As part of their socialization into the scientific community, students are introduced to language norms detailing what constitutes successful written communication at university level (Duff, 2010). One notion that is often brought up is that of formality, and students are asked to make their writing “sufficiently formal”. However, this is not always an easy task, given the elusive nature of this concept. 

This study aims to investigate to what extent factors such as register (journal articles vs. academic blog posts), prescriptive and descriptive linguistic features with two variants (e.g., split vs. non-split infinitives) affect students’ perceptions of formality. To investigate this, 117 undergraduate college students were presented with a series of binary choices that enabled a systematic comparison of all possible combinations of two registers and six linguistic features, resulting in a rank-ordered list showing which combinations of register and feature are perceived as most formal.

The results show that the combination perceived as most formal was journal articles with a formal variant, followed by journal articles with an informal variant, blog posts with a formal variant, and blog posts with an informal variant (p=0.011; Cramer’s V: 0.039). This trend was especially strong for the proscribed features exclamation points and contractions. However, the distributions for other features went against this overall trend. For example, split infinitives (to correctly interpret) were perceived as more formal than the non-split variant (to interpret correctly) regardless of register, which suggests that the effect of proscriptive norms might not be as stable as previously thought, and may in fact be changing. In bringing us one step closer to understanding how students think about formality, this study can help to inform both EAP theory and practice.