The states were divided into groups by performance levels based on the difference in the proportion of students scoring at proficient levels and above between their state and the NAEP 2005 fourth grade reading assessments and into quartiles by states' per pupil spending. These groups were compared with each other and the NAEP to see if differences existed for any of the variables. Differences were identified between several groups of states and between the NAEP and state groups for several of the variables. Stepwise regression was used to determine if any of the independent variables predicted the difference in proficiency levels between the 2005 state fourth grade reading test and the NAEP. Three models were identified accounting for 34% to 43% of the variance of the difference in proficiency levels of the test scores.

The purpose of this study was to determine if differences exist between the NAEP and state fourth grade reading assessments. Specifically, the research questions focused on whether there were differences in text difficulty, text length, and depth of knowledge requirements between the state fourth grade reading tests as well as between the state tests and the NAEP. Sample text passages from 28 states as well as the NAEP were collected. The passages were analyzed for readability levels, text length, and whether or not introductions and illustrations were included with the text passage. The related test items were analyzed to determine whether they were multiple choice or constructed response and how they assessed comprehension by two scales, the four categories used by the NAEP, and five more traditional categories of comprehension.