Posted on

The Most Infuriating Test of All

Welcome to the week after President’s Day. For many, this week brings many positive signs: Winter may soon cease its hold on the northern hemisphere. Spring is coming. Lent is here, followed by Easter sooner than you think. Baseball spring training is in-progress, so baseball season is approaching, and it’s here at the high school and college levels. The Winter Olympics are reaching their climax. Other sports are ramping up for 2026 – auto racing, tennis, golf. College basketball’s March Madness will be here soon. But for public school teachers, I have always referred to this time of year as the “Season of Hate.” The holidays were over a while ago. Spring Break is still weeks away. There’s not much novelty in classrooms at this time of year, and the end of the school year seems far in the distance. There’s basically just the daily grind of instruction — and because of recent days lost to icy roads, there is much content on which to catch up. There’s also much content to cover before district benchmarks, or before Spring Break, or before state assessments. Teachers are essentially forced to cram in lots of instruction – tons of it – probably more than they normally would, and students hate it. And truthfully, so do the teachers. It’s the toughest time of the school year, a fairly constant face-off and struggle between students and teachers.

On top of all of this hostility-infused activity, the state layers on another federally-mandated test that doesn’t get nearly enough attention. It’s the poster child for “Good Intentions Run Amok.” Texas calls it the Texas English Language Proficiency Assessment System, or TELPAS. And everybody hates it — teachers, students, administrators, you name it. Everyone hates TELPAS. You’ll hardly ever hear about it on the news because it’s not the centerpiece of the state accountability system, STAAR. But it can affect state accountability ratings, especially if you’re a campus or district with a high population of EB students.

“What the heck are ‘EB’ students?” Ah-ha, so maybe you don’t know the ins and outs of TELPAS. Here’s a quick synopsis: TELPAS grew out of the No Child Left Behind Act of 2001 — that law acknowledged that A) the U.S. has a growing population of students whose primary language is not English and B) it’s imperative that public schools help these students develop English language skills over time so that they can succeed in U.S. schools. We used to call these students English as a Second Language (ESL), Limited English Proficient (LEP), and English Language Learners (ELL), but now we call them Emergent Bilingual (EB). The federal government has a requirement for all public schools to track their English Language Proficiency (that’s the ELP in TELPAS). The State of Texas had actually begun such tracking way back in 1999. The test covers four Domains of language: Reading, Writing, Listening, and Speaking.

“Why does everybody hate TELPAS?” I would suggest a big part of it is that TELPAS is yet another intrusion on instruction right smack dab in the middle of the Season of Hate. But beyond that, it’s a challenge precisely because it is NOT required for every single student; it’s only for students designated as EB. The EB designation is already pretty vague, but once it’s attached to the student’s record, it’s also difficult to remove. Which makes TELPAS a pesky requirement, even for EB students who actually know English well. At its inception, TELPAS still affected all students, because the writing portion was required to be “authentic writing from an academic class,” BUT the assignment had to be assigned for every student. So teachers were expected to develop some contrived writing assignment for 100% of their students just so EB students could be evaluated for TELPAS. Then there was the “calibration” process for teachers, where they had to figure out how to evaluate what constituted Beginner vs Intermediate vs Advanced vs Advanced-High proficiency, specifically according to the state’s guidelines. That process, by itself, was at least two weeks of extra work for teachers just for the training. THEN they would have to assign the writing assignments in class , and get students to actually complete them (easier said than done). THEN administrators had to collect the writing samples for the EB students (and that process had a multitude of other requirements). THEN the teachers had to read all the samples and evaluate them. THEN the teachers also had to evaluate the Listening and Speaking proficiency. PLUS the EB students would have to take a basic Reading test online.

“How long would it take to do all that?” Way. Too. Long. Teacher training for TELPAS occurred at the beginning of February (for two weeks), then the “window” for completing TELPAS activities opened right after President’s Day and closed at the end of March (another SIX weeks). That’s EIGHT weeks of TELPAS-related activity, not including the time administrators spent attending trainings and preparing to coordinate all of this stuff. Naturally, campuses would do their best to complete TELPAS in the early part of the window, but there would be struggles to get students to complete the writing assignments (a minimum FIVE samples were required, with other stipulations as to what type of writing was needed). There might be students absent on the scheduled date of the reading test. Or there might be students who were chronic absentees through most of the TELPAS window. Then there was the issue of what to do with new students or students who withdrew from your campus or district during the window. It was a cacophony of never-ending activity, all in service to a test that, while well-meaning, ultimately mattered not one bit to district or, frankly, TEA officials. Students hated it. Teachers hated it. Administrators hated it. Everybody hates TELPAS. As a district testing coordinator, I literally made a joke out of how much people loathed the whole thing.

In recent years, the cacophony has quieted ever-so-slightly. For grades 2-12, TEA now offers two online assessments – one for Listening and Speaking, another for Reading and Writing. For all but K-1 teachers, the chaotic days of calibration and writing sample collection are over. But there’s still a logistical challenge, because the Listening/Speaking test requires students to use wired headsets plugged into the computing device. Yes, wired headsets — as in, headphones with a microphone attached. You can get cheap earbud-style headsets on Amazon at $1/item or less by buying in bulk, but if you want decent over-the-ear headsets that can easily be reused, you’re typically talking $10-$20 per headset. If you’re an elementary campus with only a few dozen EB students, it’s easy. But if you’re a 6A high school like where I most recently worked, with 50% EB students (yes, 1,500 students out of 3,000), that’s A LOT of headsets, and a potentially HUGE expense. To reduce the possibility of chronic absenteeism, we used to schedule a “shutdown” day for each test, but we couldn’t schedule more than 15-20 students per testing room because the speaking portion required distance between students. So you can probably imagine that planning roughly 100 testing rooms and distributing 1,500 headsets while relocating another 1,500 students, using a staff of up to 200 adults, was…interesting. And if you can imagine that, you can see why everybody STILL hates TELPAS, even though it’s now 100% online testing. Because you know what? The district and the state still don’t really care about it. Yet here we are.

“So why do they still do it?” Because of what I said earlier – “Good Intentions Run Amok.” Honestly, TELPAS is born of a well-meaning concept. If we’re concerned about the academic performance of ALL students, and academic performance hinges on the literacy rate, then there’s definitely a need for students who come from non-English-speaking households to grasp the English language and establish literacy skills in it. Trouble is, our politicians’ first thought regarding student performance has always been “Let’s have the schools stop everything and test them.” So NCLB mandated it, and ESSA (the Every Student Succeeds Act) renewed the requirement. Thus, TEA has to figure out when they can cram the test into the academic year without intruding on all the other federally-mandated tests. So their testing calendar puts it square in the Season of Hate. Fun!

“Do these EB students show improvements?” Based on my own experience in the classroom and as an administrator, I am confident that most EB students learn English reasonably well by being immersed in English-based instruction IF their teachers offer instruction that properly teaches the vocabulary of the course. (That is a massive “IF.”) That can be said for native and non-native English speakers alike. There are plenty of native English speakers out there who don’t have the first clue about how to use the English language, too. (Seriously, just read online comments for a sampling of the illiteracy problem in the U.S.) But the key to any of this is instruction, not testing. There’s not a single EB student who ever dug down and learned English better because “I’m going to ace that TELPAS test.” Not one. TELPAS is well-intended, but you simply cannot suggest its existence has improved the academic performance of EB students. It’s been around for roughly a quarter-century, but any statistical analysis of TELPAS growth results is flawed at best because of the fluctuation in EB student populations, changes across districts and campuses, changes in the format and content of the tests, and, oh yeah…these are still kids being evaluated every year from K-12. (YES, K through 12; TELPAS doesn’t wait until Grade 3.) I’m sure EB students are improving their English skills overall; I’m also sure it has nothing to do with TELPAS.

“We should get the government to abolish this test!” It wouldn’t hurt my feelings if we got rid of all of the federally-mandated testing requirements — not just TELPAS, but STAAR, the next iteration of testing coming to Texas, and everything else. (And that’s coming from a guy who spent over a decade in the testing world.) Without the mandates, schools would naturally figure out what standardized tests, if any, would generate the data that would actually help them improve instruction and learning. But the reality is that we live in the era of school accountability (another can of worms all its own), and where there is accountability, there is assessment. I don’t know of many politicians with the fortitude to admit that the extra requirements they’ve been adding for the past 30-40 years aren’t working. If anything, they double- and triple-down. Add to the equation all the marketing and lobbying efforts of the big testing firms out there who profit from the system. So it’s probably never going away in our lifetime. The best we can hope for is that the trend recedes a bit over the next decade, and then hope that any reversal gains transaction.

In the meantime, I offer my former colleagues my best wishes as they navigate these waters in the school year. I know you hate TELPAS. I know your students hate TELPAS. Everybody hates TELPAS. Have some solidarity in that knowledge and work through it together, knowing everyone’s in the same boat with similar emotions. “Survive and advance,” as they say. It’s going to suck, but it will end (…okay, just in time for the next round of testing, but still). Do your best to fight through it. If you’re not in education and have never had to deal with this type of stuff, try to have a little empathy for those in the throes of this time of the school year — students, teachers, and administrators alike. If it seems like it’s rough for them, that’s because IT IS. Recognize and appreciate that. But know this: The Season of Hate eventually ends, every year, and is supplanted by the Season of Love. (Don’t worry; I will write about that when the time comes, too, because it’s the best.) Stay strong, and don’t let the frustration consume you. EHT.

Posted on

The Least Wonderful Time of the Year

No, this post is NOT about to be 2,600+ words bashing the Christmas season for being too commercialized or sappy or anything else. I’m actually loving the holiday season this year. Anyone who’s read this blog before probably knows why.

What this post IS about is STAAR Testing – specifically, STAAR End-Of-Course (EOC) Retests for those high school students who are cursed with the requirement of taking them. I will also share a personal story about what I now believe was proverbial “Testing Hades,” and how I helped get my staff (and myself) through it using humor.

I have previously confessed on this blog that the final dozen years of my K-12 education career had me wrestling with self-loathing because, as a so-called “assessment professional,” my biggest role was training, implementation, and support of a system that is deeply flawed at best. There are a host of reasons for these flaws, many of which would involve exposition that would truly be agonizing to read. What it comes down to is primarily inconsistency.

STAAR is fundamentally inconsistent as a matter of course because everything about its construction is constantly changing. The state curriculum of Texas public schools, known as the Texas Essential Knowledge and Skills (TEKS) is regularly revised. Subject area TEKS are reviewed and adjusted at least once every 10 years on a rotating basis. Schools don’t have to contend with full-scale revisions all at once, but teachers of a given subject know that they will have to adjust planning and teaching within a decade at minimum. While a decade doesn’t seem frequent on its face, the reality in a classroom is that it may take several years after revisions are published to determine the best way to approach certain TEKS standards instructionally. It’s not like anyone can just flip a switch, make a couple of tweaks, and all students magically respond positively to whatever the new TEKS are.

And as you might expect, the process of assessing any new TEKS learning standards can also be messy and flawed over several years. This is why TEA randomly selects districts and campuses for specific field tests every year, usually in February. A typical TEA Field Test is similar to an actual STAAR test, and campuses are required to implement them with the same approach and security protocols that they use for the actual STAAR test in the spring, except there’s no payoff for a field test. Whether students do well or poorly, they will never be rewarded or punished. They will never even know how they scored. Reporting of results is minimal, because the point is literally to test the test items themselves and determine if they’re reliable and valid. If you’re thinking that your children might become lab experiments each February so that psychometricians can analyze results, you would be correct.

This process is also why actual STAAR tests in the spring include “field test items” that may or may not actually be scored. And it’s why TEA will rewrite and revise items continually between STAAR testing cycles, which occur annually each April for Grades 3-8, but every April, June, and December for high school students. There’s a continual item analysis process to seek out a reliable and valid test, and as you might expect, some STAAR tests are more reliable and valid than others. Small comfort to a high school freshman who learns he has to retake an EOC even if the test itself was poorly constructed.

The revision process, combined with the desire of schools to “teach to the test” so that scores improve has created its own cottage industry. A legion of consulting companies have developed with intent of helping schools and their teachers to analyze released items from STAAR tests, connect those items to the TEKS standards being assessed, and determine ways to adjust instruction so that the teachers may better prepare students for what they might see on the STAAR. Is it good instruction? Sometimes. Your mileage may vary, as they say. But “teach to the test” has become scientific (or pseudo-scientific, in many cases), all in the name of accountability points, property values…and, oh yeah, “for the children.”

Speaking of those children – we all want them to excel, right? We want them to score well since it’s evidence that they’re learning, it’s good for self-esteem, yada, yada, yada. But of course, it’s not always so simple. Remember that this is a test that students will see only once per year. (Theoretically three times with upcoming legislative changes.) And when they see the test next year, it’ll be a bit different from last year because Grade 8 is different from Grade 7, which is different from Grade 6, and so on, even if the subject is still Reading or Math. It is no wonder that most schools see drops in Math scores, for instance, from Grade 4 to Grade 5, year after year, even as students change, because Grade 5 Math is typically a bit more challenging than Grade 4 Math. What’s more, the actual passing standard might be adjusted thanks to the magic of the “cut score.”

When you visit TEA’s website, you might run across the STAAR Performance Standards for Grades 3-8 and EOC. You’ll see austere tables listing the Scale Scores required for a given student to reach Approaches, Meets, or Masters Grade Level performance categories as handed down by the gods themselves…er, I mean, by TEA officials. These categories and scale scores are unchanged (thanks to the auspices of either Odin or Ra; I never remember which), but the real sausage is made after students have completed the tests, psychometricians have analyzed the results, and TEA constructs what’s known as a Raw Score Conversion Table. Take Grade 5 Math, for instance, which had 42 items that were scored in 2025. TEA takes every student raw score, from 1 to 42, and links it to a given Scale Score. Those links then determine what raw score a student needed to reach the Approaches / Meets / Masters level from the Performance Standards. It’s all computed AFTER THE FACT, because TEA looks at the distribution of student raw scores statewide before deciding where, in fact, “passing” will be located. In 2025, a Grade 5 student in Math needed to get 17 of 42 items correct for Approaches, 26 for Meets, and 34 for Masters. In 2026 and beyond, those raw scores could change depending on how every student in Texas fared on that year’s test. It’s a moving target every year. It might not move much, but it can move. Is it possible that public relations and political concerns can impact where the cut scores fall? You tell me. It was certainly curious that accountability ratings in 2023 and 2024 went to court, then in 2025 most districts and campuses in the state saw their STAAR scores and ratings increase. I don’t have the time or inclination to lay out a full analysis of the data over those years, but it sure was a curious coincidence.

The bottom line is that passing standards can be, and are, adjusted year-to-year as cut scores are linked to scale scores. “Passing the STAAR” is, itself, an exercise in inconsistency. What’s more, “passing” isn’t always the end for students in high school. Current TEA guidelines require students to reach Approaches Grade Level or higher in all 5 EOC-assessed subjects – English I, English II, Algebra I, Biology, and US History. (I won’t open the Individual Graduation Committee or the Substitute Assessment cans of worms; those are other deep-dive posts.) The students currently taking December EOCs are those who have previously not passed one or more tests, or who were Absent or otherwise missed their opportunities in April and June. When a student finally achieves those scores and passes their classes, graduation is on the horizon. Perhaps college? Not so fast. “Approaches Grade Level” is passing for graduation purposes only. TEA has a whole other set of standards for what is called “College, Career, and Military Readiness” (CCMR). And Approaches on STAAR EOC ain’t one of them. In fact, STAAR scores don’t matter AT ALL for CCMR accountability standards, and they won’t grant students access to Texas public colleges and universities…at least, not without some type of remedial learning. So now, high schools in Texas offer the TSIA2, SAT, and ACT at least once to all their students in an attempt to get as many students as possible to meet CCMR requirements. More testing for our high schoolers! Isn’t it grand? In fact, it’s several grand paid each year by the taxpayers, or by the students themselves.

Lest you think we only torture the high schools and their students, there’s a whole set of other accountability measurements that primarily impact elementary, intermediate, and middle schools and their students – the Progress Measure, brought to you by TEA through each year’s STAAR tests. The intent is actually well-meaning and fairly intuitive: Students should show growth, also called academic progress, in their year-to-year performance on STAAR Reading and Math tests. Easy, right? Of course not! You might think that an intuitive approach to growth would be that a given student should score at or higher than the previous year’s raw or scale score to show progress. Or perhaps there should be a set of score ranges that might overlap so that students wouldn’t be penalized for missing one more item than last year. But you would be wrong in both cases. Instead, TEA determines progress based on the student’s performance among the Approaches / Meets / Masters standards, which we’ve already established may change thanks to cut scores. Essentially, in order for a student to “meet progress” officially, that student must match or exceed the performance category from the previous year. If they reached Approaches last year, they must reach Approaches, Meets, or Masters this year. Here’s the problem: because those performance levels are matched directly to a specific raw score, it’s possible for a student to “not meet growth” based on a single test item. Consider the Grade 5 Math scores referenced earlier, and suppose a 5th grader in 2025 got 34 of 42 items correct in 2025 to reach Masters Grade Level. It just so happens that in 2025, Masters on Grade 6 Math also required a raw score of 34 out of 43 items. BUT, in 2026, IF the Masters Level ends up being raised to 35 items after TEA’s psychometric analysis, AND this same student gets the same raw score, 34 on Grade 6 Math, the student will actually DROP to Meets Grade Level. That might seem fine, BUT this student will be deemed “Did Not Meet” for the progress measure in Math. By a single item. Even though the student passed the test easily, getting 79% of the items correct when it only takes 37% to “pass.” Is this equitable, fair, justifiable, reliable, and/or valid? You tell me.

Knowledge of such flaws and inequities are just some of the reasons that it became more and more difficult for me to justify continuing to work as an “assessment professional.” It became increasingly difficult to pretend that the system was defensible, let alone worth training teachers how to implement it appropriately. Of course, when it came to the insanity of the system, Grand Prairie ISD said “hold my beer” and added layers of local assessment to this Least Wonderful Time of Year. So began the creation of Testivus.

Here’s how it happened: The high school calendar in GPISD had students attending classes for roughly 3 weeks after Thanksgiving break. Week 1 was mostly instruction; TEA allowed districts to offer December EOCs that week, but GPISD elected to wait. Week 2 was when GPISD offered December EOCs over four days (Tuesday through Friday because TEA at the time did not allow STAAR tests on Mondays). But GPISD also added four (4!!!) additional days of local assessments – “Q2 Summative Assessments,” they were called – and required “shutdown” testing for EOC-assessed courses. What’s more, the US History Q2 Summative HAD to be given on the Friday of Week 3 because of district policy on semester exams, meaning we had to hold both the final day of EOCs AND a major local assessment on the same day. On a Friday, no less. Self-induced torture. Or should I say, district-induced torture. Somewhere in those 8 days, non-core subjects also had to offer Semester 1 Exams. As you might imagine, the schedule was somewhere between confusing and downright comical. As an administrator, it was a death-defying juggling act just to create a coherent schedule, and then we had to communicate it to the staff and students. So I had a choice: either tear out my hair, elevate my blood pressure, and otherwise stress myself out at the holidays trying to make it work, or have fun with it. I decided to have fun and approach it with humor. So Testivus was born.

Testivus, as any good Seinfeld fan would infer, was a riff on Festivus, the fictional holiday “for the rest of us” in response to the rampant commercialism of Christmas. We needed something bizarre to associate with, and yes, resist, the madness, because there was no way to comprehend it without also admitting it was strange and convoluted. I even created a logo for it that I included on documents I gave to the teachers. I made jokes about it in e-mail communication. I was also brutally honest. “This is what happens when district tries to shoehorn more than 8 tests into 8 days. You might argue it’s the counterpoint to the miracle of Hanukkah.” Eventually, my humor got me in mild trouble. It just so happened that the husband of the district’s head of data and accountability worked as the chief security officer on our campus, and he received the mass e-mails I sent to the staff. He would forward said e-mails to his wife. But rather than contact me herself, she asked the district testing coordinator (DTC) to call and badger me about my humor. It so happens that the DTC’s personality is often quite dramatic, dialing any little issue up to 11 (shout-out to Nigel Tufnel) immediately. So I was told, “Senior district administrators are reading your e-mails, and they are not amused. I guess you’re trying to be funny, but to them it sounds like you’re pitting your campus in opposition to the district.” Fine, whatever. I expressed regret to her that she was being asked to deal with it, but the reality was that district was pitting itself against campuses by inundating us with local assessments literally layered on top of state assessments. I make no apologies for calling that out, nor would I apologize for fostering some empathy and camaraderie with my campus colleagues through humor that they genuinely enjoyed. And you know what? In Fall 2025, Grand Prairie ISD nixed those plans, removed “Q2 Summative Assessments” from the calendar entirely, and instead created “Fall District Assessments” on a different week in November. Granted, they still managed to make it awkward (not worth outlining how here), but they changed their approach to local assessments. Maybe the actual message behind Testivus, and my humor, resonated with someone, after all.

So if you’re winding your way through a similar situation, navigating the nonsense passed down from someone who doesn’t realize the negative impact on your students and teachers, I wish you a Happy Testivus. Hang in there, do the best you can under the circumstances, and most of all, laugh about the laughable. Go ahead and roll your eyes at the outlandish. Give yourself permission to have fun with it instead of getting upset. And most of all, realize that none of these things…zero, nada, nenhum, nüt…will matter all that much in the grand scheme of things. Treat your students and your colleagues with dignity, empathy, and kindness in this most blessed of holiday seasons, and let the absurd wallow in its absurdity.