“…The studies typically compare preference changes between two or more conditions with different experimental manipulations (e.g., choice importance, choice reversibility, etc. ; e.g., Brehm, 1956; Brehm and Cohen, 1959; Deutsch et al, 1962; Brock, 1963; Walster et al, 1967; Greenwald, 1969; Brehm and Jones, 1970; Brehm and Wicklund, 1970; Gordon and Glass, 1970; Walster and Walster, 1970; Converse and Cooper, 1979; Olson and Zanna, 1979, 1982; Gerard and White, 1983; Frey et al, 1984; Steele et al, 1993; Heine and Lehman, 1997; Lyubomirsky and Ross, 1999; Lieberman et al, 2001; Harmon-Jones and Harmon-Jones, 2002; Kitayama et al, 2004; Hoshino-Browne et al, 2005; Harmon-Jones et al, 2008; Imada and Kitayama, 2010; Lee and Schwarz, 2010). In these studies, results on relative difference in preference change between experimental conditions might be valid as the level of noise (and thus preference change explained by the artifact) should be no different across conditions as long as participants were randomly assigned into each experimental condition.…”