“…However, as there is no fixed measure to define the consensus level (Becuwe et al, 2017), variable statistics were used for consideration of the consensus level as suggested by various studies (Richardson, de Leeuw, & Dullaert, 2016;Gracht, 2012;Harmsen et al, 2015;Tharisara, Kanchana, & Turner, 2016;Williams, 2016;Hasson, Keeney, & McKenna, 2000;Hsu and Sandford,2007). Following (Bulger & Housner, 2007;Green, Jones, Hughes, & Williams, 1999;Lohuis, van Vuuren, & Bohlmeijer, 2014;Neuer Colburn, Grothaus, Hays, & Milliken, 2016), this study uses a consensus level of four for five-point Likert scale questions. Thus, factors with the mean scores ̅ ≥ 4 were retained, while factors with the mean ̅ < 4 would be further evaluated in the second round.…”