#### Standard Error of the Mean

According to the **Central Limit Theorem**, if you take samples from the population to measure any phenomenon, and you take the average of those samples, the histogram of those averages would be shaped like a bell curve. That distribution has the same mean as the population. The standard deviation of that distribution is called the **Standard Error of the Mean.**

This **SEM decreases if you increase the sample size.** It will become zero if everyone in the population is measured. It is an indication of how well the sample (of samples) represents the population.

#### Standard Error of Estimate

A correlation coefficient represents how two variables are related in a linear fashion (if one changes, the other as well). A regression line is the line of best fit. **The stronger the relationship between two variables, the less Standard Error of Estimate**, and the closer the values to the line.

SEE is also used to determine the **Confidence Interval** around a predicted score for a criterion.

#### Standard Error of Measurement

When an instrument is used to to assess a particular aspect in a person, the score usually indicates some level of error. **The more reliable assessment will have a lower Standard Error of Measurement.** It means that if you repeat the measurement, the scores will be around the person’s true score.