Experiments with a good design and high internal validity can still have quite low external validity, so let's discuss this. Suppose that a fast food chain conducts a laboratory test of a new TV advertising campaign. The results indicate that 35% of the consumers who viewed the ad “will definitely purchase” during the following week, while only 15% of the consumers who did not view the ad “will definitely purchase” during the following week.
Let's say that the laboratory experiment was well designed and high in internal validity (i.e., exposure to the advertising truly had an effect on consumers’ intentions to purchase). What do you think about the external validity of this experiment? Should the company expect a similar change in purchase intentions in the target market, once it launches the TV advertising campaign at national level?