1: A selection bias that would threaten the internal validity of an experiment is

D: anything the experimenter does that indicates to the subjects how they should behave in the experiment

    Incorrect.

Choice D is referring to an experimenter expectancy effect.  An experimenter expectancy effect is where the experimenter consciously (less common) or unconsciously (more common) elicits certain behavior in the subject.  Experimenter expectancy effects ARE a threat to internal validity, but the question is asking for SELECTION BIASES that threaten internal validity.  Since experimenter expectancy effects are NOT selection biases, selection D is incorrect