Boston Globe analysis of demographics and 2001 MCAS scores Summary: Which school districts score above or below the level that their demographics would predict? The Boston Globe explored this question using 2001 test scores on the Massachusetts Comprehensive Assessment System tests. The two articles are attached, and here is a quick summary of the analysis. Please send any comments or suggestions to Bill Dedman, dedman@globe.com. Files: These data files are available: Predicted scores and all demographic variables are at: http://PowerReporting.com/files/factors.sav (SPSS format) http://PowerReporting.com/files/factors.xls (Excel format) http://PowerReporting.com/files/factors.txt (text format) Just the predicted and actual scores are at: http://PowerReporting.com/files/mcas.xls (Excel format) This document, in Word format, is at http://PowerReporting.com/files/mcas.doc and in text format at http://PowerReporting.com/files/mcas.txt Data sources: The state Department of Education provided MCAS scores as well as certain demographic and financial variables about each school district. The variables were: spending per pupil: actspd99 99 $ actual spending edumedia 99 $ educational media instotal 99 $ instructional spending guidance 99 $ guidance counseling insttech 99 $ instructional technology perpup99 99 $ total integrated spending princoff 99 $ principal's office printech 99 $ principal's tech profdev 99 $ prof devel psych 99 $ psych local99 99 $ local contribution superv 99 $ instructional superv teaching 99 $ teaching salaries text_equ 99 $ text acquisition school size: stdt00 00 enrollment stdt01 01 enrollment stdt98 98 enrollment stdt99 99 enrollment attendance: g1_12att attendance rate 00 poverty: li990001 3-year running average lowinc 99-01 li989900 3-year running average lowinc 98-00 liper01 2001 lowinc pct lowinc00 2000 lowinc N lowinc01 2001 lowinc N lowinc98 1998 lowinc N parents' education: (from MCAS 2000: 1 = dropout, 2 grad, 3 some college, 4 bach degree, 5 grad degree) daded_1 dad's ed level 10 momed_1 mother's education level 10 pared_1 parents' ed level 10 race variables: aspct00 asian 00 pct blkpct00 black 00 pct hispct00 hispanic 00 pct natpct00 native american 00 pct w_as_per white and asian 00 pct whpct00 white 00 pct Background: Studies around the country have shown that children from poorer families, with less well educated parents, tend to do worse on standardized tests. Not must do worse, not always do worse, but tend to do worse. We test this connection in Massachusetts, we used a regression analysis to determine which factors, of those available to us, best predicted test scores. First, we looked at whether the data we had is suited to the task. What relation is there between poverty and test scores? between parents' education and test scores? Here is a scatter plot of parents' education and 10th grade English average scaled scores for districts. English scores are on the Y axis, or vertical axis, and parents' education is on the X axis, or horizontal axis. You'll see a relatively linear relationship: The higher the parents' education, the higher the scores, with some exceptions. (Scatter plots are available only in the Word version of this document.) (image) And here is a similar scatter plot for poverty and English scores, not quite so linear: (image) Doing a simple regression analysis, we looked at the extent to which a variable predicted the resulting scores. This prediction is measured by the R-square value. Results: Parents' education level explains about 70 percent of the differences in the average test scores of districts in English, about 72 percent in math. Poverty explains about 59 percent of the differences in English, 57 percent in math. If we combine both variables, together they explain about 76 percent in English, 79 percent of math. That's in 10th grade. Throw in all the race variables, too: All combined explains about 79 percent in English (hardly any change), and 79 percent of math (no change). Let's look at each variable, using only one variable at a time, to see how much it predicts the variations in the scores in 10th grade English. A value of, say, .571, means that that variable predicts 57 percent of the variation: spending per pupil: .132 actspd99 99 $ actual spending .176 edumedia 99 $ educational media .015 instotal 99 $ instructional spending .031 guidance 99 $ guidance counseling .000 insttech 99 $ instructional technology .002 perpup99 99 $ total integrated spending .007 princoff 99 $ principal's office .006 printech 99 $ principal's tech .001 profdev 99 $ prof devel .003 psych 99 $ psych .011 local99 99 $ local contribution .001 superv 99 $ instructional superv .008 teaching 99 $ teaching salaries .039 text_equ 99 $ text acquisition school size: .112 stdt01 01 enrollment attendance: .107 g1_12att attendance rate 00 poverty: .569 li990001 3-year running average lowinc 99-01 parents' education: (from MCAS 2000: 1 = dropout, 2 grad, 3 some college, 4 bach degree, 5 grad degree) .671 daded_1 dad's ed level 10 .657 momed_1 mother's education level 10 .686 pared_1 parents' ed level 10 race variables: .002 aspct00 asian 00 pct .086 blkpct00 black 00 pct .334 hispct00 hispanic 00 pct .021 natpct00 native american 00 pct .296 w_as_per white and asian 00 pct .255 whpct00 white 00 pct Schools: Now, to cases. Which schools perform much better or worse than the poverty and parents' education would predict? Most schools perform about how the demographics would predict. It's a bit subjective to set a cutoff point for calling a school an overachieving or underachieving school. One measure might be schools that score more than one standard deviation above the predicted score. That turns out to be about seven points. We agreed on a slightly more liberal cutoff, 5 points. We highlighted those schools that scored five or more points above or below the predicted score. In other words, every other school should consider this exercise a wash. It did about how its demographics would predict. Parents and educators are cautioned against making much of small differences. Here, FYI, are schools with differences of 5 points or more: In 10th-grade English (STD 6.36) Worse than expected Predicted Actual Difference District 228.2 214.0 -14.2 Champion of Brockton Charter 242.8 230.0 -12.8 South Shore Charter 240.4 231.0 -9.4 Swansea 236.8 229.0 -7.8 Webster 236.7 229.0 -7.7 City On A Hill Charter 252.8 246.0 -6.8 Francis W Parker Charter 244.7 238.0 -6.7 Rockport 239.4 233.0 -6.4 Cambridge 250.6 245.0 -5.6 Andover 248.4 243.0 -5.4 Holliston Better than expected: Predicted Actual Difference District 229.0 239.0 10.0 Lowell Middlesex Academy Charter 235.3 244.0 8.7 Wareham 237.4 245.0 7.6 Clinton 242.8 250.0 7.2 Stoneham 252.0 259.0 7.0 Wayland 243.2 250.0 6.8 Braintree 232.0 238.0 6.0 Everett 235.4 241.0 5.6 North Brookfield 232.8 238.0 5.2 Ware 227.9 233.0 5.1 Somerville 246.0 251.0 5.0 Shrewsbury In 10th-grade math (STD 6.94): Worse than expected: Predicted Actual Difference District 240.9 224.0 -16.9 South Shore Charter 224.9 212.0 -12.9 Champion of Brockton Charter 247.4 235.0 -12.4 Pioneer Valley Performing Arts Charter 238.3 229.0 -9.3 Swansea 241.7 234.0 -7.7 Somerville Charter 236.2 229.0 -7.2 Hull 243.0 236.0 -7.0 Rockport 244.9 238.0 -6.9 Milton 234.2 228.0 -6.2 City On A Hill Charter 252.0 246.0 -6.0 Francis W Parker Charter 239.6 234.0 -5.6 Seekonk 224.4 219.0 -5.4 Springfield Better than expected: Predicted Actual Difference District 224.7 233.0 8.3 Somerville 220.7 227.0 6.3 Chelsea 244.9 251.0 6.1 Newburyport 229.9 236.0 6.1 Ware 251.1 257.0 5.9 Wayland 234.3 240.0 5.7 Provincetown 236.4 242.0 5.6 Woburn 241.4 247.0 5.6 Braintree And for curiosity's sake, here are the predicted and actual scores for 10th-grade English and math for every district. Districts are marked "N/A" when the demographic values or scores were unavailable. PreE10 ActE10 DifE10 PreM10 ActM10 DifE10 ID District N/A N/A N/A N/A N/A N/A 445 Abby Kelley Foster Regional Charter 240.3 242 1.7 238.1 242 3.9 001 Abington N/A 251 N/A N/A 248 N/A 412 Academy of the Pacific Rim Charter N/A N/A N/A N/A N/A N/A 002 Acton N/A N/A N/A N/A N/A N/A 003 Acushnet 240.2 236 -4.2 238.0 235 -3.0 005 Agawam 239.9 242 2.1 237.7 241 3.3 007 Amesbury N/A N/A N/A N/A N/A N/A 008 Amherst 250.6 245 -5.6 249.5 248 -1.5 009 Andover 244.5 245 .5 242.8 245 2.2 010 Arlington 245.0 248 3.0 243.3 248 4.7 014 Ashland N/A N/A N/A N/A N/A N/A 491 Atlantis Charter 238.7 236 -2.7 236.4 235 -1.4 016 Attleboro 242.9 242 -.9 241.1 242 .9 017 Auburn 239.1 238 -1.1 236.8 233 -3.8 018 Avon 237.0 240 3.0 234.5 235 .5 019 Ayer 241.2 241 -.2 239.1 239 -.1 020 Barnstable 249.6 251 1.4 248.5 250 1.5 023 Bedford 243.7 243 -.7 241.9 242 .1 024 Belchertown 240.1 236 -4.1 237.9 234 -3.9 025 Bellingham 249.9 253 3.1 248.8 252 3.2 026 Belmont N/A N/A N/A N/A N/A N/A 420 Benjamin Banneker Charter N/A N/A N/A N/A N/A N/A 447 Benjamin Franklin Classical Charter N/A N/A N/A N/A N/A N/A 027 Berkley N/A N/A N/A N/A N/A N/A 028 Berlin 242.8 240 -2.8 241.0 240 -1.0 030 Beverly 241.3 240 -1.3 239.2 239 -.2 031 Billerica 228.6 229 .4 225.5 228 2.5 035 Boston N/A N/A N/A N/A N/A N/A 481 Boston Renaissance Charter 240.1 241 .9 237.9 238 .1 036 Bourne N/A N/A N/A N/A N/A N/A 037 Boxborough N/A N/A N/A N/A N/A N/A 038 Boxford N/A N/A N/A N/A N/A N/A 039 Boylston 243.2 250 6.8 241.4 247 5.6 040 Braintree N/A N/A N/A N/A N/A N/A 041 Brewster N/A N/A N/A N/A N/A N/A 043 Brimfield 233.7 236 2.3 231.0 230 -1.0 044 Brockton N/A N/A N/A N/A N/A N/A 045 Brookfield 249.5 246 -3.5 248.4 246 -2.4 046 Brookline 244.6 245 .4 242.8 244 1.2 048 Burlington 239.4 233 -6.4 237.3 233 -4.3 049 Cambridge 245.8 244 -1.8 244.2 245 .8 050 Canton N/A N/A N/A N/A N/A N/A 432 Cape Cod Lighthouse Charter N/A N/A N/A N/A N/A N/A 051 Carlisle 238.8 242 3.2 236.5 238 1.5 052 Carver 228.2 214 -14.2 224.9 212 -12.9 434 Champion of Brockton Charter N/A 245 N/A N/A 243 N/A 055 Chatham 247.6 244 -3.6 246.2 245 -1.2 056 Chelmsford 224.3 226 1.7 220.7 227 6.3 057 Chelsea 233.2 231 -2.2 230.3 228 -2.3 061 Chicopee 236.7 229 -7.7 234.2 228 -6.2 437 City On A Hill Charter N/A N/A N/A N/A N/A N/A 063 Clarksburg 237.4 245 7.6 235.0 239 4.0 064 Clinton 248.8 251 2.2 247.5 251 3.5 065 Cohasset N/A N/A N/A N/A N/A N/A 440 Community Day Charter N/A N/A N/A N/A N/A N/A 067 Concord N/A N/A N/A N/A N/A N/A 439 Conservatory Lab Charter N/A N/A N/A N/A N/A N/A 068 Conway 244.3 245 .7 242.6 244 1.4 071 Danvers 241.4 239 -2.4 239.3 239 -.3 072 Dartmouth 243.2 242 -1.2 241.3 240 -1.3 073 Dedham N/A N/A N/A N/A N/A N/A 074 Deerfield 240.2 240 -.2 238.0 237 -1.0 077 Douglas N/A N/A N/A N/A N/A N/A 078 Dover 240.0 241 1.0 237.7 239 1.3 079 Dracut 249.8 251 1.2 248.6 249 .4 082 Duxbury 242.8 244 1.2 240.8 240 -.8 083 East Bridgewater 246.2 243 -3.2 244.6 240 -4.6 087 East Longmeadow N/A N/A N/A N/A N/A N/A 085 Eastham 236.8 238 1.2 234.3 238 3.7 086 Easthampton N/A 249 N/A N/A 244 N/A 088 Easton N/A N/A N/A N/A N/A N/A 089 Edgartown N/A N/A N/A N/A N/A N/A 091 Erving 232.0 238 6.0 229.0 232 3.0 093 Everett 236.8 236 -.8 234.3 231 -3.3 094 Fairhaven 229.2 229 -.2 226.0 227 1.0 095 Fall River 242.7 243 .3 240.8 237 -3.8 096 Falmouth 233.0 234 1.0 230.2 229 -1.2 097 Fitchburg N/A N/A N/A N/A N/A N/A 098 Florida 245.0 245 .0 243.3 244 .7 099 Foxborough 242.8 243 .2 241.0 242 1.0 100 Framingham 252.8 246 -6.8 252.0 246 -6.0 478 Francis W Parker Charter 245.8 247 1.2 244.2 245 .8 101 Franklin N/A N/A N/A N/A N/A N/A 442 Frederick Douglass Charter N/A N/A N/A N/A N/A N/A 102 Freetown 238.1 241 2.9 235.7 240 4.3 103 Gardner 244.7 241 -3.7 243.0 244 1.0 105 Georgetown 238.3 237 -1.3 235.9 233 -2.9 107 Gloucester 244.1 243 -1.1 242.3 243 .7 110 Grafton 241.2 242 .8 239.1 238 -1.1 111 Granby N/A N/A N/A N/A N/A N/A 112 Granville 237.1 237 -.1 234.7 237 2.3 114 Greenfield 245.9 244 -1.9 244.3 247 2.7 117 Hadley N/A N/A N/A N/A N/A N/A 118 Halifax N/A N/A N/A N/A N/A N/A 121 Hancock 245.3 248 2.7 243.7 244 .3 122 Hanover 251.4 254 2.6 250.4 254 3.6 125 Harvard 241.8 244 2.2 239.8 241 1.2 126 Harwich 246.3 248 1.7 244.8 249 4.2 127 Hatfield 238.6 236 -2.6 236.3 235 -1.3 128 Haverhill 231.6 234 2.4 228.7 226 -2.7 452 Health Careers Academy Charter N/A N/A N/A N/A N/A N/A 450 Hilltown Cooperative Charter 249.5 251 1.5 248.3 248 -.3 131 Hingham 238.8 237 -1.8 236.5 233 -3.5 133 Holbrook N/A N/A N/A N/A N/A N/A 135 Holland 248.4 243 -5.4 247.1 245 -2.1 136 Holliston 225.9 226 .1 222.4 222 -.4 137 Holyoke 243.2 242 -1.2 241.3 241 -.3 138 Hopedale 249.1 250 .9 247.9 249 1.1 139 Hopkinton 243.2 242 -1.2 241.4 239 -2.4 141 Hudson 238.5 235 -3.5 236.2 229 -7.2 142 Hull 244.8 243 -1.8 243.2 243 -.2 144 Ipswich N/A N/A N/A N/A N/A N/A 145 Kingston N/A N/A N/A N/A N/A N/A 146 Lakeville N/A N/A N/A N/A N/A N/A 148 Lanesborough 223.5 223 -.5 219.8 221 1.2 149 Lawrence N/A N/A N/A N/A N/A N/A 454 Lawrence Family Development Charter 240.4 237 -3.4 238.3 235 -3.3 150 Lee 238.8 242 3.2 236.4 240 3.6 151 Leicester 247.1 251 3.9 245.7 250 4.3 152 Lenox 236.7 236 -.7 234.1 234 -.1 153 Leominster N/A N/A N/A N/A N/A N/A 154 Leverett 252.3 253 .7 251.5 252 .5 155 Lexington N/A N/A N/A N/A N/A N/A 157 Lincoln 246.0 248 2.0 244.4 244 -.4 158 Littleton 251.1 250 -1.1 250.1 249 -1.1 159 Longmeadow 230.8 233 2.2 227.8 231 3.2 160 Lowell N/A N/A N/A N/A N/A N/A 456 Lowell Community Charter 229.0 239 10.0 225.8 229 3.2 458 Lowell Middlesex Academy Charter 240.3 241 .7 238.1 240 1.9 161 Ludlow 242.2 247 4.8 240.2 243 2.8 162 Lunenburg 230.6 230 -.6 227.6 229 1.4 163 Lynn N/A N/A N/A N/A N/A N/A 460 Lynn Community Charter 247.6 249 1.4 246.2 245 -1.2 164 Lynnfield 235.0 235 .0 232.3 234 1.7 165 Malden 243.6 245 1.4 241.8 242 .2 167 Mansfield 249.7 251 1.3 248.6 247 -1.6 168 Marblehead N/A N/A N/A N/A N/A N/A 464 Marblehead Community Charter N/A N/A N/A N/A N/A N/A 169 Marion 241.0 241 .0 238.9 241 2.1 170 Marlborough 244.6 242 -2.6 242.9 238 -4.9 171 Marshfield 238.9 N/A N/A 236.5 N/A N/A 466 Martha's Vineyard Public Charter 239.8 237 -2.8 237.7 237 -.7 172 Mashpee N/A N/A N/A N/A N/A N/A 173 Mattapoisett 241.2 242 .8 239.1 241 1.9 174 Maynard 250.0 252 2.0 248.8 249 .2 175 Medfield 238.0 236 -2.0 235.7 232 -3.7 176 Medford 246.1 247 .9 244.6 243 -1.6 177 Medway 243.8 246 2.2 242.0 242 .0 178 Melrose 238.6 236 -2.6 236.2 234 -2.2 181 Methuen 239.6 239 -.6 237.4 233 -4.4 182 Middleborough N/A N/A N/A N/A N/A N/A 184 Middleton 242.9 241 -1.9 241.0 239 -2.0 185 Milford 237.5 240 2.5 235.0 235 .0 186 Millbury 244.7 245 .3 243.0 240 -3.0 187 Millis 246.4 244 -2.4 244.9 238 -6.9 189 Milton 239.7 242 2.3 237.5 236 -1.5 191 Monson N/A N/A N/A N/A N/A N/A 435 Murdoch Middle Public Charter N/A N/A N/A N/A N/A N/A 470 Mystic Valley Regional Charter N/A N/A N/A N/A N/A N/A 196 Nahant 243.3 242 -1.3 241.4 244 2.6 197 Nantucket 245.8 246 .2 244.3 244 -.3 198 Natick 250.6 252 1.4 249.6 251 1.4 199 Needham N/A N/A N/A N/A N/A N/A 444 Neighborhood House Charter 230.3 231 .7 227.2 230 2.8 201 New Bedford N/A N/A N/A N/A N/A N/A 471 New Leadership Academy Charter 246.4 248 1.6 244.9 251 6.1 204 Newburyport 250.8 252 1.2 249.7 253 3.3 207 Newton N/A N/A N/A N/A N/A N/A 208 Norfolk 235.3 236 .7 232.7 234 1.3 209 North Adams 246.9 248 1.1 245.5 245 -.5 211 North Andover 243.3 240 -3.3 241.4 238 -3.4 212 North Attleborough 235.4 241 5.6 232.7 236 3.3 215 North Brookfield 244.4 246 1.6 242.6 247 4.4 217 North Reading 243.5 248 4.5 241.7 245 3.3 210 Northampton N/A N/A N/A N/A N/A N/A 213 Northborough 238.4 239 .6 236.1 237 .9 214 Northbridge 240.4 244 3.6 238.2 239 .8 218 Norton 249.6 248 -1.6 248.4 248 -.4 219 Norwell 243.5 248 4.5 241.7 244 2.3 220 Norwood N/A N/A N/A N/A N/A N/A 221 Oak Bluffs N/A N/A N/A N/A N/A N/A 223 Orange N/A N/A N/A N/A N/A N/A 224 Orleans 237.4 234 -3.4 234.9 230 -4.9 226 Oxford 236.9 234 -2.9 234.4 230 -4.4 227 Palmer 240.3 237 -3.3 238.1 237 -1.1 229 Peabody N/A N/A N/A N/A N/A N/A 230 Pelham N/A N/A N/A N/A N/A N/A 231 Pembroke N/A N/A N/A N/A N/A N/A 234 Petersham 248.6 245 -3.6 247.4 235 -12.4 479 Pioneer Valley Performing Arts Charter 238.4 235 -3.4 236.1 233 -3.1 236 Pittsfield N/A N/A N/A N/A N/A N/A 238 Plainville 240.5 240 -.5 238.4 237 -1.4 239 Plymouth N/A N/A N/A N/A N/A N/A 240 Plympton 236.8 241 4.2 234.3 240 5.7 242 Provincetown 236.6 239 2.4 234.1 238 3.9 243 Quincy 239.3 238 -1.3 237.1 237 -.1 244 Randolph 247.3 246 -1.3 245.9 247 1.1 246 Reading 232.0 233 1.0 229.0 232 3.0 248 Revere N/A N/A N/A N/A N/A N/A 249 Richmond N/A N/A N/A N/A N/A N/A 483 Rising Tide Charter N/A N/A N/A N/A N/A N/A 482 River Valley Charter N/A N/A N/A N/A N/A N/A 451 Robert M. Hughes Academy Charter N/A N/A N/A N/A N/A N/A 250 Rochester 237.8 238 .2 235.4 237 1.6 251 Rockland 244.7 238 -6.7 243.0 236 -7.0 252 Rockport N/A N/A N/A N/A N/A N/A 253 Rowe N/A N/A N/A N/A N/A N/A 484 Roxbury Preparatory Charter N/A 245 N/A N/A 246 N/A 446 Sabis Foxborough Regional Charter 236.0 238 2.0 233.5 233 -.5 441 Sabis International Charter 234.8 231 -3.8 232.2 230 -2.2 258 Salem 245.3 248 2.7 243.6 244 .4 261 Sandwich 239.0 237 -2.0 236.7 239 2.3 262 Saugus N/A N/A N/A N/A N/A N/A 263 Savoy 246.5 245 -1.5 245.0 245 .0 264 Scituate 241.6 240 -1.6 239.6 234 -5.6 265 Seekonk N/A N/A N/A N/A N/A N/A 486 Seven Hills Charter 251.0 251 .0 250.0 252 2.0 266 Sharon N/A N/A N/A N/A N/A N/A 269 Sherborn N/A N/A N/A N/A N/A N/A 270 Shirley 246.0 251 5.0 244.4 249 4.6 271 Shrewsbury N/A N/A N/A N/A N/A N/A 272 Shutesbury 240.8 242 1.2 238.6 242 3.4 273 Somerset 227.9 233 5.1 224.7 233 8.3 274 Somerville 243.3 242 -1.3 241.7 234 -7.7 487 Somerville Charter N/A N/A N/A N/A N/A N/A 449 South Boston Harbor Academy Charter 241.5 240 -1.5 239.5 238 -1.5 278 South Hadley 242.8 230 -12.8 240.9 224 -16.9 488 South Shore Charter N/A N/A N/A N/A N/A N/A 275 Southampton N/A N/A N/A N/A N/A N/A 276 Southborough 233.1 238 4.9 230.3 234 3.7 277 Southbridge 227.7 224 -3.7 224.4 219 -5.4 281 Springfield 242.8 250 7.2 240.9 245 4.1 284 Stoneham 241.9 243 1.1 239.9 241 1.1 285 Stoughton N/A N/A N/A N/A N/A N/A 287 Sturbridge N/A 252 N/A N/A 246 N/A 489 Sturgis Charter N/A N/A N/A N/A N/A N/A 288 Sudbury N/A N/A N/A N/A N/A N/A 289 Sunderland 243.7 246 2.3 241.8 245 3.2 290 Sutton 247.4 250 2.6 246.0 249 3.0 291 Swampscott 240.4 231 -9.4 238.3 229 -9.3 292 Swansea 235.4 234 -1.4 232.8 233 .2 293 Taunton 242.1 240 -2.1 240.1 243 2.9 295 Tewksbury N/A N/A N/A N/A N/A N/A 296 Tisbury N/A N/A N/A N/A N/A N/A 298 Topsfield N/A N/A N/A N/A N/A N/A 300 Truro 244.4 245 .6 242.6 245 2.4 301 Tyngsborough 240.5 240 -.5 238.4 239 .6 304 Uxbridge 243.5 246 2.5 241.7 240 -1.7 305 Wakefield N/A N/A N/A N/A N/A N/A 306 Wales 246.0 245 -1.0 244.4 243 -1.4 307 Walpole 237.9 237 -.9 235.5 237 1.5 308 Waltham 232.8 238 5.2 229.9 236 6.1 309 Ware 235.3 244 8.7 232.7 237 4.3 310 Wareham 240.7 240 -.7 238.6 240 1.4 314 Watertown 252.0 259 7.0 251.1 257 5.9 315 Wayland 236.8 229 -7.8 234.4 231 -3.4 316 Webster 252.3 255 2.7 251.4 254 2.6 317 Wellesley N/A N/A N/A N/A N/A N/A 318 Wellfleet 243.4 243 -.4 241.5 241 -.5 322 West Boylston 239.2 239 -.2 236.8 241 4.2 323 West Bridgewater 238.9 235 -3.9 236.6 232 -4.6 332 West Springfield 250.2 251 .8 249.1 253 3.9 321 Westborough 239.0 236 -3.0 236.8 234 -2.8 325 Westfield 247.9 248 .1 246.6 251 4.4 326 Westford N/A N/A N/A N/A N/A N/A 327 Westhampton 252.0 254 2.0 251.1 253 1.9 330 Weston 237.5 235 -2.5 235.0 231 -4.0 331 Westport 249.0 253 4.0 247.8 249 1.2 335 Westwood 239.8 238 -1.8 237.6 237 -.6 336 Weymouth N/A N/A N/A N/A N/A N/A 337 Whately N/A N/A N/A N/A N/A N/A 340 Williamsburg N/A N/A N/A N/A N/A N/A 341 Williamstown 241.8 242 .2 239.7 240 .3 342 Wilmington 235.2 236 .8 232.5 232 -.5 343 Winchendon 251.5 252 .5 250.6 251 .4 344 Winchester 240.4 241 .6 238.2 235 -3.2 346 Winthrop 238.7 243 4.3 236.4 242 5.6 347 Woburn 232.0 229 -3.0 229.2 228 -1.2 348 Worcester N/A N/A N/A N/A N/A N/A 350 Wrentham Copyright 2001 Globe Newspaper Company The Boston Globe November 2, 2001, Friday ,THIRD EDITION Correction Appended SECTION: METRO/REGION; Pg. A1 LENGTH: 1091 words HEADLINE: MCAS RESULTS / GROWING IMPROVEMENT Sandy Coleman and Megan Tench of the Globe staff contributed to this article.; SCHOOLS MAKE BIG GAINS ON MCAS BYLINE: By Scott S. Greenberger, and Bill Dedman, GLOBE STAFF BODY: BROCKTON - Nearly every Massachusetts school district had a hand in the startling statewide MCAS gains that have cut failure rates almost in half, according to detailed results released yesterday. Scores went up and failure rates went down in almost every district, and in nearly all of them more than half of the students who took the 10th-grade exam passed both the English and math sections. But once again, wealthy suburbs such as Wellesley, Lexington, and Wayland posted the highest overall scores (three-time champion Harvard slipped to number eight), while poorer city schools landed at the bottom. That pattern, typical nationwide, has endured through four years of MCAS scores. A further Globe analysis shows that some districts rose above - or sank below - their demographic destinies, but parents' education and income are still better predictors of MCAS scores than teacher salaries or spending on computers or textbooks. Family background also trumps school size, attendance rates, or students' race. In Woburn, which did better in math than its demographic profile would predict, Superintendent Louise Nolan said her district didn't use any special MCAS-prep strategy to boost scores. "We go through the [state curriculum] frameworks with a fine- tooth comb, and we teach to the standards every single day," Nolan said. "Our kids want to do well." Everett, which outperformed expectations in English, took a different approach. Each week English and math teachers at Everett High School produced challenging questions that might appear on the MCAS, and on "MCAS Mondays" students took a crack at them in MCAS strategy courses. "We did make some drastic improvements," said Richard Wallace, Everett's associate superintendent. "With this, we were hoping to make a good student a better student and make a struggling student able to pass the exam." The goal of the state's eight-year-old reform effort is to close the gap between what students get in rich suburban districts - high standards, skilled teachers, and plenty of resources - and what they get in poorer urban ones. To do it, the state has revamped curriculum, stiffened standards, and pumped more than $7 billion into districts. And it has promoted MCAS as a way to hold students and schools accountable. State education leaders are well aware of the gap, which is why they chose Brockton High School - which doesn't fit the profile of an MCAS high achiever - to release the latest results in an event resembling a pep rally. Few of the 4,000 students at Brockton High School come from wealthy families and 60 percent are minorities. (Black and Hispanics have higher failure rates than whites.) But 51 percent of the junior class passed MCAS math and English on the 2001 exam, compared to only 26 percent in 2000. The class of 2003 is the first that will have to pass MCAS English and math to graduate. Those who failed will have at least four more chances to pass before the end of 12th grade. Standing in Brockton High's cavernous auditorium, State Board of Education chairman James A. Peyser saluted the school's juniors, comparing them to Brockton's favorite sons, boxing champions Rocky Marciano and Marvelous Marvin Hagler. "There were a lot of people out there who thought you couldn't do it, but you knew better," said Peyser, who is also Acting Governor Jane M. Swift's top education adviser. "There are hundreds of new champions in Brockton today, and they're all in this room." Education Commissioner David P. Driscoll has long predicted that MCAS scores would go up once students knew they had to pass to graduate. Several Brockton juniors said their friends in the senior class told them they hadn't tried too hard on the test. "I don't think we're smarter - I think we want to leave high school," Jeanne Elie said. "We don't want to stay." Others, however, credited their teachers for preparing them. Sandy Verneus said she passed "because of all the preparation they did for us, and the special classes and the after-school programs." But Verneus and other students also said that the focus on MCAS had crowded out other subjects, such as music and science. With Driscoll and Peyser standing uncomfortably beside him and the television cameras rolling, Earnest Offley said, "we were deprived of a lot of things." "Things were being skipped over that we really need to know, but they weren't on the test so we didn't do them," Offley said of his history class. "We were pushed into things that were on the test." The improved scores haven't impressed the Massachusetts Association of School Committees, which recently reaffirmed a resolution it passed last year calling for a delay in using the test as a graduation requirement. Some have also accused the state of downplaying the impact of scoring changes on the 2001 exam. The Department of Education recalibrated the 200-280 scoring scale this year, but it affected only students at the lowest and highest ends of the spectrum. While it changed scoring averages, officials say it did not impact failure rates. As far as teaching to the test, Driscoll said Achieve, Inc., a national education think tank, recently judged the Commonwealth's curriculum and test to be the best in the nation. He angrily rejected the suggestion that the scoring change is reponsible for the gains, calling it "an insult to our children." As the scores are publicized some high-flying districts may find themselves in the unusual position of defending results that are great - but perhaps not as great as expected. Andover reduced its failure rates in both English and math, but it performed worse than expected in English. Superintendent Claudio Bach said he's pleased with his district's improvement, even if its socioeconomics suggest it should be doing better. "If we aren't doing as well as we might, it's because we could use more staff at the high school," she said. "Like many districts, we have a very large high school and larger class sizes than we would like." Harvard Superintendent Mirhan Keoseian, who always insisted he didn't care about MCAS when his district had the highest scores in the state, stuck to that line yesterday when told that Harvard had dropped into a tie for eighth. "I am very happy with my teachers and my students," he said. "We are not MCAS-driven. We don't do MCAS hysteria." PLEASE REFER TO MICROFILM FOR CHART DATA. SIDEBAR 1: HIGHEST SCORES SIDEBAR 2: EXCEEDING EXPECTATIONS SIDEBAR 3: MOST IMPROVED SIDEBAR 4: MCAS ACHIEVMENT CORRECTION-DATE: November 3, 2001, Saturday CORRECTION: Because of a reporting error, the name of Andover Superintendent Claudia Bach was misspelled in a story that ran in the MCAS section Thursday. Database Manager's Note : Original story ran on Friday, November 2, 2001 GRAPHIC: PHOTO CHART, Brockton High School student Earnest Offley, speaking yesterday at an MCAS news conference, said students were denied "a lot of things." / GLOBE STAFF PHOTO / JOHN BLANDING LOAD-DATE: November 5, 2001 Copyright 2001 Globe Newspaper Company The Boston Globe November 2, 2001, Friday, THIRD EDITION SECTION: METRO/REGION; Pg. A34 LENGTH: 486 words HEADLINE: MCAS RESULTS / GROWING IMPROVEMENT WEIGHING THE RESULTS; SOME RESULTS FROM SCHOOLS DEFY DEMOGRAPHIC FACTORS BYLINE: By Bill Dedman, Globe Staff BODY: Districts with better-educated parents and fewer poor families have the highest average scores on the MCAS tests, according to a Globe analysis of results on the 2001 exam. Just those two factors account for more than 70 percent of the differences in average district scores on the MCAS tests in English and math, the Globe found. Factors outside the control of schools - parents' education and income - are much better predictors of scores than teacher salaries or spending on computers, textbooks, and psychologists, which affect scores hardly at all. Family background also explains more of the test results than school size, attendance rates, or students' race. But upbringing is not destiny. The Globe looked for school districts whose students performed better or worse than the family backgrounds of students would have predicted. With the field leveled, Wellesley falls from first to 35th. Its students do better than their predicted scores, by nearly three points in 10th-grade math and in 10th-grade English, but many districts jumped even higher - from where they started. The state Department of Education does not do this type of analysis. The Globe predicted MCAS scores for every district for 10th-grade math and English, by using data from the Department of Education on parental education and poverty rates. The figures on parents' education come from last year, as reported by 10th-graders taking the MCAS tests. Figures on poverty rates were reported by schools; the Globe used a three-year average of poverty rates from 1999 through 2001. While family background explains much of the test scores, some students score far higher than their family background would predict, and some far lower. Demographics and scores are linked, but higher income doesn't cause high scores. Leveling the playing field before comparing schools may give school boards and parents a tool for asking public policy questions: Which schools add the most value to the children who come to their schools? Are the highest-scoring districts really those that teach best, or are they teaching the best students to begin with? Here are factors, ranked from most influence to least on 10th-grade math scores. The computations use a simple, well-accepted statistical model called a regression; the values shown are the R-squared value, which measure the percentage of the variation explained by any one factor. The factors overlap, of course - for example, educated parents tend to make more money. .699 - parent's education level 2000; .693 - father's education level 2000; .659 - mother's education level 2000; .576 - percentage of parents who are poor (three-year average, 1999-2001); .308 - Hispanic percentage 2000; .257 - white race percentage 2000; .108 - actual school spending 1999; .082 - enrollment (size of school) 2001; .137 - attendance rate; .027 - spending on teaching salaries 1999. LOAD-DATE: November 2, 2001