Validity in Image‐based Research: an elaborated illustration of the issues

Abstract

Abstract This article is concerned with the trustworthiness of products of research which are derived from image‐based data. Fundamentally it asks whether validity represents an appropriate measure of the quality of research which uses such data. Two examples of image‐based research are taken from recent work undertaken by the authors. A comparison of these examples provides a well‐grounded illustration of the issues confronted by the researchers in their attempt to consider the validity of their findings. The resulting discussion develops the argument that the concept of validity is particularly problematic in its application to image‐based research. We present a tentative strategy for ensuring trustworthiness that is based on five key criteria: completeness, adequacy of interpretation, transparency, self‐reflection and the aggregation of conflicting interpretations.