smf@hfserver.hfnet.bt.co.uk (SMF) (02/06/90)
REQUEST FOR INFORMATION ABOUT SEQUENTIAL TESTS Engineering Psychologists, and other Human Factors experts, working in the area of the usability engineering for information and communications products, have been developing methods to set minimum standards for ease of use. An element of this activity has involved establishing procedures for the reliable measurement of components of these systems with a typical operator. In order to asses the ease of use of an alphanumeric keyboard, it has been suggested that the performance of the keyboard under test be compared with that of a reference keyboard. This would be for a group of subjects performing a standardised task. The International Standards Organisation (ISO) is currently considering a draft standard based around this procedure. The proposed test for statistical difference between the two keyboards is the Barnard sequential t-test (Davis 1963). This test has the advantage of requiring a smaller number of subjects, in order to determine whether there was a statistically significant difference at a given probability level, than a conventional t-test (Brigham 1989). I am interested in obtaining information about the strengths and weakness of the use of sequential tests of significance within behavioural experimentation. I am particularly interested in identifying references were it has been used for the analysis of experimental results, or of a commentary on the use of this type of test for experimental analysis in psychology. REFERENCES Brigham, F. R. (1989) Statistical Methods for Testing the Conformance of Products to User Performance Standards, Behaviour and Information Technology, 8(4), 279-283 Davies, L. O. (1963) The Design and Analysis of Industrial Experiments, Oliver & Boyd London 2nd ed. -- Stephen Furner Tele. +44 473 643818 Industrial & Human Factors Design Divsion Fax. +44 473 637557 British Telecom Research Labs. E-mail smf@hfnet.bt.co.uk Martlesham Heath, Ipswich IP5 7RE, UK