Answer Base

I am thinking of doing my sports science undergrad thesis on the effects of alcohol on reaction time. the basic scenario is that i will perform an experiment on a sample of 6 females who will be recruited from a local gymnastics club. i will test the subject’s reaction time (measured in milliseconds) twice. prior to the first test, the subjects will not have ay alcohol at all. prior to the second test performed the following week, subjects will ingest 100 ml of whiskey. the reaction time test is based on the time it takes for the subjects to press a key on a keyboard in response to seeing an image change colour on a monitor. given the scenario, what do you see as the main threats to the internal and external validity and their possible solutions?. am thinking of a paired t test for analysis, does this seem appropriate also?

I would argue that problem to this design is issues of counterbalance. you say one week you will not give them alcohol and the nect week you will. i suggest you randomly split the participants into 2 groups: first group will have alcohol 1st week and no alcohol second week and for group 2 vice versa. i also worry that 6 participants is not ! for result to be significantly valid i suggest you use more. maybe at least 20.