NAEP Technical DocumentationNonresponse Bias Analyses for the 2010 Assessment

NCES Statistical Standards call for a nonresponse bias analysis to be conducted for a sample with a response rate below 85 percent at any stage of sampling. Weighted school response rates for the 2010 assessment indicate a need for school nonresponse bias analyses for private school samples for grades 4, 8, and 12. Also, student nonresponse bias analyses are necessary for the grade 12 public school student sample. Both the school-level and student-level analyses were conducted separately by session type (history/geography and civics) at each grade. Thus, six separate school-level analyses and two separate student-level analyses were conducted.

The procedures and results from these analyses are summarized briefly below. The analyses conducted consider only certain characteristics of schools and students. They do not directly consider the effects of the nonresponse on student achievement, the primary focus of NAEP. Thus, these analyses cannot be conclusive of either the existence or absence of nonresponse bias for student achievement. For more details, please see the NAEP 2010 NRBA report (594KB).

Each school-level analysis was conducted in three parts. The first part of the analysis looked for potential nonresponse bias that was introduced through school nonresponse. The second part of the analysis examined the remaining potential for nonresponse bias after accounting for the mitigating effects of substitution. The third part of the analysis examined the remaining potential for nonresponse bias after accounting for the mitigating effects of both school substitution and school-level nonresponse weight adjustments. The characteristics examined were Census region, reporting subgroup (private school type), urban-centric locale, size of school (categorical), size of school (continuous), and race/ethnicity enrollment percentages.

Based on the school characteristics available, for the private school samples at grades 4 and 8, there does not appear to be evidence of substantial potential bias resulting from school substitution or school nonresponse. However, the analyses suggest that a potential for nonresponse bias remains for the grade 12 private school samples. This result is evidently related to the fact that, among non-Catholic schools, larger schools were less likely to respond. Thus, when making adjustments to address the underrepresentation of non-Catholic schools among the respondents, the result is to over-represent smaller schools at the expense of larger ones. The limited school sample sizes involved means that it is not possible to make adjustments that account fully for all school characteristics. Please see the full report for more details.

Each student-level analysis was conducted in two parts. The first part of the analysis examined the potential for nonresponse bias that was introduced through student nonresponse. The second part of the analysis examined the potential for bias after accounting for the effects of nonresponse weight adjustments. The characteristics examined were gender, race/ethnicity, relative age, National School Lunch Program eligibility, student disability (SD) status, English language learner (ELL) status, and subject (only for the history/geography session type).

At first glance, the analyses for the grade 12 public school students seem to indicate a potential for nonresponse bias due to the results for a handful of characteristics. However, the student sample sizes were large, making the tests very sensitive to small changes. In fact, for the characteristics in question, the absolute value of the potential bias was actually very small--one percent or less. Further, after nonresponse adjustments, the estimated absolute and relative biases were reduced in many cases. Based on the student characteristics available, for the grade 12 public school student samples, there does not appear to be evidence of substantial potential bias resulting from student nonresponse.

Last updated 09 October 2014 (GF)

Printer-friendly Version