Task Performance Measures

It has been suggested that task performance measures can be used as objective corroborative indicators of presence (Barfield & Weghorst, 1993). Though it is generally assumed that higher levels of presence are associated with better task performance, the exact relation between presence and task performance remains unclear. There is no firm evidence indicating a causal link between he two constructs. It seems plausible that several characteristics of a VE will similarly influence presence and task performance (IJsselsteijn, 2004). Also, characteristics of the user such as ability and motivation will influence task performance (Heeter, 2001). Task performance measures are only applicable in media environments where there is a clear task that should be performed.
Many studies associating presence to task performance have been conducted. Only a few of them are described here, since it is our aim to review measures rather than draw a conclusion on the relationship between presence and task performance.

Completion Time and Error Rate

Description

The time taken to complete a task and the number of errors made in that task are the most classic ways of measuring task performance.

Research

Basdogan, Ho, Srinivasan, and Slater (2000) investigating the influence of haptic feedback on task performance and sense of togetherness of participants in a shared virtual environment (this study was described in the section Basdogan et al. Questionnaire). They calculated a task performance score based on the total amount of time taken to complete the task, and the ratio of time spent in error-free condition. The results show that haptic feedback increases both the performance score and subjective feeling of togetherness. Correlation between the measures is not reported.

Sensitivity: the performance score discriminated between two conditions
Reliability: not reported
Validity: similar pattern in performance scores and subjective feeling of togetherness

Primary Source

Number of Actions

Description

Instead of measuring the time taken to complete a task, it is also possible to count the number of actions or steps that is needed to complete a task.

Research

Slater, Linakis, Usoh, & Kooper (1996) used performance measures in a study (n=24, mixed design) in which participants played the game tri-dimensional chess. Participants were either immersed in the environment by means of a HMD (egocentric view), or viewed it on a TV (exogentric view). The environment was either realistic, or plain. The complexity of the game was slightly varied, requiring either 7 or 9 moves. Performance was measured as the number of correct moves (out of 7 or 9) that the subject made. Also, the SUS questionnaire was administered after the experience.
Results show that performance was positively associated with egocentric immersion in comparison with the exocentric screen based viewpoint. Also, performance was positively associated with a more realistic environment compared with an empty environment. Subjective presence was significantly higher for the egocentric compared to exocentric condition, but realism did not have a significant effect on subjective presence.

Sensitivity: the performance score discriminated between different conditions
Reliability: not reported
Validity: partly similar pattern in performance scores and SUS scores

Primary Source

Secondary Task Performance

Description

Performance on a secondary task can serve as a measure for the amount of effort allocated to the primary task. The more effort is dedicated to the primary task, the more performance on the secondary task will decrease. A similar argument can be made in the case of presence: if more attention is allocated to the mediated environment, performance on a secondary task will decrease. Reaction times or error rates can be used as secondary task performance measures (IJsselsteijn, 2004),

Research

Research using direct secondary task performance measures has not been reported. Nichols et al. (2000) used a secondary task paradigm; they measured background awareness by recall of background music (this study has been described in the section attention/awareness). No difference was found between awareness scores in the different conditions.

Sensitivity: the performance score did not discriminate between different conditions
Reliability: not reported
Validity: not reported

Primary Source
  • Nichols, S., Haldane, C., & Wilson , J. R. (2000) Measurement of presence and its consequences in virtual environments. International Journal of Human Computer Studies, 52, 471-491.

Transfer

Description

Performance can be measured by the degree to which a skill which is learned or practised in a VE is transferred to a real-world situation.

Research

Youngblut & Perrin (2002) pursued this approach in a study (n=40, between-subjects design) in which participants practiced an aircraft maintenance task in a VE, either for a long or a short time. Afterwards, training transfer was measured as the time taken to complete the task on a physical mock-up and the number of errors made. They also completed the PQ and the SUS. The results show no effect of practice time on any of the performance measures. Neither of the presence questionnaire scores showed a correlation with time taken, but both showed a significant negative correlation with the number of errors.

Sensitivity: the performance scores did not discriminate between different conditions
Reliability: not reported
Validity: number of errors was correlated with subjective presence measures (PQ and SUS). Validity of the time measure is not supported

Primary Source
  • Youngblut, C., & Perrin, B. M. (2002). Investigating the relationship between presence and performance in virtual environments. Paper presented at IMAGE 2002 Conference, Arizona.