“…Second, 16 of the studies relayed findings related to improvements in their participants’ BPSD. Thirteen of which used behavioural measures, 5 number of negative behaviours observed, social validation check, task-related error rate over time, number of actions or utterances made when engaging with device; ( Benveniste et al, 2012 ; Boulay et al, 2011 ; Cheng et al, 2019 ; Cheng & Lee, 2018 ; Cruz-Sandoval et al, 2018 ; Cunningham et al, 2019 ; Kwak et al, 2020 ; Lancioni et al, 2014 , 2015 , 2017 ; Manca et al, 2021 ; Murphy et al, 2018 ; Tyack et al, 2017 ) to indicate these changes in BPSD, two used cognitive measures 6 ( Azman et al, 2017 ; Han et al, 2020 ), while one other used a combination of cognitive and behavioural measures 7 ( Iacono & Marti, 2016 ). Almost all studies presenting behavioural evidence quantitatively described improvements as occurring along their behavioural measures, with reporting mixed on the types of statistics used, and whether corrections for multiple comparisons were employed.…”