Haas: "Reading Online"

Haas, Christina. “Reading Online.” Writing technology: studies on the materiality of literacy. Mahwah N.J.: L. Erlbaum Associates, 1996. 51-72. Print.



Interest: how computer technologies affect writers. Finding: a computer is not a computer is not a computer. Depends on configuration and use.


Examining Mismatches between Computers and Writers


Citing Barton, notes that dominant discourse about computers tends to be celebratory. One root is technological celebration; other may be vision of computer as streamlining. Focus of article not on benefits of computers, but the problems that users have with various aspects of the machine.


Format: three interviews, discussion of three studies, and feature analysis. Argues size, legibility, responsiveness, and tangibility significantly affect user''s experience (53).


Writers'' Reading Problems and Their Use of Hard Copy


Conducted 30 interviews over 10 years. Will focus on only one issue: difficulty in reading in writers reading their own texts online. Participants varied in age from 18-52 years. Open-ended questions to allow for unanticipated responses. Technology ranged from low end (CRTs in early 1980s) to PCs. All had some form of printer access. Many differences in levels of use, but Haas notes common positive views: positive approaches to using computers for writing; freer; better on-screen formatting; network access, file systems; "sheer fun" (54). Common negative views: causes reading difficulties. Physical space allows for laying out pages; "intimacy" (55) of pen and paper. Research demonstrates that closer reading/more effective proofing in hard copy. Writers interviewed pretty much universally used hard copies.


Four Types of Reading Problems:


  • Formatting: Especially prob for those without WYSIWYG editors
  • Proofreading: resolution and display, as well as documented issues with online proofreading errors
  • Reorganizing: More difficult to do when limited view of text (unlike when entire paper as papers before writer). Actual reorganization is easier, but conceptualization more difficult.
  • Reading for the Sense of the Text: A difficulty identifying/analyzing shape of writer''s own arguments


Three Studies of Writing-Related Tasks


Three empirical studies measuring: recalling location of information; retrieving information; and reading to revise.


Study I: Recalling the Location of Information


Sample: 10 (6W; 4 M) Grad students in TCR at Carnegie Mellon. Randomly assigned hard copy or computer condition.


Procedure: Read 1000 word article; instructed to "read normally." Then asked to recall location of 8 items.


Analysis and Results: Three scores for each question: text sequence score (difference between original page or screen) and one chosen; vertical score (difference between original line or row and one chosen); horizontal score (difference between original column or character number and one chosen). Found significant difference in vertical score.


Discussion: not a scrolling screen. Simply theorize that this may contribute to issues of readability. Why no mention of gender?


Study 2: Retrieving Information


Builds on results of study 1. Participants read texts in three conditions: printed text; CRT; higher res/more sophisticated unit (called "Andrew").


Participants: 15 in Humanities and Engineering. Why isn''t gender mentioned here? All had range of familiarity with computers, but those in 3 had some special training in Andrew in preparation for sessions.


Procedures: read 1800 self-help file on CarnMell sys. Asked 6 questions that could only be answered explicitly from reference to text and 6 that could only be answered by inference from text. Instructed to locate information even if they could recall. Timed. Incorrect responses fewer than 1 per question and distributed among the questions and conditions.


Analysis and Results: Because so few errors, only time results analyzed. Note marked difference in location time between CRT and hard copy; no statistically significant differences between hard copy and Andrew.


Discussion: Could be result of display differences, differences of user interface.


Study 3: Reading to Revise


Needed to construct meaning of a scrambled text. Several computer systems tested.


Participants: 10 entering freshman with fewer than 10 hours of computer experience. Participants trained approx 3 hours in 2 sessions.


Procedures: Five 1200 word texts (11.5-12 grade readability). Conditions: hard copy; large copy with scroll bar; large screen and function keys; small screen and scroll bar; small screen and function keys. Participants orally verbalized how text should be reorganized. Explored reading only, not ability to rework (which could be technical prowess).


Analysis and Results: hard copy and large screens similar; mean time lowest for hard copy; large screen came in second. No significant diff between scroll bars and function keys.


Discussion: Supports view that visibility/legibility key issue.


A Framework of Computer Features


Returns to interviews. three empirical studies focusing on such issues as display, speed, screen size, etc. (69). Theorizes as to reasons for responses/ results. Not a single factor, but interconnected issues. Leads to questions of how systems/interfaces should be designed.