Comparative analysis of usability between text and virtual reality based expert systems for cattle diseases diagnosis

The use of cattle disease diagnosis expert systems has the ability to improve the cattle production as they would help in disease control. Particularly if the system is usable, that is, if it is efficient, effective, satisfying and requires less effort to use.

Cattle disease diagnosis expert systems have been adopted in recent years in the cattle production industry particularly in Asian countries like China. However, the usability of these systems still remains a challenge. There is limited or no empirical research that investigates comparison of usability of a text and non-immersive virtual reality-based (multimedia) cattle disease diagnosis expert system.

This study seeks to bridge this knowledge gap by conducting a quasi-experiment research which evaluates the usability of a text and a non-immersive virtual reality­based (VR) cattle disease diagnosis expert system. Software Usability Measurement Inventory (SUMI) was used to gain insight in user satisfaction. Video-recording software (Morae TechSmith) was used to measure task efficiency and effectiveness. A self­reporting ratings using NASA Task Load Index (NASA-TLX scale) was used to gain an insight of mental effort (Cognitive workload) invested by users when performing a task. Sixty-four participants, consisting of aspiring, subsistence and commercial cattle farmers within the Palapye (Botswana) veterinary district, took part in the study. They were divided in two groups of 32, one group using a text-based expert system and other group using a VR-based expert system.

The results indicated that VR-based expert system is better than a text based expert system in terms of effectiveness, efficiency and user satisfaction. Moreover, users using a text based expert exerted less mental effort (work load) than those using the VR­based system. The overall differences indicated that VR-based ES was 75% better than the Text based system which only scored 25%. Results acquired from this study imply that software developers should develop VR-based cattle diseases diagnosis expert system that is usable for cattle farmers.