Guidelines for Analyzing Data
There are any number of calculations a dataset may require before you can extract meaningful results from it. The most important thing to keep in mind when analyzing data is what your question is. This will help you determine if statistics are necessary, and, if so, which statistical test is most appropriate.
When you have a taken a sample, but are trying to understand something at the population level, statistical tests are probably appropriate. Strictly speaking, statistics are just values calculated from sample data. Often, statistics are used to make inferences about patterns at the population level, by comparing them to meaningful probability distributions (like the bell curve). In other words, if your statistic (value calculated at the sample level) is highly unlikely to correspond to certain probability distribution (a hypothetical description of the population), you have learned something about the population(s) your samples come from.
Among other things, statistical tests can be used to infer things about the centrality and variability of a population or multiple populations. For example, a statistical test can tell you if it is more likely that two sets of samples come from one population with the same mean (a measure of centrality) or that they come from two populations with different means. The T-test is a common test used in the above situation.