OHIO LINKING STUDY A Study of the Alignment of the NWEA RIT Scale with the Ohio Achievement Assessment (OAA) December 2012 COPYRIGHT 2012 NORTHWEST EVALUATION ASSOCIATION All rights reserved. No part of this document may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without written permission from NWEA. 1
A STUDY OF THE ALIGNMENT OF THE NWEA RIT SCALE WITH THE OHIO ACHIEVEMENT ASSESSMENT (OAA) DECEMBER 2012 Recently, NWEA completed a project to connect the scale of the Ohio Achievement Assessment used for Ohio s mathematics and reading assessments with NWEA s RIT scale. Information from the state assessments was used in a study to establish performance-level scores on the RIT scale that would indicate a good chance of success on these tests. To perform the analysis, we linked together state test and NWEA test results for a sample of 3,860 Ohio students who completed both exams in the spring of 2012. The Ohio state test is administered in the spring. For the spring season (labeled current season ), an Equipercentile method was used to estimate the RIT score equivalent to each state performance level. For fall (labeled prior season ), we determined the percentage of the population within the selected study group that performed at each level on the state test and found the equivalent percentile ranges within the NWEA dataset to estimate the cut scores. For example, if 40% of the study group population in grade 3 mathematics performed below the proficient level on the state test, we would find the RIT score that would be equivalent to the 40 th percentile for the study population (this would not be the same as the 40 th percentile in the NWEA norms). This RIT score would be the estimated point on the NWEA RIT scale that would be equivalent to the minimum score for proficiency on the state test. Documentation about this method can be found on our website. Table Sets 1 and 2 show the best estimate of the minimum RIT equivalent to each state performance level for same-season (spring) and prior-season (fall) RIT scores. These tables can be used to identify students who may need additional help to perform well on these tests. Table Sets 3 and 4 show the estimated probability of a student receiving a proficient score on the state assessment, based on that student s RIT score. These tables can be used to assist in identifying students who are not likely to pass these assessments, thereby increasing the probability that intervention strategies will be planned and implemented. These tables can also be useful for identifying target RITscore objectives likely to correspond to successful or proficient performance on the state test. Table 5 shows the correlation coefficients between MAP and the state test in each grade. These statistics show the degree to which MAP and the state test are linearly related, with values at or near 1.0 suggesting a perfect linear relationship, and values near 0.0 indicating no linear relationship. Table 6 shows the percentages of students at each grade and within each subject whose status on the state test (i.e., whether or not the student met standards ) was accurately predicted by their MAP performance and using the estimated cut scores within the current study. This table can be used to understand the predictive validity of MAP with respect to the OAA. 2
TABLE SET 1 MINIMUM ESTIMATED SAME-SEASON (SPRING) RIT CUT SCORES CORRESPONDING TO STATE PERFORMANCE LEVELS MATH - Current Season s and s for each State Performance Level Grade Limited Basic Proficient Accelerated Advanced 2 <172 172 6 183 27 194 59 202 80 3 <183 183 6 195 27 206 59 214 80 4 <192 192 7 203 25 218 65 225 81 5 <209 209 21 215 34 227 66 233 79 6 <200 200 6 215 25 229 58 236 74 7 <204 204 7 219 26 241 73 248 84 8 <196 196 1 221 22 242 66 262 94 READING - Current Season s and s for each State Performance Level Grade Limited Basic Proficient Accelerated Advanced 2 <167 167 6 180 26 188 47 198 71 3 <177 177 6 190 26 198 47 207 71 4 <178 178 2 194 18 211 62 225 90 5 <192 192 8 203 26 223 77 230 89 6 <175 175 1 198 10 221 62 230 82 7 <189 189 2 207 19 225 64 233 82 8 <193 193 2 209 18 224 54 235 80 * Note: the cut scores shown in this table are the minimum estimated scores. Meeting the minimum MAP cut score corresponds to a 50% probability of achieving that performance level. Use the probabilities in Table Set 3 to determine the appropriate target scores for a desired level of certainty. Italics represent extrapolated data. 3
TABLE SET 2 MINIMUM ESTIMATED PRIOR-SEASON (FALL) RIT CUT SCORES CORRESPONDING TO STATE PERFORMANCE LEVELS MATH - Prior Season s and s for each State Performance Level Grade Limited Basic Proficient Accelerated Advanced 2 <158 158 6 170 26 181 59 189 80 3 <173 173 6 184 26 195 59 202 78 4 <184 184 7 195 25 209 65 215 80 5 <201 201 20 207 34 218 64 224 78 6 <196 196 6 209 25 222 56 229 73 7 <201 201 7 215 26 236 73 242 84 8 <193 193 1 217 22 237 66 257 94 READING - Prior Season s and s for each State Performance Level Grade Limited Basic Proficient Accelerated Advanced 2 <152 152 6 166 26 174 45 184 70 3 <167 167 6 180 25 188 45 198 71 4 <171 171 2 186 17 204 62 218 90 5 <187 187 8 198 26 217 76 224 88 6 <181 181 1 194 10 216 60 225 81 7 <188 188 2 204 19 221 63 229 81 8 <190 190 2 205 17 220 52 232 80 * Note: the cut scores shown in this table are the minimum estimated scores. Meeting the minimum MAP cut score corresponds to a 50% probability of achieving that performance level. Use the probabilities in Table Set 4 to determine the appropriate target scores for a desired level of certainty. Italics represent extrapolated data. 4
TABLE SET 3 ESTIMATED PROBABILITY OF SCORING AS PROFICIENT OR HIGHER ON THE STATE TEST IN SAME SEASON (SPRING), BY STUDENT GRADE AND RIT SCORE RANGE ON MAP ASSESSMENT MATH - Current Season Estimated Probability of Passing State Test Based on Observed MAP RIT Range 2 3 4 5 6 7 8 120 0% 0% 0% 0% 0% 0% 0% 125 0% 0% 0% 0% 0% 0% 0% 130 0% 0% 0% 0% 0% 0% 0% 135 1% 0% 0% 0% 0% 0% 0% 140 1% 0% 0% 0% 0% 0% 0% 145 2% 1% 0% 0% 0% 0% 0% 150 4% 1% 0% 0% 0% 0% 0% 155 6% 2% 1% 0% 0% 0% 0% 160 9% 3% 1% 0% 0% 0% 0% 165 14% 5% 2% 1% 1% 0% 0% 170 21% 8% 4% 1% 1% 1% 1% 175 31% 12% 6% 2% 2% 1% 1% 180 43% 18% 9% 3% 3% 2% 2% 185 55% 27% 14% 5% 5% 3% 3% 190 67% 38% 21% 8% 8% 5% 4% 195 77% 50% 31% 12% 12% 8% 7% 200 85% 62% 43% 18% 18% 13% 11% 205 90% 73% 55% 27% 27% 20% 17% 210 94% 82% 67% 38% 38% 29% 25% 215 96% 88% 77% 50% 50% 40% 35% 220 98% 92% 85% 62% 62% 52% 48% 225 99% 95% 90% 73% 73% 65% 60% 230 99% 97% 94% 82% 82% 75% 71% 235 99% 98% 96% 88% 88% 83% 80% 240 100% 99% 98% 92% 92% 89% 87% 245 100% 99% 99% 95% 95% 93% 92% 250 100% 100% 99% 97% 97% 96% 95% 255 100% 100% 99% 98% 98% 97% 97% 260 100% 100% 100% 99% 99% 98% 98% 265 100% 100% 100% 99% 99% 99% 99% 270 100% 100% 100% 100% 100% 99% 99% 275 100% 100% 100% 100% 100% 100% 100% 280 100% 100% 100% 100% 100% 100% 100% 285 100% 100% 100% 100% 100% 100% 100% 290 100% 100% 100% 100% 100% 100% 100% 295 100% 100% 100% 100% 100% 100% 100% 300 100% 100% 100% 100% 100% 100% 100% *Note: This table provides the estimated probability of passing the state test based on a MAP test score taken during that same (spring) season. Example: if a fifth grade student scored 200 on a MAP test taken during the spring season, her/his estimated probability of passing the state test is 18%. Italics represent extrapolated data. 5
READING - Current Season Estimated Probability of Passing State Test Based on Observed MAP RIT Range 2 3 4 5 6 7 120 0% 0% 0% 0% 0% 0% 0% 125 0% 0% 0% 0% 0% 0% 0% 130 1% 0% 0% 0% 0% 0% 0% 135 1% 0% 0% 0% 0% 0% 0% 140 2% 1% 0% 0% 0% 0% 0% 145 3% 1% 1% 0% 0% 0% 0% 150 5% 2% 1% 0% 1% 0% 0% 155 8% 3% 2% 1% 1% 1% 0% 160 12% 5% 3% 1% 2% 1% 1% 165 18% 8% 5% 2% 4% 1% 1% 170 27% 12% 8% 4% 6% 2% 2% 175 38% 18% 13% 6% 9% 4% 3% 180 50% 27% 20% 9% 14% 6% 5% 185 62% 38% 29% 14% 21% 10% 8% 190 73% 50% 40% 21% 31% 15% 13% 195 82% 62% 52% 31% 43% 23% 20% 200 88% 73% 65% 43% 55% 33% 29% 205 92% 82% 75% 55% 67% 45% 40% 210 95% 88% 83% 67% 77% 57% 52% 215 97% 92% 89% 77% 85% 69% 65% 220 98% 95% 93% 85% 90% 79% 75% 225 99% 97% 96% 90% 94% 86% 83% 230 99% 98% 97% 94% 96% 91% 89% 235 100% 99% 98% 96% 98% 94% 93% 240 100% 99% 99% 98% 99% 96% 96% 245 100% 100% 99% 99% 99% 98% 97% 250 100% 100% 100% 99% 99% 99% 98% 255 100% 100% 100% 99% 100% 99% 99% 260 100% 100% 100% 100% 100% 100% 99% 265 100% 100% 100% 100% 100% 100% 100% 270 100% 100% 100% 100% 100% 100% 100% 275 100% 100% 100% 100% 100% 100% 100% 280 100% 100% 100% 100% 100% 100% 100% 285 100% 100% 100% 100% 100% 100% 100% 290 100% 100% 100% 100% 100% 100% 100% 295 100% 100% 100% 100% 100% 100% 100% 300 100% 100% 100% 100% 100% 100% 100% *Note: This table provides the estimated probability of passing the state test based on a MAP test score taken during that same (spring) season. Example: if a fifth grade student scored 200 on a MAP test taken during the spring season, her/his estimated probability of passing the state test is 43%. Italics represent extrapolated data. 6
TABLE SET 4 ESTIMATED PROBABILITY OF SCORING AS PROFICIENT OR HIGHER ON THE STATE TEST IN PRIOR SEASON (FALL), BY STUDENT GRADE AND RIT SCORE RANGE ON MAP MATH - Prior Season Estimated Probability of Passing State Test Based on Observed MAP RIT Range 2 3 4 5 6 7 8 120 1% 0% 0% 0% 0% 0% 0% 125 1% 0% 0% 0% 0% 0% 0% 130 2% 0% 0% 0% 0% 0% 0% 135 3% 1% 0% 0% 0% 0% 0% 140 5% 1% 0% 0% 0% 0% 0% 145 8% 2% 1% 0% 0% 0% 0% 150 12% 3% 1% 0% 0% 0% 0% 155 18% 5% 2% 1% 0% 0% 0% 160 27% 8% 3% 1% 1% 0% 0% 165 38% 13% 5% 1% 1% 1% 1% 170 50% 20% 8% 2% 2% 1% 1% 175 62% 29% 12% 4% 3% 2% 1% 180 73% 40% 18% 6% 5% 3% 2% 185 82% 52% 27% 10% 8% 5% 4% 190 88% 65% 38% 15% 13% 8% 6% 195 92% 75% 50% 23% 20% 12% 10% 200 95% 83% 62% 33% 29% 18% 15% 205 97% 89% 73% 45% 40% 27% 23% 210 98% 93% 82% 57% 52% 38% 33% 215 99% 96% 88% 69% 65% 50% 45% 220 99% 97% 92% 79% 75% 62% 57% 225 100% 98% 95% 86% 83% 73% 69% 230 100% 99% 97% 91% 89% 82% 79% 235 100% 99% 98% 94% 93% 88% 86% 240 100% 100% 99% 96% 96% 92% 91% 245 100% 100% 99% 98% 97% 95% 94% 250 100% 100% 100% 99% 98% 97% 96% 255 100% 100% 100% 99% 99% 98% 98% 260 100% 100% 100% 100% 99% 99% 99% 265 100% 100% 100% 100% 100% 99% 99% 270 100% 100% 100% 100% 100% 100% 100% 275 100% 100% 100% 100% 100% 100% 100% 280 100% 100% 100% 100% 100% 100% 100% 285 100% 100% 100% 100% 100% 100% 100% 290 100% 100% 100% 100% 100% 100% 100% 295 100% 100% 100% 100% 100% 100% 100% 300 100% 100% 100% 100% 100% 100% 100% *Note: This table provides the estimated probability of passing the state test based on a MAP test score taken during that same (fall) season. Example: if a fifth grade student scored 200 on a MAP test taken during the fall season, her/his estimated probability of passing the state test is 33%. Italics represent extrapolated data. 7
READING - Prior Season Estimated Probability of Passing State Test Based on Observed MAP RIT Range 2 3 4 5 6 7 120 1% 0% 0% 0% 0% 0% 0% 125 2% 0% 0% 0% 0% 0% 0% 130 3% 1% 0% 0% 0% 0% 0% 135 4% 1% 1% 0% 0% 0% 0% 140 7% 2% 1% 0% 0% 0% 0% 145 11% 3% 2% 0% 1% 0% 0% 150 17% 5% 3% 1% 1% 0% 0% 155 25% 8% 4% 1% 2% 1% 1% 160 35% 12% 7% 2% 3% 1% 1% 165 48% 18% 11% 4% 5% 2% 2% 170 60% 27% 17% 6% 8% 3% 3% 175 71% 38% 25% 9% 13% 5% 5% 180 80% 50% 35% 14% 20% 8% 8% 185 87% 62% 48% 21% 29% 13% 12% 190 92% 73% 60% 31% 40% 20% 18% 195 95% 82% 71% 43% 52% 29% 27% 200 97% 88% 80% 55% 65% 40% 38% 205 98% 92% 87% 67% 75% 52% 50% 210 99% 95% 92% 77% 83% 65% 62% 215 99% 97% 95% 85% 89% 75% 73% 220 100% 98% 97% 90% 93% 83% 82% 225 100% 99% 98% 94% 96% 89% 88% 230 100% 99% 99% 96% 97% 93% 92% 235 100% 100% 99% 98% 98% 96% 95% 240 100% 100% 100% 99% 99% 97% 97% 245 100% 100% 100% 99% 99% 98% 98% 250 100% 100% 100% 99% 100% 99% 99% 255 100% 100% 100% 100% 100% 99% 99% 260 100% 100% 100% 100% 100% 100% 100% 265 100% 100% 100% 100% 100% 100% 100% 270 100% 100% 100% 100% 100% 100% 100% 275 100% 100% 100% 100% 100% 100% 100% 280 100% 100% 100% 100% 100% 100% 100% 285 100% 100% 100% 100% 100% 100% 100% 290 100% 100% 100% 100% 100% 100% 100% 295 100% 100% 100% 100% 100% 100% 100% 300 100% 100% 100% 100% 100% 100% 100% *Note: This table provides the estimated probability of passing the state test based on a MAP test score taken during that same (fall) season. Example: if a fifth grade student scored 200 on a MAP test taken during the fall season, her/his estimated probability of passing the state test is 55%. Italics represent extrapolated data. 8
TABLE 5 CORRELATION COEFFICIENTS BETWEEN MAP AND STATE TEST FOR EACH GRADE AND TEST SUBJECT Grade Math Correlation Pearson's r Reading Correlation Pearson's r 3 0.817 0.766 4 0.815 0.738 5 0.816 0.725 6 0.811 0.764 7 0.801 0.769 8 0.799 0.780 * Note: Correlations range from 0 (indicating no correlation between the state test score and the NWEA test score) to 1 (indicating complete correlation between the state test score and the NWEA test score). 9
TABLE 6 PERCENTAGE OF STUDENTS WHOSE PASS STATUS WAS ACCURATELY PREDICTED BY THEIR MAP PERFORMANCE USING REPORTED CUT SCORES Grade Mathematics Sample Size MAP Accurately Predicted State Performance MAP Underestimated State Performance MAP Overestimated State Performance 3 638 88.7% 6.0% 5.3% 4 643 88.5% 4.7% 6.8% 5 561 86.6% 6.2% 7.1% 6 670 87.8% 5.7% 6.6% 7 644 89.4% 4.7% 5.9% 8 676 86.8% 7.4% 5.8% Reading 3 636 84.4% 8.3% 7.2% 4 641 87.7% 6.2% 6.1% 5 562 84.9% 7.7% 7.5% 6 667 87.4% 6.6% 6.0% 7 632 87.7% 6.5% 5.9% 8 687 87.3% 6.6% 6.1% * Note: The third column of this table shows the percentage of students whose Pass/NotPass status was predicted accurately when their state test score was linked to their MAP score based on this linking study. The fourth column shows the percentage of students whose MAP score predicted they would not pass the state benchmark but they did pass. The last column shows the percentage of students whose MAP score predicted they would pass the state benchmark but they did not pass. Due to rounding, percentages may not add to 100%. 10