I have been researching when to classify a dataset as population vs. sample for many hours now for a statistics project I am doing. I have a dataset of NFL combine data and one thing I need to clarify is whether the dataset represents a population or a sample. It contains data from every athlete from the past 20 years to attend the combine. My question is: Because the dataset does not contain all years of the combines existence, does this constitute the data set as a sample? Or because the sample is large enough(almost 5000 rows of data) is it able to be called a population data set. Thanks for any and all help!!
3
-
$\begingroup$ The key bit of information we need to answer your question is: what are you planning to do with the data? If all you will do is describe the athletes in your sample, it's a population. If you intend to draw inferences about any aspect of your data that might extend to other people at other times, then you don't have a population. $\endgroup$– whuber ♦Commented Mar 5, 2019 at 18:10
-
$\begingroup$ @whuber I only intend to compare them to the athletes within the time span I have. Judging by your response I am assuming the population is the way to go? $\endgroup$– RyanCommented Mar 5, 2019 at 18:42
-
$\begingroup$ It's still unclear, because it could depend on the nature of the comparison. For instance, if the comparison is based on measurements, will you account for the possibility of measurement error? If so, you have a sample of all possible measurements. You need to be specific about what you intend to do with these data. $\endgroup$– whuber ♦Commented Mar 5, 2019 at 18:52
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
It's not the size of the dataset that matters, it's what you want to do with it.
Really, it depends on what you want to try and learn from the data. Are you interested in questions about only the participants in the combine dataset? Then it is your population. Are you interested in inferring something about hypothetical participants, not in your dataset? Then it is a sample taken from the population about which you wish to infer something.
Take, for example, athletes who were in the combine before 20 years ago OR athletes who will be in the combine in the future. Since neither of those is in your dataset, if you want to make inferences that includes them then you'd treat your data as a sample from a larger population that includes the sample (observed) and out-of-sample (unobserved) participants.
