6. Limitations

The representatives for the semi-structured interviews were selected by NORDHORCS, which introduces a selection bias, and the views and practices detailed by the representatives may not encapsulate all possible perspectives within each council or the broader academic and research communities within the Nordic countries. Also, the study relies on self-reported data obtained through the interviews and this data is subject to individual biases. There is also the inherent risk of misinterpretation or omission of details, especially since the interviews are conducted in a multi-lingual environment, which may influence the study’s conclusions. To mitigate this, all respondents were given the opportunity to fact-check and comment on this report before publication.

Regarding the methodology, the comparative analysis aims to systematically identify similarities and differences in research impact philosophies. However, this method relies on the assumption that the categories of comparison are equivalent across contexts, which may not account for the unique cultural and operational nuances of each Nordic country. Furthermore, as the interviews were conducted via virtual meetings, there may be limitations in communication, such as missing out on non-verbal cues, which can be significant in qualitative research, are harder to discern in virtual settings.

Further, AI did set a new dimension to consider, as its use is experimental. To mitigate the risk of AI generated mistakes or misinterpretations, all steps in the process were documented and all AI supported conclusions and findings were crosschecked by a human analyst.

Recognising these limitations is important for critically assessing the research and for understanding the scope within which the findings may be applied. Future research may address these limitations by, for example, incorporating quantitative data, broader participant selection, or through longitudinal studies that capture changes over time.