In particular, researchers should evaluate tests in these terms before selecting them for use in their studies. At a minimum, measures should demonstrate reasonable internal reliability (Cronbach alpha coefficients of about > 0.7), and have evidence of construct validity (Kline, 2000).