Fitts Law and argue why IBM can still be right. (8)

a) Computer-aided design editing: (D=16, W=1)

iod = log₂(2*16/1) =5. T^CAT = C₁^CAT + C₂^CAT * iod = 0.1 + 0.2 * 5 = 1.1

Icon selection: (D=16, W=4)

iod = log₂(2*16/4) =3. T^CAT = C₁^CAT + C₂^CAT * iod = 0.1 + 0.2 * 3 = 0.7

Menu selection: (D=4, W=1)

iod = log₂(2*4/1) =3. T^CAT = C₁^CAT + C₂^CAT * iod = 0.1 + 0.2 * 3 = 0.7

b) The assumptions underlying the Fitts law are:

• One-dimensionality
• Moderate distances (D)
• Gross arm movement

Hence, IBM could still be right because the only type of task users have to spend more time to point is when D = 16 and W = 1, (i.e long distance to object and high precision). Besides all the above tasks occur typically in two-dimensional space.

Q2. Critique the problems in interface design of the following dialog box. Describe 10 points in your critique precisely with specific examples from the dialog (20).

• Window title and dialog title mismatch
• Button text styles are different
• Button coloring inconsistent
• Button alignment not appropriate
• Too many words in a button
• Assistant help is too anthropomorphic
• Age textbox unaligned
• SSN boxes inappropriately separated
• Text style for Mr / Mrs are different, Ms missing.
• Grouping for Personal Information box inappropriate
• Format for Name, Date of Birth unknown.
• Age and Date of Birth fields ambiguous
• Age slider is horizontal rather than vertical

Q3. Find the problems in each of the following error messages produced by a sour-code compiler. Briefly state your reasoning and suggest a better message. (15)

1. SYNTAX ERROR (A. Too general, unspecific. Alternative: EXPRESSION FORMULATION IN LINE 13 DOES NOT MATCH GRAMMAR)
2. OUT OF RANGE. (A. Unspecific. Field and appropriate range unknown. Alternative: ACCEPTABLE AGE FIELD RANGE IS 18-55)
3. INCORRECTLY FORMED RHS OF EXPRESSION. BAILING OUT ON LINE 16. (A. Abbreviation unknown. Unspecific. Wording hostile. Alternative: RIGHT-HAND SIDE OF EXPRESSION ON LINE 13 DOES NOT MATCH GRAMMAR)
4. ERROR STRG12, CONSULT THE MANUAL. (A. Unspecific. Manual reference should be avoided. Alternative: STRING NAME SHOULD BE LESS THAN 5 LETTERS.)
5. ILLEGAL OPERATION. PROCESS KILLED!!! (A. Wording too hostile. Unspecific operation. Alternative: OPERATION ENDED DUE TO UNKNOWN COMMAND: LLS)

Q4. A group decision system is being built over a network, which has inherent delays of due to network delays. Make an argument of the effects of system response times from 0.1 to 100 seconds on each of the following tasks: (12)

1. System login, password check (A. Expected delay typically ranges from 1-5 sec. Response times less than this would make users uncomfortable regarding the security of the system. Response times larger than these would decrease satisfaction. Variability in response times would make the user think about the system load.)
2. Synchronous group editing of a document (A. Expected delay typically ranges from 0.1 - 1sec. Response times between 1-5 seconds would decrease user satisfaction considerably as in editing users expect real-time performance. For repetetive tasks such as editing variation in response time also makes the user uncomfortable using the system. Depending on the editing task (e.g. text, image, graph, etc.) users expectation of response times vary.
3. Asynchronous critiquing of other participants' work (A. This is practically a problem solving task thus expected delay typically ranges 1sec - 10sec. Delay much longer decrease user satisfaction.)
4. Voting on serious issues. (A. Expected delay typically ranges 10-100sec. Previous experience with the system may invoke suspicion on the user if the response times tend to vary.)

Q1. A window manager is a software package that helps the user control the layout of multiple windows on the screen. Proponents of tiled window layout strategies claim that tiled window managers improve novice user performance (speed and accuracy) for most task complexities (small, medium, large number of windows) for search tasks, because less window housekeeping activities are performed. Proponents of overlapping window layout strategies claim that expert users have more capability in window arrangement and thus yield higher user performance. However, it is typically an issue when designing an application interface as to which strategy to follow.

Assuming substantial resources, design a controlled experiment to resolve this issue.

1. Describe the hypotheses of the experiment. (5)

1. Although novice users perform search tasks faster with less number of errors in tiled window layout strategies, expert users perform faster in overlapped windows. However, as the number of windows increase tiled windowing strategies yield faster and more accurate performances.)

1. Describe the independent and dependent variables and their treatments. (6)

1. IV's: 1. Task complexity (# of windows), with 3 treatments: S (2), M (6), L (18)

DV's: 1. Task completion time

2. Number of errors

1. Describe the subjects, tasks and controls. Be very specific. (9)

1. Subjects: 60 Subjects will be distributed into four groups: Novice-Tiled, Novice-Overlapped, Expert-Tiled, Expert-Overlapped, with 15 in each group. Novice users in both categories should have no exposure to either of the systems. Expert users should be using their systems for more than 5 years.

1. Briefly describe which statistics should be used and why. (5)

(A. two 2x2x3 ANOVA designs for both speed and accuracy, since there are 3 IV's with 2, 2, and 3 treatments, and 2 DV's.)

2. Predict the results of the experiment and discuss them in a graph. (8)