Introduction to Statistical Terminology

Statistics - the process of collecting, analyzing and interpreting data through experimentation.

Data - facts and measurements that are collected through experimentation for the purpose of analysis.

Variable - any aspect of a study that can be measured.

Population - all individual members of a group.

Sample - a portion of a population that should represent the distribution of variables of the population.

Accuracy - refers to how close the measurement of a variable is to the true value of that variable. The goal of any measurement is to be as accurate as possible by ensuring that all measurements are taken carefully and correctly with the measuring tool.

Precision - refers to how close measurements are to one another. The goal of any experiment is to have high precision. If the measurement can be replicated not only by the experimenter but also by others, the measurement of the variable can be verified. 

Random Error - unpredictable aspects of an experiment, measuring device or interpretation of the measuring device that cause slight fluctuations in a measurement.  The fluctuations of the measurement then cause the variable to be artificially too high or too low. If the random error is small, all measurements should be relatively the same (the measurements will have high precision). Repeating the measurement of the variable several times allows for a mean of the variable to be determined as the slightly different measurements will average out, reducing the random error of the measurement.

Systematic Error - aspects of the experiment that have a strong impact on the accuracy of the measurement of a variable. The errors are usually the result of the experimenter not following a procedure correctly or consistently performing a measurement incorrectly (e.g., not measuring from the zero mark on a meter stick). If a measurement is always performed incorrectly, the measurement will always have low accuracy.

Statistical Significance - a claim that compares the sample data to an acceptable level of uncertainty to determine if the data is the result of a specific cause and not caused by chance. Two common methods to use the concept of statistical significance is during the analysis that occurs when two groups are compared to one another and during correlation studies.