Representing Software Project Vision by Means of Video: A Quality Model for Vision Videos

TL;DR

This paper introduces a comprehensive quality model for vision videos in Requirements Engineering, aiming to improve the creation and evaluation of videos that effectively communicate software project visions.

Contribution

It proposes a novel quality model with 15 characteristics for vision videos, including hierarchical evaluation and production guidance, validated through an empirical study.

Findings

Six quality characteristics significantly correlate with perceived video quality

The quality model covers all essential aspects of vision video quality

Empirical validation supports the model's relevance and applicability

Abstract

Establishing a shared software project vision is a key challenge in Requirements Engineering (RE). Several approaches use videos to represent visions. However, these approaches omit how to produce a good video. This missing guidance is one crucial reason why videos are not established in RE. We propose a quality model for videos representing a vision, so-called vision videos. Based on two literature reviews, we elaborate ten quality characteristics of videos and five quality characteristics of visions which together form a quality model for vision videos that includes all 15 quality characteristics. We provide two representations of the quality model: (a) A hierarchical decomposition of vision video quality into the quality characteristics and (b) A mapping of these characteristics to the video production and use process. While the hierarchical decomposition supports the evaluation of vision videos, the mapping provides guidance for video production. In an evaluation with 139 students, we investigated whether the 15 characteristics are related to the overall quality of vision videos perceived by the subjects from a developer's the point of view. Six characteristics (video length, focus, prior knowledge, clarity, pleasure, and stability) correlated significantly with the likelihood that the subjects perceived a vision video as good. These relationships substantiate a fundamental relevance of the proposed quality model. Therefore, we conclude that the quality model is a sound basis for future refinements and extensions.