A Floor, But Not a Ceiling

Welcome to “Part II” of our analysis of this year’s CAT4 results!

In last week’s post, we provided a lot of background context and shared out the simple results of how we did this year.  Here, in our second post, we are now able to begin sharing comparative data, focusing on snapshots of the same cohort (the same group of children) from 2019 to 2021 (with bonus data from 2018’s Grade 3).  In other words, for now based on which grades have taken the CAT4 when, we can only compare at the cohort level from 2019’s Grades 3-5 to 2021’s Grades 5-7 to 2022’s Grades 6-8.  [Remember, that we did not take them at all in 2020 due to COVID.]  In the future, this part of the analysis will only grow more robust and meaningful.

Here is a little analysis that will apply to all three snapshots:

• Remember that any score that is two grades above ending in “.9” represents the max score, like getting a “6.9” for Grade 5.
• Bear in mind, that the metric we are normally looking at when it comes to comparing a cohort over time is whether or not we see at least one full year’s growth (on average) each year – here we are factoring an expected two full year’s growth between 2019 and 2021.  As we discussed last year, that might have been the place one would have expected to see a “COVID Gap” (but we largely did NOT) and between 2021 and 2022 one might expect to see a “COVID Catch-Up”.

Here are the cohort snapshots:

What does this snapshot of current Grade 6s reveal?

• They consistently function a full grade if not not more above the expected grade level.
• That even with COVID we consistently see at least a year’s worth of growth each year across almost all the topics.
• Technically, there is only six month’s worth of growth “Mathematics” (6.9 to 7.5) from 2021 to 2022, but that is already significantly above grade level.
• The one domain, Computation & Estimation, where they are barely below grade level (6.0), we can now properly contextualize by noting that they grew from 4.4 in 2021 to 6.0 in 2022 – more than a year’s worth of growth in a year (the year we would expect a bit of “COVID Catch-Up”.  This means, that they should be more than on track to match all the rest of their scores being significantly above grade level when they take the text in 2023.

All in all…excellent news and trajectory for our current Grade 6s.

What does this snapshot of current Grade 7s reveal?

Not much!  This cohort has maxed out their scores in almost every domain in almost each year!  And in the few places they did not, they were still above grade level – like “Spelling” (4.9) and “Computation & Estimation” (5.5) in 2019 – and grew at least a full grade level each year so that by now, in Grade 7, it is max scores all across the board!  That is pretty awesome to see.

What does this snapshot of current Grade 8s reveal?

This class had a bit of stranger trajectory, but essentially ends where we would like.  “Spelling” took a strange path, beginning way above grade level, plateauing with a dip where we should have seen two years worth of growth, and now fully rebounding to grade level.  “Computation” had a more normal curve, but went from being consistently a year below grade level before completely catching up and now being well above.

To sum up this post, we have a lot to be proud of in the standardized test scores of these particular cohorts over time.  The two areas (Spelling and Computation & Estimation) that were worthy of prioritization the last couple of years (this year’s Grades 6 & 8) were indeed prioritized.   We began providing professional growth opportunities for language arts teachers in our school on Structured Word Inquiry as part of larger conversation about the “Science of Reading”.  [Please check out our Director of Special Needs, Sharon Reichstein’s recent post on this issue, which I’ll also have more to say about in Part III.]  With regard to Computation & Estimation, we discussed it during last year’s November PD Day which focused on “Data-Driven Decision Making” and it has continued to be a point of emphasis.  The results indicate that these efforts have borne fruit.

The Math and Language Arts Teachers in Grades 3-8 have now begun meeting to go through CAT4 results in greater detail, with an eye towards what kinds of interventions are needed now – in this year – to fill any gaps (both for individual students and for cohorts); and how might we adapt our long-term planning to ensure we are best meeting needs.

Stay tuned next week for the concluding “Part III” when we will look at the same grade (different students) over time, see what additional wisdom is to be gleaned from that slice of analysis, and conclude this series of posts with some final summarizing thoughts.

As committed to “transparency” as I am, I find myself growing more and more ambivalent each year about how to best facilitate an appropriate conversation about why our school engages in standardized testing (which for us, like many independent schools in Canada, is the CAT4), what the results mean (and what they don’t mean), how it impacts the way in which we think about “curriculum” and, ultimately, what the connection is between a student’s individual results and our school’s personalized learning plan for that student.  It is not news that education is a field in which pendulums tend to wildly swing back and forth as new research is brought to light.  We are always living in that moment and it has always been my preference to aim towards pragmatism.  Everything new isn’t always better and, yet, sometimes it is.  Sometimes you know right away and sometimes it takes years.

I have already taken a blog post that I used to push out in one giant sea of words, and over time broke it into two ,and now three parts, because even I don’t want to read a 3,000 word post.  But, truthfully, it still doesn’t seem enough.  I continue to worry that I have not done a thorough enough job providing background, research and context to justify a public-facing sharing of standardized test scores.  Probably because I haven’t.  [And that’s without factoring in all the COVID gaps that come along with it.]

And yet.

With the forthcoming launch of Annual Grades 9 & 12 Alumni Surveys and the opening of the admissions season for the 2023-2024 school year, it feels fair and appropriate to be as transparent as we can about how well we are (or aren’t) succeeding academically against an external set of benchmarks, even as we are still just freshly coming out facing extraordinary circumstances.  That’s what “transparency” as a value and a verb looks like.  We commit to sharing the data and our analysis regardless of outcome.  We also do it because we know that for the overwhelming majority of our parents, excellence in secular academics is a non-negotiable, and that in a competitive marketplace with both well-regarded public schools and secular private schools, our parents deserve to see the school’s value proposition validated beyond anecdotes.

Now for the annual litany of caveats and preemptive statements…

We have not yet shared out individual reports to our parents.  First our teachers have to have a chance to review the data to identify which test results fully resemble their children well enough to simply pass on, and which results require contextualization in private conversation.  Those contextualizing conversations will take place in the next few weeks and, thereafter, we should be able to return all results.

There are a few things worth pointing out:

• Because of COVID, this is now only our fourth year taking this assessment at this time of year.  We were in the process of expanding the range from Grades 3-8 in 2019, but we paused in 2020 and restricted last year’s testing to Grades 5-8.  This means that we can only compare at the grade level from 2019’s Grades 5-8 to 2021’s Grades 5-8 to 2022’s Grades 5-8.
• And we can only compare at the cohort level from 2019’s Grades 3-5 to 2021’s Grades 5-7 to 2022’s Grades 6-8.
• This is the first year we have tested Grades 3 & 4 on this exam at this time of year.
• From this point further, assuming we continue to test in (at least) Grades 3-8 annually, we will soon have tracking data across all grades which will allow us to see if…
  • The same grade scores as well or better each year.
  • The same cohort grows at least a year’s worth of growth.

• The last issue is in the proper understanding of what a “grade equivalent score” really is.

Grade-equivalent scores attempt to show at what grade level and month your child is functioning.  However, grade-equivalent scores are not able to show this.  Let me use an example to illustrate this.  In reading comprehension, your son in Grade 5 scored a 7.3 grade equivalent on his Grade 5 test. The 7 represents the grade level while the 3 represents the month. 7.3 would represent the seventh grade, third month, which is December.  The reason it is the third month is because September is zero, October is one, etc.  It is not true though that your son is functioning at the seventh grade level since he was never tested on seventh grade material.  He was only tested on fifth grade material.  He performed like a seventh grader on fifth grade material.  That’s why the grade-equivalent scores should not be used to decide at what grade level a student is functioning.

Let me finish this section by being very clear: We do not believe that standardized test scores represent the only, nor surely the best, evidence for academic success.  Our goal continues to be providing each student with a “floor, but no ceiling” representing each student’s maximum success.  Our best outcome is still producing students who become lifelong learners.

But I also don’t want to undersell the objective evidence that shows that the work we are doing here does in fact lead to tangible success.  That’s the headline, but let’s look more closely at the story.  (You may wish to zoom in a bit on whatever device you are reading this on…)

A few tips on how to read this:

• We take this exam in the “.2” of each grade-level year.  That means that “at grade level” [again, please refer above to a more precise definition of “grade equivalent scores”] for any grade we are looking at would be 5.2, 6.2, 7.2, etc.  For example, if you are looking at Grade 6, anything below 6.2 would constitute “below grade level” and anything above 6.2 would constitute “above grade level.”
• The maximum score for any grade is “.9” of the next year’s grade.  If, for example, you are looking at Grade 8 and see a score of 9.9, on our forms it actually reads “9.9+” – the maximum score that can be recorded.
• Because of when we take this test – approximately two months into the school year – it is reasonable to assume a significant responsibility for results is attributable to the prior year’s teachers and experiences.  But, of course, it is very hard to tease it out exactly, of course.

What are the key takeaways from these snapshots of the entire school?

• Looking at six different grades through six different dimensions there are only five instances out of thirty-six of scoring below grade-level: Grade 3 (Vocabulary 2.2, Writing Conventions 2.5, and Spelling 2.6), Grade 5 (Computation Estimation 4.6), and Grade 6 (Computation Estimation barely falling short at 6.0).
• I’m not quite sure what to make of Grade 3’s Language Arts scores altogether.  Reading and Writing has been the most notable lagging skill for the Grade 3 cohort since their entry into Grade 2.  This is in part due to disruptions to their learning through their foundation-building years in Kindergarten and Grade 1. In Grade 2, this cohort’s remediation was heavily focused on closing the gaps in reading and comprehension abilities, as developmentally this is what comes first.  The remediation focus has shifted to writing at the start of Grade 3, as this is a lagging skill that was already identified prior to the CAT-4 testing.  Supports and interventions have already been put in place to address this lagging skill and we have seen academic growth in these areas.  To put it more simply: These are our youngest students whose early learning was the most disrupted by COVID and they have never taken a standardized test before in their lives.  It will become a baseline that I imagine us jumping over quickly in the years to come – I’m inclined to toss them out as an anomaly.
• Importantly, tracing the trajectory from our 2019 results to our 2021 results to 2022’s, we can now more conclusively state that Spelling and Computation & Estimation are no longer globally lower as a school relative to the other dimensions.  I will have more to say about why we believe this to be true in Parts II & III.

What stands out the most is how exceedingly well each and every grade has done in just about each and every section.  In almost all cases, each and every grade is performing significantly above grade-level.  This is a very encouraging set of data points.

Stay tuned next week when we begin to dive into the comparative data.  “Part II” will look at the same cohort over time.  “Part III” will look at the same grade (the same group of students) over time and conclude this series of posts with some additional summarizing thoughts.