Why have qualitative researchers historically been hesitant to adopt computational methods like NLP, and how have LLMs changed this?

Historically, qualitative researchers hesitated to adopt NLP because computational methods were seen as incompatible with interpretive analysis, and early NLP models lacked the ability to understand context and nuance. These limitations meant NLP was useful mainly for preprocessing large datasets, not for the close, context-sensitive reading qualitative scholars value. LLMs changed this because they can capture semantic meaning, subtle nuance, and context from vast unstructured texts, making them more compatible with inductive and interpretive qualitative approaches and shifting researcher acceptance.

Qualitative researchers traditionally prioritize close contact with data, attending to nuance, subtleties, and context. Computational methods were therefore perceived as incompatible with this interpretative orientation. Early NLP models, which operated on statistics, syntax, grammar rules, and dictionaries, could only identify general themes or predefined categories. As a result, qualitative researchers viewed them mainly as preprocessing tools for thematic analysis of large datasets, not as tools that could meaningfully support interpretive analysis. The limited capacity of these early models to grasp context and nuance reinforced the reluctance to adopt them. With recent advances in large language models, however, LLMs can capture semantic meaning, subtle nuance, and context across large unstructured text corpora. This overcomes the key limitation of earlier NLP models, allowing qualitative researchers to move beyond traditional automatic thematic analysis and explore new inductive and mixed-method approaches. Consequently, acceptance of these technologies has shifted, and some qualitative scholars now advocate for LLMs as rigorous methodological complements rather than replacements for interpretative work.