Assessing the value of research results is long-known to be a difficult task. The problem lies in distinguishing between positive and negative results, because this demarcation always depends upon the underlying value system. Regardless of this philosophical difficulty, research assessments largely focus on positive results (i.e. positive impacts of research). For example, UK's 2014 Research Excellence Framework (REF) reported that an "impressive [array of] impacts were found from research in all subjects" (REF 2015); accordingly, no negative impacts were reported. This effectively marginalises negative results in favour of positive results, inadvertently deeming them 'valueless'. As a marketing strategy for research, it is indeed a powerful approach. However, as an objective scientific standard to justify what research gets funded (or not), due to this one-sided focus, it is less useful. However, negative results can be cognitively and sociologically extremely beneficial (cf. Pinker 2002, Taleb 2014). The paper explores the construction of REF's impact assessment in the case of tourism studies. We show that the impact criteria not only shift the emphasis on positive results, but also emphasize economic gains and short-term impacts. By unpacking the underlying values implicit within the REF, we propose a new socio-material approach that does not marginalise the value of negative results. By using Collins and Evans' (2007) notion of 'interactional expertise', we argue that the underlying value problem can be addressed sociologically.