Technical Translation: Usability Strategies for Translating Technical Documentation

Developing a Questionnaire 
Kirakowski (2000:6) maintains that developing a questionnaire requires “a 
lot of time, patience and resources”. In addition, considerable knowledge of 
psychological measurement and statistics is essential. Kirakowski also main-
tains that unless all of these prerequisites are met, serious questions arise as 
to the validity and reliability of the questionnaire. As a result, he recom-
mends that a pre-designed, pre-tested questionnaire should be used instead. 
In view of this, we will examine a number of commonly available usability 
questionnaires and select a suitable one. These range from freely available, 
public-domain versions to sophisticated commercial varieties. 
A number of the available models are simply unsuitable for testing 
documentation, e.g. 
WAMMI
1
which is designed for testing websites. Oth-
ers, such as 
IsoNorm
2
(developed by Jochim Pümper on the basis of 
ISO9241-10) are only available in languages other than English which 
could not be understood by participants.
Of the remaining models, the following criteria can be used to select the 
most appropriate one: 
cost
reliability 
data validation & standardisation 
flexibility/customisation 
The first such model is the 
Computer System Usability Questionnaire 
or
CSUQ
3
. This questionnaire is available free of charge and incorporates psy-
chometric reliability properties. However, the test is run using a web page 
which is submitted to another server for processing. While this removes the 
1
http:// www.wammi.com 
2
http://www.sozialnetz-hessen.de/ergo-online/Software/Isonorm-Workshop.htm 
3
http://www.acm.org/~perlman/question.cgi?form=CSUQ 
192
•
•
•
•

---

Survey Methods
effort of processing data, it also leaves the test administrator without a per-
manent record of the answers provided. Another problem with this is the 
fact that questions are phrased in a way that could lead or prompt specific 
responses. Obviously, this can distort responses and lead to unreliable data. 
Furthermore, the questionnaire does not refer to documentation and it 
cannot be customised or modified. 
Another questionnaire is the 
Software Usability Measurement Inventory 
or SUMI
4
developed by the 
Human Factors Research Group
at University 
College Cork. This questionnaire overcomes the problems of CSUQ in 
which eliminates the need for test administrators to perform complex calcu-
lations. However, samples of SUMI available on the Internet fail to include 
documentation. Above all, the high cost of SUMI can prove prohibitive. 
The questionnaire selected for this study was the 
Questionnaire for User 
Interaction Satisfaction 
or QUIS
5
developed by Kent Norman and his re-
search team at the 
University of Maryland at College Park
. Using a 9-point 
Likert scale, QUIS features a section to determine overall levels of satisfac-
tion and hierarchical sections which examine satisfaction with specific as-
pects of the interface (see Table 1). 
screen factors 
learning factors 
technical manuals 
multimedia
virtual environments 
software installation 
terminology and system feedback 
system capabilities 
online tutorials 
voice recognition 
Internet access
Table 1: QUIS Question Categories 
What makes this model so attractive is that is can be modified and custom-
ised to suit the needs of a particular study. Thus, entire blocks of questions 
can be omitted. Just as importantly, all questions are phrased in a neutral 
4
http://sumi.ucc.ie 
5
http://www.cs.umd.edu/hcil/quis/ 
193
that questions are worded in a variety of ways so as to avoid leading res-
pondents. SUMI is supplemented with a data analysis and reporting service
•
•
•
•
•
•
•
•
•
•
•

---

Assessing Usability 
questionnaire. 
QUIS also comes with instructions for analysing and processing data (see 
Norman 2003). Indeed, this is reinforced by references to other research 
projects which have used QUIS. From a cost point of view, QUIS is com-