Overview
A Mixed-Methods Approach
This research project employed a mixed-methods approach, combining qualitative and quantitative data collection and analysis techniques to provide a holistic understanding of the AI landscape. '
This approach, inspired by the principles of Research-Driven Innovation (RDI) and established mixed-methods methodologies, aimed to move beyond surface-level observations and uncover more nuanced insights. Specifically, we utilized an exploratory sequential design, prioritizing qualitative exploration to inform subsequent, targeted inquiries.
Our methodology incorporated the following key elements:
Qualitative Data Collection (Primary): We conducted 42 in-depth, semi-structured interviews with a diverse group of AI stakeholders: primarily AI researchers, along with startup founders and industry experts. This qualitative phase was central to our exploratory approach, allowing us to gather rich, contextualized data about participants' experiences, challenges, and perspectives.
Quantitative Data Collection (Secondary/Contextual): While our primary focus was qualitative, we drew upon existing quantitative data sources (industry reports, published research statistics) to provide context and, where relevant, to corroborate or contrast with our qualitative findings.
Triangulation: We used a triangulation protocol, gathering data from multiple sources (researchers, founders, experts) and using different data types (interview transcripts, industry reports) to cross-validate our findings and ensure the robustness of our conclusions.
Data Analysis and Synthesis:
Mixed Methods Matrix: We organized and synthesized our findings, initially primarily qualitative insights from the interviews.
Whiteboarding: Initial thematic analysis was conducted through collaborative whiteboarding sessions, visually mapping out connections and identifying emergent themes.
Cluster Analysis: We employed manual cluster analysis to systematically group and regroup related findings, identifying higher-level patterns and non-obvious relationships within the data. This inductive process helped us move from specific interview data points to broader, more generalizable insights.
Iterative Refinement: The research process was iterative, with initial findings informing subsequent interview questions and analysis, allowing us to progressively refine our understanding of the key issues.