What is Standard Error of Measurement (SEM) and how is it used in score interpretation?

What is Standard Error of Measurement (SEM) and how is it used in score interpretation?

• A. SEM is the average error across test items and has little practical use in interpretation.
• B. SEM reflects score precision and is used to create confidence intervals around observed scores to estimate probable true scores.
• C. SEM measures the test's difficulty level.
• D. SEM indicates the maximum possible error for a single administration.

Answer: (B)

The main idea is that Standard Error of Measurement (SEM) tells you how precise a test score is. It represents the amount of random error that can be expected in a single observed score given the test’s reliability. In practice, SEM is used to create a confidence interval around an observed score to estimate where the true score would likely fall.

If you know the test’s reliability and its standard deviation, you can estimate SEM with SEM = SD × sqrt(1 − reliability). When reliability is high, SEM is small, meaning the observed score is a more precise estimate of the true ability. Conversely, lower reliability leads to a larger SEM and less precision.

Using SEM helps with interpretation: you can say the true score for a person is likely inside a range around the observed score (for example, roughly within ±1 SEM for about 68% confidence, ±2 SEM for about 95% confidence, assuming a normal distribution). This is also why change over time is interpreted with measurement error in mind—if the difference between two observed scores is greater than about two SEMs, it’s more likely to reflect real change rather than just error.

SEM isn’t about average item error, nor does it measure test difficulty or the maximum possible error on a single administration. It’s a measure of score precision and the basis for estimating probable true scores around each observed score.