HCI Research in MIS

1
HCI Research in MIS

• Dennis Galletta
• University of Pittsburgh
• AMCIS 2002 Dallas

2
ForewardDallas, We Have a Problem!

• Some might think that interfaces have finally
  been fixed.
• Truth is, even current systems are problematic
• What kind of system, for use by the general
  public would puzzle a PhD (me) and two employees
  of the firm?

3
Dart Fares

• Fare types/ zones
• Local/1
• Premium/2
• Reduced/all
• Goal clear

4
Dart Screen

• New types and different zones
• full/all
• full/1
• reduced/all
• Where is
• premium 2?

5
1. Positioning

• In common with other fields
• Usually individual level of analysis
• Usually experiments
• Often address user attitudes, behavior,
  performance
• Training is often a key area
• Sometimes develop the interfaces and sometimes
  evaluate or compare them

6
Positioning

• Different from other fields
• We seldom (never?) develop hardware
• We seldom (never?) develop the operating
  environment
• We less often build an interface
• We have more pressure to show organizational
  impact
• We have less pressure (and ) from vendors
• We often focus on decision making
• Interestingly Our research community has greater
  expectations for theoretical justification and
  large samples

7
Positioning-Illustrations

• Fairly typical example of what youd see at an
  HCI conference (from ACM-CHI)
• A business-related example from the Research
  Channel
• Shneidermans chapters reveal OUR bias

8
Shneidermans HCI

• System design issues
• Guidelines and tools for development
• Managing the design process
• Testing, reviewing, surveying tools
• Software tools for building an interface
• Devices and software interfaces
• Direct Manipulation
• Menu, form, dialog design
• Command languages
• Devices (keyboards, mice, etc.)

• us them

9
Shneidermans HCI

• Research-intensive topics
• Response time/display rate
• Presentation styles (positive, understandable,
  nonanthropomorphic, clear display design)
• Documentation (manuals, help, tutorials)
• Multiple-Window Strategies
• CSCW
• Search and Visualization
• Hypermedia and the Web
• Societal Issues

10
Interesting paradox

• Some of the most well-known HCI studies
• Publishable in an MIS journal?
• probably
• But not heavily referenced!

11
GOMS (Card et al 83)

• (Goals Operators Methods Selection rules)
• Extensively (!) used by HCI researchers
• Much less commonly used by MIS researchers
• A family of models that permits modeling of Goals
  and prediction of a users path toward the goals
• Most heavily used model is the Keystroke Model

12
Screen vs paper (Gould et al.)

• IBM researchers tried to discover why people read
  more quickly and effectively on paper versus
  screen
• No overriding theoretical model or basis just
  kept guessing what factor was responsible
• Finally discovered one

13
2. My applications

• Primarily interested in user abilities ( fit)
• Cognitive Fit influences
• training psych course I took at Minnesota (it
  depends)
• PhD seminar question given to us by Gary Dickson
• a Herb Simon CMU seminar Iris and I attended
• More recently examined spreadsheet detection
  ability
• Most recently investigated fit between varying
  web site characteristics (speed trades off
  against depth and familiarity)

14
Applications

• More successful
• Had theoretical justification
• Had results
• Had implications
• For researchers
• For practitioners

15
Applications

• Less successful
• Thesis study system failed
• CSCW study no results
• Stress study few results

16
As an AE/Reviewer

• Commonly, failed studies
• Could mislead the reader/field
• Control was lacking
• Confounding effects were contained
• Substantial, generalizable sample was not tapped
• Measures were not appropriate
• Contained difficulties in scope
• Address a truly miniscule issue
• Dont do what they say they will do in the
  introduction
• Were poorly written
• Dont justify or set a context
• Lack important disclosure
• Dont communicate

17
3. Journals and Editors

• Is an HCI study hard to publish?
• I did a quick review
• Answer It Depends
• MISQ 12-15
• ISR 22-25

18
ISR
19
MIS Quarterly
20
4. Future

• Bright.
• HCI is a strong area of MIS
• Web developers abound
• More people are using systems
• More people fact ever-more terrible systems
• Opportunities
• Usability as a driver of system development
• Arrangement and categorization
• User training is alive and healthy