Validity and Reliability in Visual Methods

Validity

Validity issues begin with the selection of appropriate methods which are capable, in theory, of fulfilling the needs of the research project. Once visual methods are selected, other issues of validity arise in relation to subject matter, context and framing, and the extraction of information from what has been captured (Adelman 1998: 151). For this note, the focus is on a research taking still photographs for the information that can be extracted from them but the same principlies apply to other forms of visual method. It is fundamental to validity that all imagery used in the project is relevant to it.

Context and framing require forethought. In designing the research project the researcher will have developed a view on when and where it is most likely that what is being investigated will be observable and being in the right place at the right time is part of building validity (Adelman 1998: 150). The researcher still has to make decisions about what aspects to photograph and where to position self as photographer and these decisions are aided by observation in advance of taking photographs (Adelman 1998: 150). Having decided what to photograph the process then moves to how to frame the image and how much contextual background to include in it. (Adelman 1998: 155)

In the extraction of information from the image the researcher is at an advantage over the viewer because of additional contextual information the researcher derives from having been at the scene. Whatever the researcher claims to be represented in the image must be there to be observed by another viewer, accompanied by captioning or other contextual background information that may be necessary (Adelman 1998: 158).

All these validity points relate to the internal validity of the imagery in the context of the specific project. External validity arises from being able to extrapolate from the visual evidence to make valid generalisations. Validity is also closed tied to reliability (Adelman 1998: 158).

Reliability

Reliability goes hand in hand with validity. Although reliability can be achieved in isolation, it is of no relevance unless it is linked to validity in relation to the project. Reliability has nothing to do with the reliability of the technology but is to do with the principle that repeating an action many times under the same conditions leads to concentration of results from which the dominant position can be derived (Becker cited in Harper 1998: 29). In visual methodology terms, taking many photographs under similar conditions will reveal what is dominant and prevent the researcher’s interpretation and conclusions being influenced by the exceptional. The more photographs that are taken the stronger the concentration of results will become. Depending on the nature and scope of the research, the same process may need to be repeated at different times and in different locations.

Adelman, C. (1998) ‘Photocontext’. In Image-based Research: A Sourcebook for Qualitative Research. ed. by Prosser, J. Abingdon: RoutledgeFalmer: 148-161

Harper, D. (1998) ‘An Argument for Visual Sociology’. In Image-based Research: A Sourcebook for Qualitative Research ed. by Prosser, J. Abingdon: RoutledgeFalmer: 24-41