The State Released 2026 Test Scores. Where D38 Finished Among Its Peers.

The Colorado Department of Education released its 2026 test results on Aug. 19, and the tables that follow show every Tri-Lakes school in them. But a ranking of raw scores is close to a ranking of household income, and the state’s own data makes that plain: across the 1,148 Colorado schools that report both figures, the share of students eligible for free or reduced-price meals explains 63 percent of the variation in reading proficiency. Measured that way, the two schools at the extremes of Lewis-Palmer School District 38 are not the ones the raw table names. Palmer Lake Elementary, which posts the lowest reading score of the district’s neighborhood elementaries, teaches the poorest student body in the district and beats what that composition predicts by nine points. Monument Academy, which teaches the least disadvantaged student body of any school in the comparison, falls twenty-one points short of the same prediction — a gap wider than all but 29 of the 1,148 Colorado schools the state measures both ways.

What the state actually measures

Three separate tests sit inside the August release, and they do not measure the same children.

CMAS English language arts and mathematics covers grades three through eight. Every D38 elementary appears in it, along with Lewis-Palmer Middle School and both Monument Academy campuses. No high school does.

CMAS science is given in grades five, eight and eleven, which is the only place Lewis-Palmer High School and Palmer Ridge High School appear in the CMAS files at all.

PSAT and SAT are the high school assessments proper — PSAT in grades nine and ten, SAT in grade eleven. They were released the same day.

What the number in every table means. Colorado places each student’s result in a performance level and publishes the share reaching the top two — met expectations and exceeded expectations. CMAS uses five levels: did not yet meet, partially met, approached, met, exceeded. The PSAT and SAT use four, without the partially-met band. Either way the published figure is the same idea, and it is the state’s equivalent of “proficient,” though Colorado does not use that word.

Two things about that percentage are easy to misread. Its denominator is the students who sat the test and received a valid score, not the students enrolled — which is why a participation rate is printed beside it wherever the state supplies one. And it applies to the PSAT and SAT as well as to CMAS: Colorado scores those tests against its own expectations levels rather than against the College Board’s separate college-readiness benchmark, so an 84 in the PSAT tables is not a percentage of students judged college-ready. Where the state withheld a figure to protect student privacy, the tables show a dash. A dash is not a zero.

Two cautions, before the tables

Rankings carry less than readers ask of them. A curriculum leader in the region, asked about composition-adjusted measures for this article, said he puts no stock in any of them: much of what is published rests on subjective choices, and the methods make the outcomes suspect. He has never been a fan of rankings in education, he said, because so many factors outside a school’s control feed the result. His example was Colorado’s first-place finish in U.S. News‘s 2026 education table, which folds K-12 together with higher-education completion and the share of Colorado adults holding degrees — including the ones who earned them elsewhere and moved here. The Independent set out that breakdown in August: first overall, third on the higher-education half, ninth on the Pre-K-12 half.

That caution applies to what follows as much as it does to U.S. News. The tables below fit one variable across one year. They are a way of reading the state’s numbers, not a score.

And none of this is an indictment of D38. By the state’s own measures this is one of Colorado’s most advantaged communities: a median household income of $160,525 against $95,470 statewide, and the third-lowest share of students eligible for free or reduced-price meals of the 147 Colorado districts that report one — behind only Aspen and Douglas County. It funds its schools like one of the least advantaged. Counting the state’s formula and the local override together, D38 raises $11,769 a student — fourth-lowest of Colorado’s 179 districts, and $2,733 behind Cheyenne Mountain. The district’s own finance office reaches the same conclusion, ranking D38 last of the sixteen districts it treats as peers. D38 collects a fifth of the override state law would allow it. Cheyenne Mountain collects 97 percent. Most of the variation in Colorado school results tracks circumstances a district does not choose, and the tables below record where schools finished rather than how hard anyone worked.

Why The Classical Academy appears in these tables

The Classical Academy is a charter school in Academy District 20 rather than in D38, and on a map it sits outside the Tri-Lakes. In practice the district line is more porous than that.

In the October 2025 count, 704 students whose parents live inside Lewis-Palmer School District 38 attended school in Academy 20 — more than went to every other district in the state combined, and 59 percent of all D38 residents attending school outside their own district. Academy 20 sends 391 students the other way, leaving a net outflow of 313 from a district that enrolls 6,318 in total. The next largest destination, an online BOCES program, takes 255.

The state’s file records the district a student attends, not the school, so it does not establish how many of those 704 sit at The Classical Academy specifically. What can be said is that Academy 20 is overwhelmingly where Tri-Lakes families go when they choose a school outside D38, that TCA is Academy 20’s largest charter, and that its northern campuses are the ones closest to Monument. Its results belong in these tables because a substantial number of local children are in them.

The finding: composition first

Colorado publishes results by student group, which makes it possible to calculate what share of each school’s tested students is eligible for free or reduced-price meals — the standard, if crude, measure of school poverty. Set that against reading proficiency for every Colorado school and the relationship is the strongest in the data by some distance.

Two entries on that list deserve attention because both are commonly offered as explanations locally and neither survives.

Class size does not predict results. The small positive correlation above is a well-known artifact — affluent suburban districts tend to run larger classes. Hold poverty constant and the relationship disappears entirely, to a correlation of −0.02. On the state’s own numbers, how many children sit in a Colorado classroom tells you nothing about how they score.

Nor does the share of students who sit the test. Opt-outs are a live subject in higher-income districts, and D38 leaves more of its students unscored than either of its peers. But across the state there is no relationship between participation and results, and once poverty is held constant the relationship is mildly positive: schools that test more of their students score slightly better, not worse. The participation figures in this article are published because a percentage should carry the number of children behind it, not because they explain a result.

What does carry independent weight, beyond poverty, is chronic absenteeism — a student missing a tenth or more of the school year. It remains a real signal after poverty is controlled for, and it is the one measure on this list that separates the local schools from each other. It returns below.

Where each school lands against its own composition

The state relationship predicts a reading score for any school from its poverty share alone. The gap between that prediction and the actual result is the useful number.

Monument Academy sits at the third percentile. Of the 1,148 Colorado schools with both figures, only 29 fall further below what their student composition predicts. Its secondary campus sits at the fourth.

The comparison that makes this concrete is with schools like it. There are 129 Colorado schools where fewer than 15 percent of students are eligible for free or reduced-price meals. Their median reading score is 70.2 percent. Monument Academy posts 45.3 and its secondary campus 47.1. Those are raw scores rather than adjusted ones, and only about one school in twenty in that group scores lower.

This forecloses the most natural objection to the charter’s numbers, which is that it serves a harder population than the schools around it. It does not. On the state’s measure it serves the least disadvantaged student body in the district — 8.2 percent eligibility against a Palmer Lake figure of 40.6 and a state median of 43.6.

Palmer Lake Elementary is the inverse. Ranked by raw score it is the weakest of D38’s neighborhood elementaries in reading. It also teaches, by a wide margin, the district’s poorest student body: four times Prairie Winds’ eligibility rate and nearly five times Monument Academy’s, and close to the statewide middle. Adjusted for that, it is the 81st percentile in the state in reading, and beats its predicted mathematics result by 13.2 points, among the strongest figures in the region. A ranking by raw score reverses both schools.

One caution, stated plainly. These are correlations drawn from a single year, and a gap between prediction and result describes a difference rather than diagnosing a cause. Nothing here identifies why any school lands where it does. A charter’s intake is also self-selected: Monument Academy’s eligibility rate reflects the families who apply and enroll, not a catchment, and families who choose a school differ from those who do not in ways this measure does not capture. That cuts both ways — it is the reason the low figure is meaningful, and the reason it is not the whole picture.

The results, by subject and level

Every table below is ranked by the 2026 result, highest first.

CMAS English language arts — districts

D38’s reading result has fallen three years running and is now the lowest in the series the Independent has published. Its lead over the state narrowed from 14.4 points to 12.7. Of the three districts the paper has tracked since the 2025 accreditation downgrade, D38 was the only one to lose ground.

CMAS English language arts — elementary schools

Prairie Winds and Bear Creek posted the two largest gains in the district. The Classical Academy’s elementary reading has now fallen two years running, from 74.6 to 64.4, on participation that has barely moved — the region’s highest-performing charter moved further than any D38 school.

CMAS English language arts — middle schools

Lewis-Palmer Middle recorded the largest single-school decline in the region. Nearly all of it sits in one grade: seventh-grade reading fell from 71.6 to 56.2, while eighth grade moved 54.6 to 52.2.

CMAS mathematics — districts

D38’s mathematics result rose, but by less than the state and by a quarter of Academy 20’s gain. The two districts were level at 51.9 percent a year ago.

CMAS mathematics — elementary schools

CMAS mathematics — middle schools

Lewis-Palmer Middle’s mathematics rose while its reading fell nine points, which argues against a general decline at the school and for something specific to reading.

CMAS science — districts

Science was D38’s strongest showing in the release and the one least likely to be reported, because it sits outside the headline files. Eighth-grade science rose 11 points to 60.7, more than 25 points clear of the state, and within 1.6 points of Cheyenne Mountain — the only measure anywhere in the release where D38 comes close to its strongest peer.

Grade-11 science carries a caveat that runs in D38’s favor. Roughly half of Colorado’s eleventh-graders sit the test at all. D38 tested 64.3 percent of its juniors, against 40.5 percent in Cheyenne Mountain and 25.9 percent in Academy 20. The three districts’ scores are closer than those participation rates make comparable, and D38 reached its figure from much the broadest slice of its students.

CMAS science — elementary schools (grade 5)

Grade-five science is the least stable column in the release. These schools tested between 37 and 222 children each, and the swings run 30 points in both directions. Ray E. Kilmer’s fall came on 37 tested students. None of these single-school movements should be read as a verdict on a school.

CMAS science — middle schools (grade 8)

CMAS science — high schools (grade 11)

Lewis-Palmer High School’s 2026 science figure is withheld by the state, as is its 2024. What the state does publish for the school is a mean scale score of 742, twelve points above the state average of 730, and a bottom band containing 16.4 percent of its tested juniors against 38.5 percent statewide. It also tested the largest share of its juniors of any school in the table.

SAT grade 11, reading and writing — districts

SAT grade 11, mathematics — districts

D38’s high schools sit far above the state on both halves of the SAT — 18.1 points clear in reading and 17.6 in mathematics — and its mathematics result recovered 5.3 points after a sharp fall the previous year, overtaking Academy 20.

SAT grade 11, reading and writing — high schools

SAT grade 11, mathematics — high schools

Palmer Ridge recorded the largest single-school gain in the set, 7.2 points in mathematics, after an 11-point fall the year before. Lewis-Palmer High gained 4.3. Both remain below their 2024 figures.

Monument Academy Secondary’s SAT rests on eighteen students. It tests every one of them, the highest participation rate in the table, but a cohort that size will swing sharply from year to year on the movement of one or two children and should not be read as a measure of the school.

PSAT reading and writing — districts

PSAT mathematics — districts

The ninth and tenth grades moved in opposite directions in D38. Ninth-grade mathematics gained seven points, the largest district movement anywhere in the PSAT set, against a state gain of three. Tenth grade fell in both subjects while the state and both peer districts rose in both.

Absence, and the school where it concentrates

The one factor besides poverty that carries independent weight is chronic absenteeism — the share of students missing a tenth or more of the school year, excused and unexcused together. Statewide it explains 44 percent of the variation in reading results on its own, and a meaningful share after poverty is accounted for. The state median is 24.3 percent.

Chronic absenteeism — elementary schools

Chronic absenteeism — middle schools

Chronic absenteeism — high schools

Every local school sits below the state median, most of them far below. The two that come closest are the district’s middle-grade campuses, and Lewis-Palmer Middle is the highest of any school in the region.

That places three facts at the same school in the same year: a quarter of its students chronically absent, a reading result 7.4 points below what its composition predicts, and a fifteen-point fall in seventh-grade reading. The data cannot say whether they are connected. The district is in a position to know.

The high schools carry the other figure worth noticing. Lewis-Palmer High reports 4.4 percent of its students chronically absent; Palmer Ridge reports 21.9 percent — five times the rate, in the same district, in the same year. Chronic absenteeism usually rises through the grades, which makes Lewis-Palmer’s figure the unusual one rather than Palmer Ridge’s. It is what the state publishes for the school, and the two schools’ attendance practices are worth asking about before too much weight is placed on either number.

The conditions the results were produced in

Test scores are the output of a staffing arrangement, and the Independent set out D38’s last month. The state’s staffing files were published before these results and are not part of the August release, so the figures below are not new. What is new is the company they keep.

Staff turnover, 2025-26

Average salary, 2025-26

Monument Academy pays its teachers $19,163 less than the district-run schools it is measured against, and roughly $10,000 less than Academy 20’s charters. That is a fact about the school with the weakest composition-adjusted result in the region, and it is worth stating alongside that result. It is not an explanation of it: this article has no evidence bearing on what any individual school’s pay does to its scores, and the series has tested and rejected simpler versions of that argument before.

The teacher gap is closing. D38’s district-run teacher salary rose 5.3 percent this year, the largest gain of the three and against 0.2 percent in Cheyenne Mountain. The gap to Cheyenne Mountain has narrowed from $11,571 to $8,484 in a single year, and the gap to Academy 20 from $7,401 to $6,061. Teacher turnover has fallen in step, from a peak of 27.6 percent three years ago to 18.8 percent now, and the district’s teacher headcount grew again, from 382 to 393.

The building-leader figures have not moved the same way. D38 is the only one of the three districts paying its principals under $100,000, and it sits $44,256 below Cheyenne Mountain. Seven of its 23 principals left over the year, against one of Cheyenne Mountain’s 13. With numbers that small the comparison cannot carry statistical weight — a single departure moves D38’s rate more than four points and Cheyenne Mountain’s nearly eight — so the counts are given rather than the ranking. What the longer run supports is firmer. As the Independent reported last month from the same state file, 30.4 percent is the highest principal turnover D38 has recorded in the seven years CDE publishes, and its seven-year average of 26.2 percent runs close to double Academy 20’s 14.2 percent and Cheyenne Mountain’s 15.3 percent. Teacher turnover improved; this did not.

None of this explains a test score. Pay and turnover are conditions, not causes, and the series has been careful not to claim otherwise — an earlier article tested and rejected the idea that regional pay differences track how hard a district is to work in. But Lewis-Palmer Middle is the school in this release with the weakest adjusted result and the highest absence rate, and it sits in the district that pays its principals least and replaces them fastest. Those facts belong on the same page.

What the numbers support

Read as a league table, the 2026 results say roughly what a map of household income says. Sixty-three percent of the variation in Colorado school reading scores tracks a single measure of family circumstance, and the schools near the top of the state’s table are, with few exceptions, the schools whose families have the most. That is worth saying plainly at the start of every discussion of test scores in a district like this one, because a high score in Monument is partly a fact about Monument.

Measured against that baseline, the local picture is both better than the raw table suggests and more specific. In mathematics, five of D38’s six neighborhood schools beat the result their student composition predicts. Palmer Lake Elementary, which finishes last of the six on the raw numbers, beats its prediction by nine points in reading and thirteen in mathematics — the strongest showing in the district, and one a ranking by score conceals completely. Bear Creek gained six points in reading and fifteen in fifth-grade science. Prairie Winds gained seven in reading and nineteen in third grade. These are the schools most Tri-Lakes children attend, and they are getting more out of their students than the scores alone indicate.

Two results run the other way, and both are particular rather than districtwide. Monument Academy teaches the least disadvantaged student body in the district and posts a reading result that fewer than three percent of Colorado schools fall further below their own prediction. That holds at both campuses, and it is not explained by who enrolls there. Lewis-Palmer Middle has a quarter of its students chronically absent, the highest rate of any school in the region, and a seventh-grade reading result that fell fifteen points in a single year.

The release also disposes of the two explanations most often reached for locally. Class size has no relationship to Colorado school results once family circumstance is held constant. Neither does the share of students who sit the tests; if anything the schools that test the most of their children do slightly better. Whatever accounts for the two outliers, it is not crowded classrooms and it is not opt-outs.

A single year of scores cannot identify a cause, and this one does not. What it can do is direct attention, and it directs it at two schools rather than at a district. Growth data — the measure of how far a child moved rather than where they finished — is the metric that would answer the question these tables can only pose, and it arrives later in the accountability cycle. It is also the measure on which D38 lost its “Accredited with Distinction” rating in 2025. When it lands, it will be worth reading against everything above.

Until then, four things sit with the district and the two charters rather than with the files: why Monument Academy tests barely half its secondary students, what happened in one grade at Lewis-Palmer Middle, why two high schools in the same district report chronic absenteeism rates that differ fivefold, and what The Classical Academy makes of a ten-point fall in elementary reading across two years. The state’s files record where every school finished. Why each landed there is a local question, and it has local answers.

Sources & further reading

The Independent’s own coverage: Blame the Buildings? What D38’s Test Scores Actually Show — the 2025 results and the apartment question · A Quiet Downgrade: How D38 Lost Its “Distinction” — the growth measure this release does not contain · Four Million Dollars, Frozen — the levy and teacher turnover · Safety Does Not Set Teacher Pay in El Paso County — and what the “#1 in education” ranking actually measures · Who We Are, by the Numbers — the district’s income, age and education profile

Assessment results: CDE — CMAS Data and Results · CDE — PSAT and SAT Data and Results

Attendance, enrollment and staffing: CDE — Attendance Statistics · CDE — Pupil Membership Statistics · CDE — School/District Staff Statistics

Entities: Lewis-Palmer School District 38 · Monument Academy · The Classical Academy · Academy District 20 · Cheyenne Mountain School District 12

Methodology

All results are from the Colorado Department of Education’s district and school files released Aug. 19, 2026, read directly rather than from any summary. Colorado assigns every result a performance level and the percentages here are CDE’s own “Percent Met or Exceeded Expectations” column, which combines the top two levels over the number of valid scores. CMAS carries five levels (did not yet meet, partially met, approached, met, exceeded); the PSAT and SAT carry four, without the partially-met band. That scale is applied to the PSAT and SAT as well as to CMAS; the College Board’s college-readiness benchmark is a separate measure carried in a different field of the state file and is not used anywhere in this article. Prior-year columns are CDE’s own, carried inside the 2026 files. State figures were checked against a separate CDE state-summary workbook and matched at every grade.

The composition analysis is the Independent’s, not the state’s. CDE publishes results disaggregated by student group; the share of each school’s tested students eligible for free or reduced-price meals was calculated from that file as eligible divided by eligible plus not eligible, among students with valid scores, and verified against raw counts. That share was set against 2026 reading proficiency across the 1,148 Colorado schools reporting both, producing the relationship used to predict each school’s result. The residual standard deviation is 11.1 points, so differences smaller than about five points are within normal variation and are described here as performing about as predicted. Correlations for chronic absenteeism, attendance, truancy, school size and pupil–teacher ratio were calculated the same way from CDE’s attendance and staffing files; partial correlations hold the poverty share constant. Attendance and absenteeism figures are 2024-25, one year behind the assessment results. The district ranking cited earlier in the article was calculated the same way from the district rows of the same file: of the 147 Colorado districts reporting both counts, Lewis-Palmer 38 has the third-lowest eligible share, at 16.3 percent. That rank does not move whether or not the smallest districts are included. Funding figures are from CDE’s FY2025-26 District Funding Calculation Worksheet, updated Jan. 14, 2026: total program per-pupil funding plus the voter-approved override, divided by each district’s funded pupil count. On the state formula alone, before any override, both Academy 20 and Cheyenne Mountain receive less per pupil than D38; the money that separates the three is the local override. D38’s own management reporting reaches the same ranking on a slightly different basis — school finance act, override and half of Impact Aid — and places the district last of sixteen self-selected peers. That document adds that Pueblo County 70 is the only district in the state it outranks; on CDE’s figures, three districts fall below D38 rather than one, the difference being the Impact Aid adjustment.

Eligibility for free or reduced-price meals is a coarse measure of family circumstance and is not comparable across years, because federal pandemic meal waivers and a change in how Colorado certifies eligibility both interrupted the series. The comparison here is between schools within a single year under identical collection rules, which is the use the measure supports. A charter school’s intake is self-selected and is not equivalent to a neighborhood catchment.

Correlation is not causation. A gap between a predicted and an actual result describes a difference; it does not identify a cause, and nothing in a single year of cross-sectional data can. Small cohorts are noted where they matter: Monument Academy Secondary’s grade-11 SAT rests on 18 students, and the grade-five science figures on 37 to 66 students per school.

Staffing figures are from CDE’s 2025-26 personnel turnover and average salary workbooks, the most recent available; both compare December 2024 with December 2025 and were read directly for this article rather than carried from earlier reporting. Teacher salaries are given on a district-run basis, excluding charter schools, because charter teachers are paid substantially less and the three districts carry very different charter loads — a comparison on the all-schools basis would understate D38 relative to Cheyenne Mountain, which has no charter school. Turnover is departures divided by prior-year headcount, CDE’s own method, and is published only at district level. The principal and administrator rates rest on headcounts between 10 and 24, where a single departure moves a rate by four to ten points; those figures are reported as counts rather than as rankings, and no comparison between districts on them reaches statistical significance.

Cross-district attendance figures are from CDE’s Student October 2025 count, in the sheets recording non-resident students by the district where their parents live. That file identifies the district a student attends, not the individual school, so it supports a statement about how many D38 residents attend Academy 20 and does not support one about how many attend The Classical Academy.

Cells marked with a dash are withheld by the state to protect student privacy and are not zeros. Where the state suppresses a school’s percentage it often still publishes a mean scale score, which is used here in place of the missing figure.