Additionally, if we were to conduct the same analysis on states that spent less than $4,000 and more than $6,000, our regression analysis shows that the relationship is still weak regardless of the obvious increase in the R-square value and the decrease in the F-value. See table below.
(table)
In general, deleting the states with spending less than $4,000 a year and more than $6,000 per pupil would also not adversely affect the data. We have already established that between participating states and then between states spending specific amounts of money that the correlations are very weak. Since we know that other factors must be contributing to the relationship between the two variables, omitting and adding various states based on spending or scores will not change the strength of the relationship.
Our last step is to focus on the 13 non-participating states. The following is a list of the composite scores for the states that did not participate in the NAEP program. We must note that based on the insignificance of the relationship between the variables, these computed composite scores are not reliable. Regardless, we used the original equation from the participating sample, Y = 587.32 + 0.0087x to determine our composite score values. For example, to obtain the composite score for Idaho, we would apply the following equation: Y = 587.32 + 0.0087(3602). The composite score for Idaho is 619. The table below shows the rest of the composite scores for the 13 non-participating states.
Composite Scores for Non Participating States
(table)