Measurement matters. Addressing the importance of measurement in olfactory research

In our talk, we will address one often-neglected aspect of the credibility crisis: the measurement crisis. To make valid conclusions and avoid what some authors have called questionable measurement practices (QMP, Flake, and Fried, 2019), we have to use precise and accurate measures of operational variables. Common QMPs are for example i) the use of sum scores without testing the assumptions that underlie the use of sum scores (McNeish & Wolf, 2020) and ii) differential item functioning (DIF) and measurement invariance, which may invalidate the results from group comparisons. Given the extraordinary attention that the COVID-19 pandemic has posed on smell abilities, here we focus on the assessment of the abovementioned common QMPs in chemosensory research, by investigating the psychometric properties of a widely used test aimed to measure olfactory function, the Sniffin’ Sticks Test (Hummel et al., 1997). We investigated i) whether the Sniffin’ Sticks score (TDI), calculated as the sum of the scores of its three subtests (Threshold, Discrimination, Identification), is justified; and ii) whether items from the identification test showed some differential item functioning. With the caveat of a relatively small sample size, preliminary results from the ongoing data collection suggest that some, but not all, of the assumptions appear to be tenable. This suggests a promising validity and reliability of the Italian version of the Sniffin’ Sticks.