Results of an experiment may be biased if all individuals in one treatment group (e.g. the control group) are not treated _____except for the application of the actual treatments.