Dependent Variable 2: Strength of Expressed Opinions on Questionnaire.

Variables

Direct Intervention -- Test subjects will be exposed to conversation at the dinner table directed to them and including them on three separate occasions one week apart about the importance of consuming alcohol responsibly and about the dangerous consequences of drinking irresponsibly.

Indirect Intervention -- Test subjects will be exposed to modeled behavior of adults expressing concern over matters such as designating a non-drinking driver and avoiding excessive intoxication without any conversation about them or directed to them three separate occasions one week apart.

Controls

There will be a control condition in the post-test version of the experimental design. Specifically, the responses of test subjects will be compared to a comparable control group that is not exposed to the specified intervention.

Manipulation Check

The parents will be instructed to either engage in verbal conversation directly with the test subjects or to refrain from any involvement with the test subjects while engaging in conversation with another adult about adults in their presence. The parents will be asked to confirm their adherence to this protocol after the fact. To ensure that the test subjects heard the conversation directed at them, they will be asked a direct question. To confirm that the subjects heard the conversations not directed at them, they will be engaged in a casual comment and glance in which the...

The test samples will be procured by soliciting parents directly in conjunction with a description of the purpose of the experiment and of the possible benefits and any other conceivable consequences to their children and their families.
External and Internal Validity

The principal threat to external validity is the fact that the results for the tested sample may not necessarily represent a wider pattern that is applicable to other groups of similarly chosen sample groups. External validity could be improved by repeating the experiment with other sample groups.

The principal threat to internal validity is the fact that the results may also reflect variables that are not accounted for besides the independent variable specifically tested for. Internal validity could be improved by repeating the experiment with other sample groups as well as by conducting a survey after the conclusion of testing to identify any conceivable extraneous variables that could have been reflected in the outcomes of the experiments.