HCI401 Introduction

A brief overview of HCI401 module.

Transcript

1. 1 There are two industries that refer to their customers as users…. Computers and drugs. Libraries refer to their customers as ‘patrons’.
2. 2 HCI Introduction   Definition Human-computer interaction is a discipline concerned with the design, evaluation and implementation of interactive computing systems for human use and with the study of major phenomena surrounding them. (CDG/ACM 1988)   Interpretations of human and machine.   Multidisciplinary   Computer Science   Sociology and Anthropology   Psychology   Industrial Design  
3. 3 Another definition: “Human-computer interaction is concerned with: the joint performance of tasks by humans and machines; the structure of communication between human and machine; human capabilities to use machines (including the learnability of interfaces); algorithms and programming of the interface itself; engineering concerns that arise in designing and building interfaces; the process of specification, design, and implementation of interfaces; and design trade-offs. Human-computer interaction thus has science, engineering, and design aspects.”
4. 4 Why study HCI?   Interface is major aspect of software development   50%-60% of overall effort, up to 90%   Software is programmed once; but endured by users for lifetime   Cost savings1   Marketing   Bad interfaces cost money and lives.   Therac-25   Phobos I   USS Vincennes (story)   London Ambulance Service, 1993   Others (incl. Cali Crash)   User interfaces are difficult!   Hard theory elusive   People are varied, unpredictable and moving targets2 1 www.useit.com/papers/guerrilla_hci.html 2 Read Landauer, T.K. (1991). Let’s Get Real: A Position Paper on the Role of Cognitive Psychology in the Design of Humanly Useful and Usable Systems. In J.M. Carroll (Ed.), Designing Interaction, Cambridge University Press, pp. 60-73  
5. 5 Serious Life-Threatening Errors   Analysis of transcript of 911 call announcing bomb in Centennial Park at Atlanta Olympics indicated that 20 minutes were needed to call dispatchers   Dispatch system required an address for Centennial Park   Dispatch operators could not find anyone who knew address   Bomb was set to go off 30 minutes after call   Airline crashed in 1996 into a mountainside in Colombia killing all aboard   Pilot typed in “R” rather than full name of airport   Guidance system took first airport in the list beginning with “R” which was the wrong airport   Plane ran into mountain   Draken drop-tanks
6. 6 How not to do good design!
7. 7 Why is HCI Important?
8. 8 Ballot Problems   The instructions are misleading   Use of the phrase “vote for group” is misleading   Should say “vote for one”   Instructions only on lefthand side   Implies righthand side is different   The interleaving of holes is misleading   Only the President page has this layout   Other offices are one per page (with appropriate instructions)   The sample ballot looks different   No holes – the source of the problem   Did not lead to complaints
9. 9 Other Ballot Issues   No user testing   No training for ‘chad’ removal   People vote infrequently   Have to re-learn the system each time   Rushed, uncomfortable circumstances   Palm Beach Demographics: Elderly
10. 10 An Informal Usability Study Barbara Jacobowitz, CHI-WEB, Nov 10, 2000   “I was able to print 10 different sample ballots from various sources. Last night, I ran them all by my mother (81) and a group of her friends (70- something to 80's). All are bright, literate, and none are legally blind.   They did reasonably well on 9 of the ballots. On one, 6 marked it incorrectly and didn't realize it, 2 did it correctly, but very slowly, and 2 had to ask me what to do. Guess which ballot it was?.”   Summary of a more formal study of punch-card voting:   http://www.osu.edu/units/research/archive/votedes.htm
11. 11 More evidence that the ballot is misleading (New York Times, Nov 9, 2000)   Percent of ballots thrown out in Palm Beach County for the error of "overvoting" on Presidential candidates: 4.1% (19,120)   Percent of ballots thrown out in Palm Beach County for the error of "overvoting" on Senatorial candidates: 0.8% (3,783)   Percent of ballots thrown out in Sacramento County (CA) for the error of "overvoting" on Presidential candidates: 0.29% (1,147)   Percentage of (unofficial) re-count votes in Gore's favor: 70% (2,520)   Percentage of (unofficial) re-count votes in Bush's favor: 30% (1,063)   Story
12. 12 Jacques Carelman’s Coffeepot for Masochists “Why do we put up with the frustrations of everyday objects…” (Donald Norman)
13. 13 POET (1988) – Donald Norman “The human mind is exquisitely tailored to make sense of the world. Give it the slightest clue and off it goes, providing explanation, rationalization, understanding. Consider the objects - books, radios, kitchen appliances, office machines, and light switches - that make up our everyday lives. Well-designed objects are easy to interpret and understand. They provide visual clues to their operation. Poorly designed objects can be difficult and frustrating to use. They provide no clues - or sometimes false clues. They trap the use and thwart the normal process of interpretation and understanding. Alas, poor design predominates. The result is a world filled with frustration, with objects that cannot be understood, with devices that lead to error.”
14. 14 Therac-25   Cancer radiation therapy machine (1986)   PDP-11, VT100   Two modes of operation (unique!)   x-ray   maximum power through thick metal plate   radiate tumors inside the body   electron beam   low power, focussed beam   localized coverage
15. 15 Therac-25 - Design   As designed   type “x” (x-ray)   high power beam (25,000 rads), metal plate in place   type “e” (electron beam)   low power beam (200 rads), metal plate retracts   type “b” (turn beam on)
16. 16 HCI is about Minimizing Errors Story: http://courses.cs.vt.edu/~cs3604/lib/Therac_25/Therac_1.html Recently repeated X Up arrow E Enter Beam Ready B Malfunction 54 Clear
17. 17 Therac-25 - Use   Type “x” oops! <edit mode> “e” “b”   The result   high power beam, metal plate retracted   Malfunction 54 displayed   type “b” Malfunction 54 “b” Malfunction 54   patient (and others) died of radiation poisoning
18. 18 Therac-25 - Problem   Input sequence never tested!   user made entry, noted error, corrected error   a very common event!   Feedback not clear   Malfunction 54 may mean something to the designer, but not to the user   Concomitant factors   Audio intercom broken   Video hookup disabled on the day
19. 19 HCI is about Minimizing Errors   60% of aircraft accidents are due to “human error”   60% of undesirable manufacturing outcomes are due to “human error”   Think about the examples we just covered   These aren’t “human error”   They are DESIGN ERRORS!   The goal of HCI is to minimize errors by minimizing design errors
20. 20 Interaction design in business   Increasing number of ID consultancies, examples of well known ones include:   Nielsen Norman Group: “help companies enter the age of the consumer, designing human-centered products and services”   Swim: “provides a wide range of design services, in each case targeted to address the product development needs at hand”   IDEO: “creates products, services and environments for companies pioneering new ways to provide value to their customers”
21. 21 User Interface Economics   Good user interface may result in:   Increased productivity   Reduced training costs   Preventable user errors   Reduced employee turnover   User satisfaction   Higher quality products produced
22. 22 Show me the money!   Australian insurance company1 redesigns its application forms to make customer errors less likely   Cost of usability project <A$100,000   Annual savings A$536,023   Old forms resulted in average of 7.8 errors per form, needing >1 hr. per form to repair   Boeing 757 flight deck2   Improved interface allowed operation by 2 rather than 3 pilots   Reduction from 3000 words to 150 words of instructions needed to operate a paging device2   Navigation features accounted for 61% of the variance of sales for one e-commerce study3 1 Fisher and Sless, (1990). Information design methods and productivity in the insurance industry. Information Design Journal 6,2,103-129 2 Harris, D.H. (1984) Human factors success stories. Proc. Human Factors Society 28th Annual Meeting, 1-5 3 http://www.ascusc.org/jcmc/vol5/issue2/lohse.htm
23. 23 Those involved...   A team of specialists (ideally)   graphic designers   interaction / interface designers   psychologists/human factors specialists   technical writers   marketers   test engineers   software engineers
24. 24 Ensuring Usability Through Design   Design is driven by requirements   what the artifact is for   not how it is to be implemented   Design represents the artifact   for UIs these include   screen sketches or storyboards   flow diagrams showing task structure   executable prototypes   representations always simplify
25. 25 Taxonomy of HCI (Greenberg/CDG)
26. 26 Higher Quality Products   User spends less time on interface and more on solving problem, e.g., one command compiles and executes program   Interface matches the way user thinks about problem, e.g., spreadsheet looks like accounting sheets   Interface adds value to problem solution, e.g., multiple ways to view data
27. 27 User Interface Economics   Good user interfaces sell systems!   Windows is a copy of the Mac interface   The Mac interface is a copy of Bravo - developed at Xerox PARC   User interface capabilities and awareness help get contracts   Poor user interfaces can cripple a system that is outstanding in all other respects
28. 28 Ubiquity of Computers   Computer driven interfaces placed in most mechanical products we know   Classic problem of users not being able to set the clock on their VCR / microwave / car   Users can often not use a duplicating machine, a fax machine, a cash register, a candy machine, a bank machine or even a telephone   Cars will eventually be computer driven
29. 29 Why Are User Interfaces Poor?   Inadequate training of people developing interfaces   Diversity of knowledge required to design good interfaces   hard to find good people   huge market for people with user interface design skills   Rapid technological advances   Reluctance of companies to commit resources   not that true anymore   Poor management - programmers do not talk to user design team and vice versa
30. 30 Lack of Real Engineering of The User Interface   User Interface specialists rarely involved   The "bricklayers" (programmers) are left to do the user interface architecture by default   “Ignorance by software engineers of usability and how to measure it is roughly equivalent to an electronics engineer not knowing what volts and watts are and how to measure them."
31. 31 Current Design Practice   Prevailing practice and theory in HCI Design models 1.  Early focus on users and tasks   Involve uses as much as possible   Readings Norman (UCSD)   Integrate knowledge from different disciplines 2.  Empirical Measurement   Simulate, observe, measure 3.  Highly iterative   Use to check key design decisions   Analogous with programming   Recognition that first time will not succeed   Heed findings of Gould & Lewis re: system designers (1985)   16% of developers identified all 3 techniques   26% mentioned none!   More recent study found…   21% Danish developers never heard of thinking-aloud method and only 6% used it.