只去除一个变数影响（作为常量）的偏相关。

• a partial correlation in which the effects of only one variable are removed (held constant).