I am studying US public company financial performance for the last 2 decades. I am searching for evidence of "sustainable superior performance", thus to isolate those companies that display such characteristics. I will use the Compustat database.
- Is the total population the universe of US public company with their past, present and future performance or the population is the subset of those displaying superior performance (subject of the study)?
- Because all the companies will be observed to spot those with certain characteristics (i.e. superior performance), what is the sample? Can I consider the company present in the Compustat database (independently if they have the traits I am seeking or not) to be the sample?
- What is the Instrument I need to ensure validity and reliability? -The Compustat database with its screening capabilities? -The criteria I will use to assess "sustained, superior performance"? (like sales growth above industry average and so on)
- The statistical technique used to find such outlier? (like Cluster Analysis for instance)