A fishbowl draw (or a similar method of physically randomizing the selection of numbers from the population contained in the table, such as dropping a coin from high above an enlarged version of the table and allowing it to bounce onto selected numbers) would be preferred for this study. This means of selection would fully randomize the selection, ensuring the validity of the sample, while at the same time would be fairly easy to accomplish. The creation or utilization of a computer program to randomly select numbers form this population would also be a highly valid and reliable way of accomplishing the necessary sampling, but this would be much more time-intensive as it would require the inputting of the entire population into the computer before the sample could be created. Analysis would be...

The population could be stratified based on factors such as the starting digit or whether the number is odd or even, for example, but such divisions would be meaningless and would not require separate evaluation (assuming the sample population was created at random, of course). Simple sampling is not only easier but is more appropriate in this case.