Qualitative data can be gathered from an array of rich
sources of research information. One of the popular ways to collect
this data is by interviewing a range of experts on the topic,
followed by transcription, resulting in a database of written
documents, often supplemented by other documented data that informs
the topic. Thematic or Content Analysis can then be used to explore
the data and identify themes of meaning that enlighten the research
topic, with the themes being gathered into nodes. The researcher now
has an array of nodes, which needs to be organised into a coherent
model, and more importantly, one that represents the views of the
research informants. To do this with some degree of rigour, the
researcher needs some way of ranking the nodes in terms of their
relative importance. The node ranking can be based on experience, or
on the literature, but neither of these approaches looks to the data
itself. If the database contains new or unexpected knowledge,
neither experience nor the literature will guide us to it, and vital
new insights may easily be missed. The framework outlined in this
paper aims to provide a sound first‑cut analysis of the data, based
on the evidence in the research interviews themselves. Clearly the
literature and research experience have an important role to play in
shaping the results of any research. However this paper argues that
one should proceed only after the data itself has been offered "the
first chance to speak".The node classification matrix detailed here,
identifies distinct node categories, each ranging in significance
and with particular characteristics that reveal key aspects of the
informants' views. In this way the researcher can use the nodes to
reveal the voice of the experts, and build a scientifically rigorous
set of results from a qualitative database.