This preview shows page 1. Sign up to view the full content.
Unformatted text preview: 7. Three percent of the mathematics assessment at age 13 was also administered to both 9and 17-year-old students.
Figure TN-2. Percentage distribution of NAEP mathematics assessment questions administered at and across
student age groups: 2012 Age
NOTE: Detail may not sum to totals because of rounding. The 1973 Mathematics Results
The mathematics trend scale was developed in 1986 for all the assessment years up to that point.
Because the 1973 mathematics assessment had too few questions in common with the 1978, 1982, and
1986 assessments, results from the 1973 assessment were placed on the same 0 to 500 mathematics
scale using mean proportion correct extrapolation. Estimates were extrapolated from the data so that
average mathematics scores could be reported for the nation and by race/ethnicity and gender at all
The extrapolated estimates for each age level were obtained by assuming a linear relationship between
a student group’s average scale score and the logit transformation of the group’s average percentage
of correct responses. The same linear relationship was assumed to hold across assessment years and
student groups within an age level. For more information, see the Mathematics Data Analysis chapter
in Expanding the New Design: The NAEP 1985-86 Technical Report. Because of the need to extrapolate
the average scale scores, caution should be used in interpreting the pattern of trends across those
assessment years. SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2012 Long-Term Trend
Mathematics Assessment. 52 THE NATION’S REPORT CARD TECHNICAL NOTES Interpreting Statistical Significance
NAEP reports results using widely accepted statistical standards. Significance tests in the form of
t-tests are conducted to determine if estimates are significantly different from each other. Findings
are reported based on a statistical significance level set at .05 with appropriate adjustments for
multiple comparisons. Comparisons over time or between groups are based on statistical tests that
consider both the size of the differences and the standard errors of the two statistics being compared.
Standard errors are margins of error, and estimates based on smaller groups are likely to have larger
margins of error. The size of the standard errors may also be influenced by other factors, such as how
representative the assessed students are of the entire population.
When an estimate has a large standard error, a numerical difference that seems large may not
be statistically significant. Differences of the same magnitude may or may not be statistically
significant depending upon the size of the standard errors of the estimates. Standard errors for
the estimates presented in this report are available at http:/
When making multiple simultaneous comparisons, error rates are controlled to ensure that
significant differences in NAEP data reflect actual...
View Full Document
This document was uploaded on 02/28/2014.
- Spring '14
- The Land