*2.3. Data Analysis*

The decision to develop and validate a taxonomy suitable to code the events dataset was made due to other commonly utilised taxonomies being identified as lacking applicability for retrospective analyses. Consequently, initial qualitative analysis was required in order to create a basis for the development of this taxonomy.

Thematic analysis, which Braun and Clarke [11] explain, is a method for identifying, analysing and reporting patterns (or themes) within data, was chosen as the primary qualitative analysis method for this study. A specific type of thematic analysis, known as template analysis, was utilised for the purpose of taxonomy creation and development.

Brooks et al. [12] describe template analysis as:

*"a form of thematic analysis which emphasises the use of hierarchical coding but balances a relatively high degree of structure in the process of analysing textual data with the flexibility to adapt it to the needs of a particular study."*

The coded themes which emerge from the data during this analysis are known as the "template" [13] and it is this template which forms the impetus for the taxonomy creation. The template analysis structure tends to be hierarchical with sub-themes emerging within themes [13], ideal for the development of a taxonomy.

NVivo 12 Plus qualitative analysis software was selected as the main tool for the template analysis. The events contained within the dataset were uploaded into the software as individual "cases", where each event was analysed for key themes, known as "nodes". This inductive thematic analysis is described by Braun and Clarke [11] as a *"process of coding data whereby no attempt is made to fit it into a pre-existing coding frame, or the researcher's analytic preconceptions".*

During the generation of the baseline themes, the language of other taxonomies was consciously not utilised, in order to encourage the development of a new taxonomy which would classify the event categories purely from the narrative of the official accident/serious incident reports, and without interpretations or assumptions.

Once the baseline themes were identified within NVivo, as the template, it was then possible to begin creation of the taxonomy, named 'MxFACS' (Maintenance Factors Analysis and Classification System). This process initially involved separation of the node coding into a three-level hierarchy:
