To ensure that your results are valid, consider the following guidelines when you collect data, perform the analysis, and interpret your results.
In statistics, random samples are used to make generalizations, or inferences, about a population. If your data are not collected randomly, your results may not represent the population. For more information, go to Randomness in samples of data.
For observations to be independent, the probability of a particular outcome does not depend on any previous outcome. For example, if you flip a coin twice and record whether heads or tails is face up, the outcome of the second flip does not depend on the outcome of the first flip. If your observations are not independent, your results may not be valid. For more information, go to What are independent trials?.