Some Basic Components of Good Experiments That Involve Human Subjects
•Subjects are not to know the purpose(s) of the experiments.
•Subjects are to be randomly assigned to the control group or the treatment group(s).
•Subjects are not to know the experimental condition to which they belong.
•Infant consent forms are to be obtained from their parents or guardians.
•Variation (individual difference) within groups is to be regarded as error.
•Outliers may need to be trimmed, transformed, or replaced by less extreme values, determined by the principal experimenter.
•Means are unbiased estimates of true scores.
•Where the reliability is of concern, subjects may be assessed repeatedly. Where the development or change of behaviors is of concern, subjects may be assessed repeatedly as well.
•Data of subjects who do not follow the given instruction are to be discarded, determined by the principal experimenter.
•To control for potential sources of error, it is required that subjects are to be restricted in the laboratory during the entire seesion of the experiments.
•Power equal to or greater than .80 is always recommended.
文章定位: