Ericsson employees). Three were females, five were males (mean age 32 years). Three had technical background, whereas five had a non-technical background. Average computer use amounted to >10 hours/week. Six of them were classified as Pioneers, one as Materialist and one as Sociable, according to Ericsson's user segmentation model based on the individual's socio-cultural values. The segmentation model gives a way of classifying and selecting the users that belong to the right target group for a new product.
The Particular environment was the following: Each user was given a scenario that described the value-added name retrieval-by-voice feature in general terms. The user was given 7 minutes to familiarize her/himself with the prototype. No manual regarding the use of the added feature was provided. Each user was asked to bring his/her regular mobile phone, including the SIM card along to the usability evaluation. Their own SIM card was taken out of their regular Ericsson phone and put into the prototype during the usability trial.
The Specified goals consisted of 7 typical "voice" tasks that were presented in "logical" order: Record (using voice) the names to two previously text-entered names on the user's own SIM card (Task 1). Text-enter three new names (and accompanying phone numbers) adding voice as well to the Phone book so that they can be retrieved later, using voice-only (Task 2). Text-enter a new name entry to the Phone book, omitting recording the person's name (Task 3). Call one of the previously stored names, using voice-only (Task 4). Substitute a previously recorded voice tag ("Robert Book") for "Robban" (Task 5). Listen to all voice-recorded names (Task 6). Add a new person to the Phone book, storing the name in voice as well (the phone was taken out of the usability lab and additional names were voice-stored by the experimenter, so that the voice memory was full) (Task 7). The presentation order for Task 1 and Task 2 was balanced out. Half of the user group received Task 2 before Task 1.
Usability criteria for Effectiveness (the accuracy, completeness, precision, number of errors with which the specified goals are achieved) and Efficiency (the cognitive resources expended in relation to the accuracy and completeness of goals achieved) were set according to ETSI's suggestions ( Gleiss 1992). Since ETSI does not suggest an Efficiency criterion, this was set for each task by the authors as:
Efficiency (Mean task completion time) < 3 × Optimum path completion time Optimum path (Op) completion time ( Mohageg 1992) is the time it takes to accomplish a task correctly (at a moderate pace) for an experienced user. The multiplication factor (3 × Op) was chosen in order to make allowance for the initial exploration time a voice-na>ve user would need. The following Efficiency criteria were computed: Task 1 (should be completed in less than) < 4 minutes,