CONNECTICUT LINKING STUDY A Study of the Alignment of the NWEA RIT Scale with the Connecticut Mastery Test (CMT) March 2013 COPYRIGHT 2013 NORTHWEST EVALUATION ASSOCIATION Al l rights reserved. No part of this document may be reproduced or utilized in a ny form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without written pe rmission from NWEA.
A STUDY OF THE ALIGNMENT OF THE NWEA RIT SCALE WITH THE CONNECTICUT MASTERY TEST (CMT) MARCH 2013 Recently, NWEA completed a project to connect the scale of the Connecticut Mastery Test (CMT) used for Connecticut 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 7,559 Connecticut students who completed both exams in the spring of 2012. The Connecticut 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 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 CMT.
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 Below Basic Basic Proficient Goal Advanced 2 <173 173 7 180 18 187 38 198 70 3 <184 184 7 191 18 199 38 210 70 4 <192 192 7 199 17 208 37 219 68 5 <199 199 7 206 16 214 32 225 61 6 <201 201 6 210 17 220 36 232 66 7 <206 206 8 215 19 227 42 239 69 8 <207 207 6 218 18 230 40 246 74 READING - Current Season s and s for each State Performance Level Below Basic Basic Proficient Goal Advanced 2 <178 178 22 185 39 190 52 204 83 3 <188 188 22 195 39 200 52 213 83 4 <195 195 20 200 32 206 48 220 83 5 <202 202 23 206 33 211 46 225 81 6 <202 202 16 207 26 212 38 227 77 7 <203 203 12 208 21 212 30 227 69 8 <207 207 15 212 24 217 36 231 72 * 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.
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 Below Basic Basic Proficient Goal Advanced 2 <159 159 7 166 17 174 37 185 70 3 <173 173 6 180 17 188 37 198 68 4 <184 184 7 191 16 199 36 210 68 5 <192 192 7 199 16 206 31 217 61 6 <196 196 6 205 17 214 36 226 66 7 <202 202 8 211 19 222 42 234 69 8 <204 204 6 214 17 226 40 241 74 READING - Prior Season s and s for each State Performance Level Below Basic Basic Proficient Goal Advanced 2 <164 164 22 171 38 176 50 190 82 3 <178 178 21 185 37 190 50 204 83 4 <188 188 20 193 32 199 48 213 82 5 <196 196 22 201 33 205 44 219 80 6 <198 198 16 203 26 208 38 223 77 7 <199 199 11 205 21 209 30 223 68 8 <204 204 15 209 24 214 36 228 72 * 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.
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 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% 0% 0% 0% 0% 0% 150 5% 2% 1% 0% 0% 0% 0% 155 8% 3% 1% 1% 0% 0% 0% 160 12% 4% 2% 1% 1% 0% 0% 165 18% 7% 3% 2% 1% 1% 0% 170 27% 11% 5% 3% 2% 1% 1% 175 38% 17% 8% 4% 3% 2% 1% 180 50% 25% 13% 7% 5% 3% 2% 185 62% 35% 20% 11% 8% 5% 4% 190 73% 48% 29% 17% 12% 8% 6% 195 82% 60% 40% 25% 18% 12% 9% 200 88% 71% 52% 35% 27% 18% 14% 205 92% 80% 65% 48% 38% 27% 21% 210 95% 87% 75% 60% 50% 38% 31% 215 97% 92% 83% 71% 62% 50% 43% 220 98% 95% 89% 80% 73% 62% 55% 225 99% 97% 93% 87% 82% 73% 67% 230 99% 98% 96% 92% 88% 82% 77% 235 100% 99% 97% 95% 92% 88% 85% 240 100% 99% 98% 97% 95% 92% 90% 245 100% 100% 99% 98% 97% 95% 94% 250 100% 100% 99% 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% 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 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 the state test is 35%. Italics represent extrapolated data.
READING - 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 3% 1% 1% 0% 0% 0% 0% 155 5% 2% 1% 1% 1% 0% 0% 160 8% 3% 2% 1% 1% 1% 1% 165 12% 5% 3% 2% 1% 1% 1% 170 18% 8% 5% 3% 2% 2% 1% 175 27% 12% 8% 4% 4% 4% 2% 180 38% 18% 12% 7% 6% 6% 4% 185 50% 27% 18% 11% 10% 9% 6% 190 62% 38% 27% 17% 15% 14% 10% 195 73% 50% 38% 25% 23% 21% 15% 200 82% 62% 50% 35% 33% 31% 23% 205 88% 73% 62% 48% 45% 43% 33% 210 92% 82% 73% 60% 57% 55% 45% 215 95% 88% 82% 71% 69% 67% 57% 220 97% 92% 88% 80% 79% 77% 69% 225 98% 95% 92% 87% 86% 85% 79% 230 99% 97% 95% 92% 91% 90% 86% 235 99% 98% 97% 95% 94% 94% 91% 240 100% 99% 98% 97% 96% 96% 94% 245 100% 99% 99% 98% 98% 98% 96% 250 100% 100% 99% 99% 99% 99% 98% 255 100% 100% 100% 99% 99% 99% 99% 260 100% 100% 100% 100% 100% 99% 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 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 the state test is 35%. Italics represent extrapolated data.
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 2% 0% 0% 0% 0% 0% 0% 130 3% 1% 0% 0% 0% 0% 0% 135 4% 1% 0% 0% 0% 0% 0% 140 7% 2% 1% 0% 0% 0% 0% 145 11% 3% 1% 0% 0% 0% 0% 150 17% 5% 2% 1% 0% 0% 0% 155 25% 8% 3% 1% 1% 0% 0% 160 35% 12% 4% 2% 1% 1% 0% 165 48% 18% 7% 3% 2% 1% 1% 170 60% 27% 11% 5% 3% 2% 1% 175 71% 38% 17% 8% 5% 3% 2% 180 80% 50% 25% 13% 8% 4% 3% 185 87% 62% 35% 20% 12% 7% 5% 190 92% 73% 48% 29% 18% 11% 8% 195 95% 82% 60% 40% 27% 17% 13% 200 97% 88% 71% 52% 38% 25% 20% 205 98% 92% 80% 65% 50% 35% 29% 210 99% 95% 87% 75% 62% 48% 40% 215 99% 97% 92% 83% 73% 60% 52% 220 100% 98% 95% 89% 82% 71% 65% 225 100% 99% 97% 93% 88% 80% 75% 230 100% 99% 98% 96% 92% 87% 83% 235 100% 100% 99% 97% 95% 92% 89% 240 100% 100% 99% 98% 97% 95% 93% 245 100% 100% 100% 99% 98% 97% 96% 250 100% 100% 100% 99% 99% 98% 97% 255 100% 100% 100% 100% 99% 99% 98% 260 100% 100% 100% 100% 100% 99% 99% 265 100% 100% 100% 100% 100% 100% 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 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 spring season, her/his the state test is 52%. Italics represent extrapolated data.
READING - 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 4% 1% 0% 0% 0% 0% 0% 145 7% 2% 1% 0% 0% 0% 0% 150 11% 3% 1% 1% 0% 0% 0% 155 17% 5% 2% 1% 1% 1% 0% 160 25% 8% 4% 2% 1% 1% 1% 165 35% 12% 6% 3% 2% 2% 1% 170 48% 18% 9% 4% 4% 3% 2% 175 60% 27% 14% 7% 6% 5% 3% 180 71% 38% 21% 11% 9% 8% 5% 185 80% 50% 31% 17% 14% 12% 8% 190 87% 62% 43% 25% 21% 18% 13% 195 92% 73% 55% 35% 31% 27% 20% 200 95% 82% 67% 48% 43% 38% 29% 205 97% 88% 77% 60% 55% 50% 40% 210 98% 92% 85% 71% 67% 62% 52% 215 99% 95% 90% 80% 77% 73% 65% 220 99% 97% 94% 87% 85% 82% 75% 225 100% 98% 96% 92% 90% 88% 83% 230 100% 99% 98% 95% 94% 92% 89% 235 100% 99% 99% 97% 96% 95% 93% 240 100% 100% 99% 98% 98% 97% 96% 245 100% 100% 99% 99% 99% 98% 97% 250 100% 100% 100% 99% 99% 99% 98% 255 100% 100% 100% 100% 99% 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 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 spring season, her/his the state test is 48%. Italics represent extrapolated data.
TABLE 5 CORRELATION COEFFICIENTS BETWEEN MAP AND STATE TEST FOR EACH GRADE AND TEST SUBJECT Math Correlation Pearson's r Reading Correlation Pearson's r 3 0.825 0.810 4 0.849 0.825 5 0.867 0.840 6 0.875 0.824 7 0.894 0.816 8 0.897 0.832 * 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).
TABLE 6 PERCENTAGE OF STUDENTS WHOSE PASS STATUS WAS ACCURATELY PREDICTED BY THEIR MAP PERFORMANCE USING REPORTED CUT SCORES Mathematics Sample Size MAP Accurately Predicted State Performance MAP Underestimated State Performance MAP Overestimated State Performance 3 1021 90.1% 5.0% 4.9% 4 1255 88.3% 5.6% 6.1% 5 1382 91.2% 4.2% 4.6% 6 1380 87.9% 5.7% 6.4% 7 1152 88.3% 6.0% 5.7% 8 1228 90.2% 4.7% 5.0% Reading 3 1012 84.1% 7.8% 8.1% 4 1252 84.8% 7.6% 7.6% 5 1378 86.2% 6.0% 7.8% 6 1353 87.2% 6.7% 6.1% 7 1153 87.0% 6.3% 6.7% 8 1185 86.5% 6.7% 6.8% * 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%.