“…The presence of an out-group experimenter (e.g., men or Whites) might enhance the relevance of stereotype threat cues given to women or ethnic minorities in testing situations, whereas the presence of an in-group experimenter might create more credibility to a message of stereotype threat-removal. Although the predictive effect of experimenter expectancy bias might be small (e.g., less than 3% of the variance in standardized test results; Janssen, 1973), it is possible that African Americans or women might subconsciously experience discomfort with out-group experimenters according to research on in-group/out-group preference (e.g., Pedersen, Walker, & Glass, 1999). So far little evidence has been gathered to address these specific questions within a stereotype threat paradigm, although a few investigators safeguarded against experimenter effects by using experimenter(s) of the same race/ethnicity or gender as that of stereotype threat targets (e.g., Blascovich, Spencer, Quinn, & Steele, 2001, Study 1; Cadinu et al, 2003; Croizet & Claire, 1998; Nguyen, O'Neal, & Ryan, 2003), using double-blind procedures (e.g., Aronson et al, 1999; Quinn & Spencer, 2001; Shih et al, 1999), or minimizing experimenter presence in test administration (e.g., Inzlicht & Ben-Zeev, 2000, 2003).…”