| Markup languages have become an important component of computing systems design and development, especially with the rapid growth of the Internet, World Wide Web, and electronic publishing. Through the use of markup languages such as HTML, SGML, and XML, users can not only format text, create web pages and sites, but can also apply meaning and structure to text and data.; The objective of this research was to study the impacts of different interaction modes, help system, and user background, on performance, learning, satisfaction, and time perception as it relates to markup languages. The markups used were a set of new markup extensions, SQML (Survey Questionnaire Markup Language), designed by the authors (Hsu and Turoff, 1996), which focused on question types for survey questionnaires. The study was conducted using MLSS (Markup Language Software System), a Java-based application which runs on a Windows PC. This software interprets SQML markups and produces on-line surveys, and collects data on the user's interaction using a transparent log file feature.; The design of the experiment is a 2 x 3 factorial, crossing interaction mode (Command, Wizard) with help system (Offline, Online, Visual). The role of a subject's skill level (novice, experienced) is also studied. The experimental task was to create an online version of a printed questionnaire using SQML. The markup code created, the log file transcript, and the responses to the Post-Test Questionnaire were analyzed. Subjects were primarily undergraduate students in Computer Science, Engineering, and Accounting at NJIT and Queens College.; The results indicate that the choice of mode and help system can produce significant differences in performance, learning, and satisfaction. For example, novices learned best in modes/help systems which they liked the least, such as command mode with offline help, and have poor perceptions of their own learning. Wizards, which increased satisfaction and speed, did not improve learning for novices, but tended to equalize the performance between skill levels. These and many other conclusions are summarized and presented as recommendations for researchers and practitioners in human-computer interaction, system design, and training. |
