What kind of assistance do users really need?

Windebank, James (2010) What kind of assistance do users really need? MA dissertation, University of Portsmouth.

[img] PDF
Restricted to Registered users only

Download (1857kB)

    Abstract

    User assistance for software applications is widely criticised because many users have low expectations of its effectiveness and believe the time spent in a help system brings little return. This raises the question as to whether technical communicators who design user assistance actually understand why users turn to their assistance. Do they understand what kinds of problem users are having, do users need complete guidance, or are the users simply seeking a small fragment of information that will help them complete their task? This research aims to discover the questions that users really ask through the means of a practical pilot study. The study initially asked 20 participants to tackle three tasks on three different types of software applications. The participants were not allowed to use any form of help and if they had any questions, they were encouraged to ask a moderator. Each participant was recorded to capture when they asked a question and what types of questions they asked. The research begins with a literature review that considers the different design attributes in creating user assistance which is retrievable, comprehensible, and usable. The literature review draws on expert opinion from the world of technical communication, with focus on current user assistance issues. Throughout the literature review, many experts claim that a thorough user and task analysis will enable help authors to understand the audiences. By understanding the users and their needs, authors will then have a greater possibility of effectively predicting the user’s questions. The study has revealed that confirmation questions are the most common question type asked. Confirmation questions were asked 47% of the time, with task questions in second at 24%, and location questions in third at 15%. Therefore, if user assistance is to be successful, it should aim to answer these types of questions, in order to be 86% effective to the users. The gender of the user did not appear to affect this outcome. However, the level of prior knowledge of a software application and the task level were shown to influence the type of questions asked. As the other four question types (Meaning, Reason, Response, and Identity) make up the remaining 14%, it can be argued whether help authors should pay particular attention to answering them. The contents of this research report may interest professional technical communicators who wish to design and develop effective user assistance. Furthermore, software developers may also have interest in this research as the integration of user assistance into an application is a key feature of today’s software production.

    Item Type: Dissertation
    Departments/Research Groups: Faculty of Humanities and Social Sciences > School of Languages and Area Studies
    Depositing User: Jane Polwin
    Date Deposited: 20 Jan 2011 12:49
    Last Modified: 28 Jan 2015 11:16
    URI: http://eprints.port.ac.uk/id/eprint/884

    Actions (login required)

    View Item

    Document Downloads

    More statistics for this item...