Abstract The main thrust of this study is to investigate how users of mass-market word-processing software go about using the built-in help systems provided in application packages as they work: how frequently they seek help with the functions they need, how they approach and access this help, how they perceive it, and in what ways their performance on tasks is improved as a result. The study concludes by suggesting how users’ confidence, effectiveness and approach to using on-line help systems might be improved, as well as recommending further investigations. The help facilities provided in computer software can take on varying forms: hints, help and coaching. To begin, I take hints to be small pieces of advice offered generally, at no particular time, but possibly ahead of the performance of the activity to which they relate. So the kind of information counted as hints would include what Word refers to as ‘tool tips’—short messages that appear when the mouse pointer rests on a visible item, such as a button or an item on a window; as well as the ‘tip of the day’ which is displayed when the program starts up. As well as these explicit hints, there are also implicit ones, such as those embodied in the names of menu items (if any one doubts their effect, let them try using a version of a program intended for speakers of a different language, where the menu names and the names of menu items can be of very little help. For example, under the French ‘Fichier’ menu, there might be the following items: Nouveau, Ouvrir…, Ouvrir l’élément recent, Fermer, Enregistrer. Enregistrer sous, Tout enregister, Format d’impression, Imprimer… ). Even more abstractly, it could be claimed that the whole ‘look and feel’ of a program continually prompts the user with hints. Help is the general focus of the bulk of this thesis, taking in any feature that is meant to respond to a user’s request for assistance. I need discuss this here no further as it will be comprehensively covered later. Finally, coaching can be defined as any activity that is intended to improve performance or to correct errors or bad habits. In contrast to help, it is often prompted by the user’s behaviour (“It looks like you’re writing a letter. …”), or by the error itself. The coloured underlining done by the spelling and grammar checkers could also be called coaching. And the most obvious coaching lessons are those provided by the various ‘wizards’ that can be invoked via menu items to guide the user through complex procedures, such as executing a mail merge. The literature that was surveyed ranges over a number of areas of relevance, several of which were explored in some detail, such as: references to help systems specifically, the way that beginners typically fail to progress beyond a modest level of proficiency, the human–computer interface (HCI), the efficacy of help systems and the factors influencing how users approach and value them, the categories of complaints frequently made about help, computer-mediated learning, the heuristics of usability testing and what evaluative approaches are possible. The literature provided valuable background for formulating and implementing an experimental program and in evaluating and analyzing its outcomes. In particular, it was realized at an early stage that several techniques commonly employed in usability testing were beyond the financial and staffing resources available; for example video recording of users’ interactions with software, or monitoring of these interactions by expert, trained observers and evaluators. An alternative approach was developed, which drew on the resources of the student subjects, who worked in pairs so that one member of the pair tackled the set tasks while the other observed and took records on pro-formas devised for this study. This approach, using pairs (or ‘dyads’) turned out to be successful. Since an action-research paradigm was adopted, in which a process of evaluation and modification was applied to each successive experiment, it was possible to progressively refine the approach with respect to both the exercises given to the participants and the procedures and pro-formas used to monitor and record the results. An article on the approach was written for the British Journal of Educational Technology, and this is included as an appendix to this thesis. The experimental subjects (those who contributed to surveys, plus a total of 176 who participated in the initial focus groups and the task-based experiments) were recruited from the students, mainly in first year, of the School of Information Technology and Electrical Engineering at the University of Queensland. The assumption was made that this sample would not differ substantially from the general population of those who use word processors for a variety of purposes but who are not professional clerical workers. The word processor with which almost all of them were already familiar was Microsoft Word, and this was adopted as the vehicle for the experimental program After an initial phase in which students’ attitudes and practices were explored to provide a context, using focus groups and an on-line survey, a series of experiments was designed and conducted, using an action research paradigm to refine the techniques and materials used as the program proceeded. A set of tasks was devised for each experiment, ranging from the commonplace (such as changing font and text size) through the more demanding (such as manipulating tables) to the less routine (such as insertion of graphics in the text and as background). The tasks were selected so that they would require most students to seek help at some point, while not being exceptionally challenging. Interesting and potentially valuable results have come out of these experiments: many of the students had been using the software for a number of years, starting at school, and were currently word-processing for several hours a week, but surprisingly they were typically using just a minimal sub-set of the capabilities of the software, given that they could satisfy most of their writing needs with them. There appeared to be a common tendency to avoid seeking help, since when the set tasks called for more advanced techniques a substantial proportion of users preferred to browse the menus, or use trial-and-error, rather than take full advantage of the help system. However these students, when forced into it by the terms of the experiments, were generally able to find and use the help that they needed to perform well in the more demanding tasks. This study suggests that many users are coping reasonably well with their day-to-day word processing but might save time and effort in the longer term (and produce more rewarding outcomes) if they could only take advantage of additional features of the software. One factor hindering them is their reluctance to use help, whether this arises from unfamiliarity, lack of confidence, unfortunate earlier experiences, sheer prejudice, or simple inertia. As well as the potential improvements in users’ facility with the packages that might be brought about by training, which was once general but has recently been largely discarded, there are ways of making help more acceptable and thus more immediately usable for these people: help systems could be made more palatable and at the same time more educative, without being patronizing or paternalistic, and this depends in part on achieving a closer alignment between the mental models of the software held by various disparate groups of users and those used by its designers. In summary, and on careful inspection, Microsoft Word is revealed as a comprehensively accomplished application that can appear daunting at first. It is a pity that, according to this study, there are many users who never progress beyond the unambitious use of a modest set of its capabilities. This thesis concludes by suggesting that barriers to the full exploitation of the powers of Word and similar software packages may be overcome by a few simple strategies, including simplifying the already plentiful repertoire of customisation tools and devising training aimed specifically at enriching and smoothing the beginners’ learning curve.

Thesis (Ph. D.)--George Mason University, 2007. / Title from PDF t.p. (viewed Jan. 17, 2008). Thesis director: Shobita Satyapal. Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Physical Sciences. Vita: p. 217. Includes bibliographical references (p. 202-216). Also available in print.

Thesis (M.S.)--Ohio University, August, 2000. / Title from PDF t.p.

Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: Cam; Tension-Member; Cam Follower; Metal Tape; Metal Belt. Includes bibliographical references (leaves 79-80).

Thesis (M.Sc. in Microbiology) -- Mahidol University, 1976.

Thesis (M.S.)--Auburn University, 2006. / Abstract. Vita. Includes bibliographic references (p.37).

Inaug.-Diss.--Bern. / "Literatur": 1 p. at end.

Thesis (Ph.D.)--University of Texas at Dallas, 2009. / Includes vita. Includes bibliographical references (leaves 224-234)

Thesis (Ph. D.)--Lehigh University, 1997. / Includes vita. Bibliography: leaves 109-114.

Thesis (M.S.)--Virginia Polytechnic Institute and State University, 1993. / Vita. Abstract. Includes bibliographical references (leaves 61-70). Also available via the Internet.