The LSAT, like any standardized test, is not a perfect measuring instrument. One way to quantify the amount of
measurement error associated with LSAT scores is through the calculation of the standard error of measurement. The
standard error of measurement provides an estimate of the average error that is present in test scores because of the
imperfect nature of the test. An error-free score, called a true score, could only be obtained from a hypothetical test
that contained no measurement error. The standard error of measurement is used to construct score bands, which are
used in score reports to quantify the uncertainty inherent in individual test scores.
Many factors besides measurement error can also affect an individual’s test performance on a particular day
(e.g., motivation, physical and mental health, or work and family responsibilities). These other factors are not
explicitly taken into consideration when calculating score bands.
What is a score band?
LSAT scores are estimates of a test taker’s actual proficiency level in the skills tested. Score bands represent a range
of scores that has a certain probability of containing the test taker’s actual proficiency level. The score bands reported for
the LSAT are designed to include the test taker’s actual proficiency level in approximately 68 percent of cases.
In other words, there is a 68 percent level of confidence that the test taker’s true score actually falls within the band.
How is a score band calculated for individual test scores?
Score-band calculations are based on the standard error of measurement. The standard error of measurement for the
LSAT is very stable, and tends to be about 2.6 scaled score points. A score band with a 68 percent confidence level
can be constructed by subtracting the standard error of measurement from the scaled score to obtain the lower value
and adding the standard error of measurement to the scaled score to obtain the upper value. Therefore, the width
of the score band is approximately 7 scaled-score points, after rounding.
The 68 percent (or approximately two out of three) level of confidence used by LSAC
for reporting purposes is a commonly used standard. To obtain a 95 percent level of confidence, the standard error
of measurement can be doubled before constructing the score band. Therefore, a 95 percent confidence band would
be approximately twice as wide as a 68 percent confidence band. Likewise, a 99 percent confidence band would be
approximately three times as wide as a 68 percent confidence band.
LSAC employs a more complicated calculation to accommodate scores that lie at the upper and lower extremes
of the LSAT score scale.
How is a score band calculated for the average test score?
The standard error of measurement is used in a similar way as that described for individual scores in calculating the
score bands for the average test score. However, there is less measurement error associated with an average score
than there is with a score earned on a single day of testing. The standard error of measurement is adjusted to take into
account the number of scores earned by the candidate in calculating the score band for an average score, resulting in a
somewhat more narrow band.
Does the test date (when I took the test) have an effect on the score band?
The standard error of measurement used to construct the score bands has been very stable from one LSAT
administration to the next. Because score bands are rounded to whole score points, your test date should have
little or no effect on the resulting score band.