ValidityPost-Publication Study DataTable 28:Validity Coefficients a of STAR Early Literacy with Itself at Later Time Pointsand with STAR Reading over a Two-Year Period for all Cohort GroupsCombined and SeparatelyGrade LevelGroupAdministrationTimeSTAR Early Literacy bSTAR Reading cWinter Spring Fall Corrected dTotal Sample Fall r e 0.75 0.69 0.64 0.71N f 1,265 1,382 1,740Winter r – 0.78 0.67 0.76N – 2,022 2,182Spring r – – 0.70 0.79N – – 2,810K to 2nd Fall r 0.52 0.50 0.47 0.56N 511 543 572Winter r – 0.64 0.57 0.70N – 922 880Spring r – – 0.59 0.73N – – 1,0951st to 3rd Fall r 0.66 0.56 0.54 0.67N 600 663 811Winter r – 0.68 0.62 0.76N – 881 1,012Spring r – – 0.66 0.78N – – 1,2552nd to 4th Fall r 0.56 0.58 0.51 0.61N 154 176 357Winter r – 0.75 0.55 0.64N – 219 290Spring r – – 0.58 0.68N – – 460a. All coefficients are statistically significant (p < 0.001).b. STAR Early Literacy Enterprise scores were taken within the same school year.c. STAR Reading was taken two-years after the fall administration of STAR Early Literacy.d. Corrected for measurement error.e. “r” indicates the validity coefficient rounded to two decimal places.f. “N” indicates the number of students used to calculate the validity coefficient.In summary, the data displayed in Tables 27 and 28 of this section provide supportfrom a substantial longitudinal study, for both the validity of STAR Early Literacyas a measure of developing skills, and for its long-term validity for predicting laterreading achievement.STAR Early Literacy<strong>Technical</strong> <strong>Manual</strong>76
ValidityConcurrent Validity of Estimated Oral Reading ScoreConcurrent Validity of Estimated Oral Reading ScoreDuring the fall of 2007 and winter of 2008, 25 schools across the United States thatwere using both STAR Early Literacy and DIBELS Oral Reading Fluency (DORF) forinterim assessments were contacted and asked to participate in research toprovide evidence supporting the validity of STAR Early Literacy’s Estimated OralReading Fluency (Est. ORF) score. The schools were asked to ensure that studentswere tested on both STAR Early Literacy and DORF within a 2-week time intervalduring September and January. In addition, schools were asked to submit fall,winter, and spring interim assessment data from the previous school year, and anystudent that had a valid STAR Early Literacy and DORF assessment within a 2-weektime span was used in the analysis. Thus, the research involved both a currentsample of students and also historical data from those same schools. No schoolsassessed 1st-grade students in the fall on the DIBELS passages, so there was no falldata for grade 1 in the analysis.The analysis was undertaken to estimate the extent to which the Est. ORF scoreson STAR Early Literacy accurately predicted the observed DIBELS Oral ReadingFluency scores. Both the Est. ORF score on STAR Early Literacy and DORF provideestimates of the students’ oral reading fluency expressed as the number of wordsread correctly within a minute (WCPM) on a grade-level-appropriate connectedtext passage. The Est. ORF score is an estimate based on the student’sperformance on STAR Early Literacy, while the DORF score is a direct measurefrom a set of standardized grade-level passages.Analysis was done on each grade independently because DORF passages areassigned to specific grade levels and therefore are not interpretable across grades.Within each grade, correlations between the DORF WCPM score and theunderlying STAR Early Literacy Rasch score for each student were calculated to getan estimate of the relation between the two measures.The 25 schools in the sample came from nine states: Alabama, Arizona, California,Colorado, Delaware, Illinois, Michigan, Tennessee, and Texas. This represented abroad range of geographic areas, and resulted in a large number of students(N = 3,221). The distribution of students by grade was as follows: 1st grade 2,028,2nd grade 729, and 3rd grade 464. The sample was composed of 61% of studentsof European ancestry; 21% of African ancestry; 11% of Hispanic ancestry; and theremaining 7% of Native American, Asian, or other ancestry. About 3% of thestudents were eligible for services due to limited English proficiency (LEP), andabout 14% were eligible for special education services.STAR Early Literacy<strong>Technical</strong> <strong>Manual</strong>77