Question #1: "Does reducing screen exposure two hours before bedtime decrease the time it takes to fall asleep?" would be best answered by an experiment. This question can be tested by manipulating one variable (screen exposure) and measuring its effect on another variable (time taken to fall asleep), allowing for a controlled setup and the possibility of establishing a cause-and-effect relationship.
Questions #2 and #3, while interesting, would be more suited for observational studies or surveys rather than experiments. Question #2 compares two age groups without manipulating variables, and Question #3 seeks an average measurement without an experiment to manipulate factors affecting sleep.