Table of Contents  |  Search Technical Documentation  |  References

NAEP Weighting Procedures → Weighting Procedures for the 2000 State Assessment → Characteristics of Nonresponding Schools and Students → Characteristics of Schools Related to Response → Results of the Logistic Regression Analysis of School Nonresponse, Grade 4 Science State Assessment: By Jurisdiction, 2000

NAEP Technical DocumentationResults of the logistic regression analysis of school nonresponse, grade 4 science state assessment: By jurisdiction, 2000

State	School participation rate after substitution (percent)	Percent of population covered by model (percent)	Model with all variables			Significant variables	Test: Yij's=0
			Degrees of freedom	P-values			Degrees of freedom	P-values
Arizona	86.83	84.10	7	p=0.987		0	4	p=0.887
Arkansas	85.15	100.00	11	p=0.750		0	4	p=0.996
California	75.89	93.83	8	p=0.355		0	4	p=0.112	5
Idaho 1	74.87	90.51	6	p=0.030		nonresponse class 5 p=0.008	2	p=0.447
Illinois	73.10	67.91	6	p=0.927		0	4	p=0.753
Indiana	70.03	100.00	11	p=0.097	2	0	4	p=0.559
Iowa	70.60	92.75	7	p=0.120		percent Black p=0.025	4	p=0.215
Maine	85.49	83.24	8	p=0.967		0	4	p=0.898
Michigan	83.48	64.35	7	p=0.298	3	0	4	p=0.306
Minnesota	82.98	89.79	8	p=0.714		0	4	p=0.279
Montana	77.47	93.97	6	p=0.735		0	4	p=0.874
New York	72.16	100.00	7	p=0.507		0	4	p=0.397
North Dakota	88.91	75.73	9	p=0.180		0	4	p=0.369
Ohio	81.70	93.36	10	p=0.913	4	0	4	p=0.471
Oregon	74.18	86.79	8	p=0.393		percent Hispanic p=0.018	4	p=0.201
Vermont	74.66	100.00	7	p=0.800		0	4	p=0.682
1 The model for Idaho does not include the variables for percent Black or percent Hispanic since these variables were always zero. 2 Also ran the logistic regression in SAS due to unstable standard errors within the nonresponse classes; in SAS the model with all variables resulted in p<0.0001. However SAS generally tends to overstate the significance. 3 Also ran the logistic regression in SAS due to unstable standard errors within the nonresponse classes; in SAS the model with all variables resulted in p<0.0001. However SAS generally tends to overstate the significance. 4 Also ran the logistic regression in SAS due to unstable standard errors within the nonresponse classes; in SAS the model with all variables resulted in p<0.0001. However SAS generally tends to overstate the significance. 5 While the test was not significant (p=0.112) the p-value for the score statistic was significant with p=0.033. SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2000 Science Assessment.

Printer-friendly Version