Technical Translation: Usability Strategies for Translating Technical Documentation

Foss et al. 1981 
Basing their work on the claim that previous work in the area of text acces-
sibility and usability was vague and often contradictory, the authors set out 
to “understand better the acquisition, representation, and utilization of 
knowledge by novice or occasional users” of software and to “test some 
ideas derived from current views of memory and attention” (Foss 
et al. 
1981:332). 
The basic hypothesis is that users learn more effectively when they 
understand what they are doing. To test this, a comparative experiment was 
carried out using two versions of a computer manual: both versions were 
essentially identical in content but one was restructured
to present
information in order of progressing complexity. Two groups were formed 
and each was given one of the two versions of the manual. 
In addition to the original and revised manuals, the authors produced 
planation and description of the basic characteristics of the software being 
used. It did this in general terms describing key parts of the software with-
out referring to specific elements of the software. The 
Advance Organiser 
was given to half of the subjects in each of the two groups.
In conducting the experiment, which took three hours per subject, subjects 
were first given general information about the experiment and their typing 
speeds were measured. Selected subjects were given the 
Advance Organiser
and told to study it. Subjects were then given one of the two manuals and 
told to study it for 15-30 minutes. Once this had been completed, a book-
let containing nine text editing exercises was distributed to subjects. 
The subjects were told to rely as much as possible on the manuals and 
that they could only ask the experimenter as a last resort. During the course 
of the experiment, the experimenter remained in the room noting certain 
aspects of the subjects’ behaviours such as the amount of time they spent 
performing each task, the number of interactions between the subject and 
203
Sullivan and Chapanis in 1983 and was concerned specifically with reengin-
neering a user guide to take into account best practice in terms of technical
communication and human factors. The following sections describe 
what they call an “
Advance Organiser
”. This document consisted of an ex-

---

Assessing Usability 
the experimenter as well as the number of tasks completed within the allot-
ted time. Although subjects were told not to ask the experimenter for help, 
some did in fact ask for help. Such instances were dealt with according to a 
strict procedure. Firstly, subjects were told that the information they re-
quired was contained in the manual. If this was unsuccessful, the experi-
menter pointed out the specific chapter in the manual. If the subject still 
experienced difficulties, the experimenter gave explicit instructions.
Such an approach may seem inappropriate when the purpose of the ex-
periment is to determine the performance of subjects using the manual. 
However, if we look at this in a wider context we can see that if a manual 
is effective, there should be no need for subjects to ask questions. Thus, a 
subject being forced to ask certain types of question indicates some form of 
problem in the manual. Sometimes during the experiment, the experi-
menter intervened when it was apparent that the subject was struggling to 
complete tasks, e.g. the subject sat for long periods without accomplishing 
anything. 
In addition to the manual recording of subjects’ behaviours by the ex-
perimenter, the experiment also utilised a rudimentary (by today’s standards 
at least) keystroke logging application which recorded the type and number 
of commands used to perform each task. 
On the basis of the data collection techniques employed in the experi-
ment, it was possible to analyse performance on the basis of the following 
criteria: 
number of tasks completed 
average time to complete a task 
average number of commands used to complete a task 
average number of verbal interactions during a task 
number of errors at the point where the subject said the task was com-
plete 
The results of the experiment showed that the organisation of a manual can 
dramatically improve user performance. It also showed that providing ad-
vance information allowed users to “scaffold”
6
their learning, thereby mak-
ing learning more efficient. While this study does not relate directly to the 
6
For a more detailed discussion of scaffolding in the context of social constructivist 
theories of learning, see Kiraly (2000). 
204
•
•
•
•
•

---

Usability Evaluation Procedures
current research in that it involves restructuring the sequence of informa-
tion in a manual and the use of external materials, it does provide a useful 
insight into usability evaluation procedures involving documentation. 
Another case study which relates more closely to this research in terms of 
goals and materials is the one carried out by Sullivan and Chapanis in 1983.