Greg's Tutoring SHSAT Tutoring Greg's Tutoring SHSAT Tutoring Greg's Tutoring SHSAT NYC Specialized HS Tutoring Greg's Tutoring SHSAT NYC Specialized HS Tutoring Greg's Tutoring SHSAT NYC Specialized HS Tutoring Greg's Tutoring SHSAT NYC Specialized HS Tutoring

Greg's Tutoring SHSAT NYC Specialized HS Tutoring Greg's Tutoring SHSAT NYC Specialized HS Tutoring

Greg's Tutoring NYC has created a number of YouTube livestreams and other info pertaining to the SHSAT, NYC's Specialized HS Admissions Test.

Please subscribe and turn on notifications so you'll stay in the loop as we add new content about the digital adaptive SHSAT.

And don't hesitate to email us if you have any questions or concerns about SHSAT admissions or prep! Have fun and enjoy!

Details on, and resources for, the upcoming digital ADAPTIVE SHSAT for 2026 and beyond:

March 12, 2026 NOTE: The DOE will release some online digital ADAPTIVE practice exams soon.

My advice would be to HOLD OFF ON ACTUALLY DOING THE ONLINE ADAPTIVE EXAMS FOR NOW. Even once they release them.

On this note, BE CAUTIOUS OF 3RD PARTY DIGITAL ADAPTIVE EXAMS that may be made available, as getting questions right is not as easy as it seems, and YOU WANT TO BE SURE YOU'RE WORKING WITH QUALITY QUESTIONS on these new question types, and even of the previous PBT (paper based test) questions.

I expect be posting videos with an unboxing and rip down of the exam shortly.

Any questions feel free to message me.

The Adaptive SHSAT is coming!
New video series!
Subscribe and set Notifications!
The Adaptive SHSAT, Let's talk, PART 1
March 20, 2026
The Adaptive SHSAT
Explore WHY it's NOT like the Adaptive SAT
PART 2
March 22, 2026
The Adaptive SHSAT
is NOT i-Ready "deep dive"
PART 3a
March 23, 2026
The Adaptive SHSAT
is NOT i-Ready "lesser dive"
PART 3b
March 23, 2026
The Adaptive SHSAT
is NOT i-Ready "executive summary"
PART 3c
March 23, 2026
i-Ready
======= Adaptive SHSAT
============== EITHER adaptive exam:
=================
The Adaptive SHSAT, AMA (Ask Me Anything) Livestream, Part 4
April 2, 2026
The Adaptive SHSAT
Recent Updates
PART 5
July 28, 2026
UPDATED Info for Fall 2026 Adaptive SHSAT
The Adaptive SHSAT
How Much Harder Is It?
PART 6
August 8, 2026 
The new adaptive SHSAT for fall 2026+
A CATASTROPHIC CONSTANT SERIES OF HARD QUESTIONS AS THE EXAMS ADAPTS?
Parents, students, and even many tutors have jumped down rabbit holes of misinformation.

In the following I will be showing some calculations.
They are exact computations but not exact numbers for use.
We're modeling and having a discussion.
That is, we're going to do back of the envelope, and possible talking points that may yield.
Just using some fuzzy math to get some ballpark information.

Oh, I am not the DOE.
We are not the DOE.

As well, the notion of getting all questions correct may not be every student's reality.
-> Consider it rare.
-> You've been warned.

Consider the following from the "NYC Guide to the SHSAT for 2027 Admissions":
"...A CAT is an exam that selects the questions
 that you will see based on your ongoing performance.
 For example, IF YOU ANSWER A MODERATELY CHALLENGING ITEM CORRECTLY,
 YOU WILL THEN RECEIVE A MORE DIFFICULT ITEM;
 IF YOU DO NOT ANSWER SUCH AN ITEM CORRECTLY,
 YOU WILL THEN BE ASKED A LESS DIFFICULT QUESTION."
(item/question, receive/be asked)

I've participated in many conversations that hence go thus:
Let's say it makes the question twice as hard,
won't that get too hard for most students?

Hmmm.
If a question gets say twice as hard each time it provides a more difficult item,
how does that unfold across the exam (per section)?
Well, after 49 times, it is 562,949,953,421,312 times harder
approximately 5.63 × 10¹⁴ or 563 trillion times harder.
Um. oookay.
I mean, was this ever really even a logical question or thought to raise or ask?
It's QUITE misguided.

Let's back off of even that:
Instead of twice, how about 10%?
If something is .1 harder each time, how much harder is it after 10 times
Additive: .1 x 10 = 1 so 2 times (100%) harder
or
Compounded: 1.1¹⁰ =  2.59 times (159%) harder, ala
 0 1.0
 1 1.10  1.10
 2 1.20  1.21
 3 1.30  1.33
 4 1.40  1.46
 5 1.50  1.61
 6 1.60  1.77
 7 1.70  1.95
 8 1.80  2.14
 9 1.90  2.36
10 2.00  2.59

Compounded,
If something is .1 harder each time, how many times will it take to be twice as hard?
  It takes about 7.27 times (or 8 full steps) for the entity to be twice as hard
That seems reasonable?

But now consider:
If something is .1 harder each time, how much harder is it after 49 times
Additive: .1 x 49 = 1 so 4.9 times (390%) harder
or
Compounded: 1.1⁴⁹ =  98.7 times (9770%!) harder ≈ 100 times harder!
Stop.  Think of how much harder that problem would actually be?
Everybody: Ouch!

If 10% is too high, what's pragmatic?
If something is .06 harder each time, how much harder is it after 49 times, multiplicatively
1.06⁴⁹ ≈ 17.38
That's a lot better than ~100 times harder.
But is ~20x harder realistic?

If something is .05 harder each time, how much harder is it after 49 times, multiplicatively
1.05⁴⁹ ≈ 10.92

If something is .04 harder each time, how much harder is it after 49 times, multiplicatively
1.04⁴⁹ ≈ 6.83

If something is .03 harder each time, how much harder is it after 49 times, multiplicatively
1.03⁴⁹ ≈ 4.25

Is ~4-5x harder realistic?
(Don't mix up 4x vs 4%.)

If we look at this in reverse:
If something is 10 times harder multiplicatively,
after 49 times how much harder is it between each iteration (question)?
10¹ᐟ⁴⁹ = ~1.04811 so 1.04811⁴⁹ = 10

Note this is 4-5% harder 4/100.
Stop.  Think of how much harder that problem would actually be?

5¹ᐟ⁴⁹ = ~1.03339 so 1.03339⁴⁹ = 5
4¹ᐟ⁴⁹ = ~1.02869 so 1.02869⁴⁹ = 4
3¹ᐟ⁴⁹ = ~1.02267 so 1.02267⁴⁹ = 3
2¹ᐟ⁴⁹ = ~1.01424 so 1.01424⁴⁹ = 2

Coming full circle, let's reconsider what we're even doing.
Did our back-of-the-enveloping go too far?
Are we off-course?  Way off?
Is our fuzzy math... too fuzzy?
Is what's harder for me harder for you?
What does hard even mean?

AND REMEMBER THE CAVEAT WARNING FROM MUCH EARLIER:
the notion of GETTING ALL QUESTIONS CORRECT may not be every student's reality.
-> As we know, it's rare (maybe 2/1000).
So such an impact is going to even be less so in-the-wild.
Less for ALMOST EVERY STUDENT, among thousands.
The "average" student will get questions correct and incorrect,
both,
even of top students.

Well, what have we been told/informed about?
Now, consider the implication of:
* If a student ANSWERS a question CORRECTLY,
  their estimated SCORE WILL GO UP,
  and they MAY then RECEIVE A MORE CHALLENGING QUESTION.
* If they answer INCORRECTLY,
  their estimated SCORE WILL GO DOWN,
  and the next question MAY have a LOWER DIFFICULTY level.
Note MAY.
This is a different statement from the statement earlier.
It is different even if we acknowledge that the other statement
began with "For example."

Moreover, while everybody focuses on the above aspects of the narrative,
actually it's not about difficulty as such.
We'll explore this further at some point.

ALSO
A N D  T H I S  I S  HHUUGGEE:

***Question difficulty is not measured on a ratio scale.***
Say what?
There is NO MEANINGFUL STATEMENT such as "this question is twice as difficult as that one."
100 degrees is not twice as hot as 50 degrees.
It's 50 degrees *hotter*.

If you're known to pole vault 10 ft they're not going to automatically move it to 150 ft.
Probably not even 15 ft.

A sidebar lense:
Consider a PBT (Paper-based Test):
Some tests are written with questions that go from easy to hard
(for some definition of hard),
so you may have:
... E E E M M M M H H H ...
slowly "ADAPTING"
-> The PBT is adapting!
Doesn't that already have a psychological impact?

And some PBTs are written with random difficulty, so you may have:
... E M E M H H E H M M ...
This adaptation is already built into the ebb and flow of questions.
Students must already deal with this adaptation.
Doesn't that already have a psychological impact?

-> This is significant selective ignorance.
-> Don't test by bling factors.

Given the roughly 4-phases of a per-question adaptive exam:
* locate quickly where a student is
* core measure where a student is, collect bulk of info here
* refine the estimate 
* stabilize the estimate
and assuming you get every math question correct:
Q#50 difficulty = Q#1 difficulty × 1.075^7 × 1.035^12 × 1.02^20 × 1.015^10 ≈ 4.32
So about 4.32 times as difficult, or about 332% more difficult than Question 1
Not for every question,
but across the whole exam,
at most.

~Averaging
4.32¹ᐟ⁴⁹ = 1.0303127 per step to reach a total of 4.32 over 49 steps.
(See earlier math.)

BUT

Students won't get every question correct, generally.

FURTHERMORE

A per-question adaptive test is not "making questions harder and harder."
(See discussion.)

Instead, A PER-QUESTION ADAPTIVE EXAM such as an ADAPTIVE SHSAT IS
TRYING TO PRESENT QUESTIONS CLOSE TO WHERE IT CURRENTLY BELIEVES YOUR ABILITY IS.
And uses questions that inform it thus.
🚨 THIS is its job.

If your current estimated ability is X,
giving you a question Y times as easy means it it likely you will get it correct.
It already knew that -> That's not informative.

If your current estimated ability is X,
giving you a question Y times as hard means it it likely you will get it incorrect.
It likely already knew that -> That's not informative.

MORE INFORMATIVE are questions close to where your (current estimated) ability is.
Not making long shots in the dark but nearby.
IF a student got all questions correct, they would likely reach the top of the question bank,
upon which the test cannot keep increasing difficulty indefinitely.
The SHSAT is a grade level exam and won't continue into HS, college, or beyond.

Actually around mid-test, the student might already be near the test's difficulty ceiling.
Questions after that would likely be at a similar upper difficulty band
-- perhaps reaching a plateau near the top of the bank --
and NOT CONTINUE TO BECOME DRAMATICALLY HARDER.
(See the 1.015 above.)
Instead it would begin selecting among the hardest available questions.
So it would just stay there?
Not quite.
I would also be satisfying topic content requirements.
It would also consider satisfying content that separates
to rank and
for diminishing sweet spots and
allowing eventual spreads for cutoff scores.
After all, this is a competitive exam so
🚨 THIS too is its job.

Regarding from earlier: 1.075^7 × 1.035^12 × 1.02^20 × 1.015^10
The test difficulty rises quickly at first.
It then levels off because the test has already found your approximate ability.
That expresses not difficulty but MOVEMENT ALONG THE DIFFICULTY scale;
it is NOT A MULTIPLICATIVE INCREASE in "hardness."
The algorithm MOVES along a CALIBRATED difficulty scale, and then some,
and not by multiplying question difficulty.
(See discussion above.)

So regarding Question X must be "twice as hard" as Question Y
(whether the first vs the last question, or questions next to each other),
A CAT DOESN'T WORK THIS WAY EITHER.
IOWs, Question 50 is targeted much higher on the calibrated difficulty scale than Question 1.
To wit, it raises the "bar" quickly until it finds roughly where your limit is,
and then it keeps presenting questions near that level
rather than making each new question dramatically harder than the last.
-> It's trying to be informed.
🚨 THIS is its job.

For a "modern per-question CAT," the largest upward (or downward)
JUMPS ARE USUALLY CONCENTRATED IN THE EARLY PART OF THE TEST,
because that's when it knows the least about you.
(Remember it's trying to find out information about you,
so it will do what is most informative, NOT strictly most difficult.)
(And this includes downward too, see above about X moving Y.)

Imagine the computer starts everyone at "average."
It doesn't know you from a hole in the wall.
But the SHSAT is known as a competitive exam,
and will start as such accordingly and adjust (adapt) from there.

The DOE says for the ELA RC section it might present a "start-up" passage.
So IN THE BEGINNING after a handful of responses, 
it can -- and will -- move the target DIFFICULTY RELATIVELY QUICKLY.

If after say 25 questions it believes you're an exceptionally strong student,
and you suddenly get say five very informative questions wrong,
it will reconsider.
This can happen.
That could produce a larger-than-usual adjustment, even in the middle of the test.
Maybe it gave you the benefit of the doubt.
Did it overestimate?  Did you get lucky?

From what we've discussed earlier,
a modern CAT such as the adaptive SHSAT is likely to:
* make biggest adjustments early
* make relatively small adjustments later
* avoid wild swings in difficulty
* keep presenting questions near the student's estimated ability.

So A STUDENT IS UNLIKELY TO EXPERIENCE:
Question 24 was easy, but then Question 25 is impossible!
Right?  Think it through.

However, a student might FEEL:
that the questions all FEEL pretty hard.
or
that the questions all FEEL about the same.
Surprises are minimal.
THIS IS ALL ACTUALLY BY DESIGN.
🚨 THIS is its job.
There could be a feeling from a PBT too.
(See selective ignorance above.)

This moves us along to...
The PLACEBO EFFECT is "a phenomenon where a patient's physical or mental health
 IMPROVES AFTER receiving a FAKE or inactive TREATMENT
 (like a sugar pill or saline injection).
 It is DRIVEN BY the person’s BELIEF in the treatment and the EXPECTATION of feeling better.
Far from being purely "imaginary," the placebo effect
 TRIGGERS MEASURABLE PHYSICAL CHANGES IN THE BODY."

Which moves us along to...
The NOCEBO EFFECT.  "Just as positive expectations can lead to healing,
 NEGATIVE EXPECTATIONS can cause a person to EXPERIENCE ADVERSE SYMPTOMS
 —even if the treatment is inactive.
 This is known as the nocebo effect (from the Latin word for "I will harm").
If a patient EXPECTS to experience SIDE EFFECTS like headaches or nausea
 from a fake treatment, there is a HIGHER CHANCE THEY WILL ACTUALLY OCCUR."

-> If a student ANSWERS a question CORRECTLY,
 their estimated SCORE WILL GO UP, and they WILL then RECEIVE A MORE CHALLENGING QUESTION.
-> An adaptive exam is supposed to be hard, real hard, and it shall impact you psychologically.
-> Who said?
-> Are we projecting?

Objectively, THE EMOTIONAL IMPACT is probably LESS SEVERE THAN MANY PEOPLE IMAGINE.
This is the NOCEBO MENTIONED EARLIER, and the math earlier supports the misperception.
There is NO WAY the exam will be as hard as the numbers shown earlier.
AND (knowing this) it can be made to be less severe!

With a PBT, students know everyone receives the same test (ignoring forms for now).
They might consider they got an easy question, a challenging question, an "impossible" question.
BUT
they generally don't attribute those differences to themselves.
Instead, they attribute the differences to the test.
Even some PBTs are known to start easy and end hard.
After all, it "just" has easy, medium, and hard questions.
So much so often with a PBT a student ignores the consideration exists!
(Whether doing so is justified is another issue altogether.)
-> Conclusion: with a PBT a student usually considers the source of the variation external.
-> Either way, focus on the question before you.
-> Continue to attribute the differences to the test, not you.

But with a per-question adaptive exam,
a student may think:
* "This question seemed easier..."
-> "Did the computer lower me?"
* "This question  is really hard..."
-> "Maybe I'm doing well."
* "Now this question feels easier again..."
-> "What happened?"
Now the PERCEIVED source of the variation becomes internal.
And they may think:
"The computer is responding to me."
That shift in attribution can change the emotional experience.
-> This need not be a roller coaster
-> Go with the flow as on a PBT

Notice that the inference may be wrong
Here's the interesting part.
While the student says:
* "This question is easier."
The algorithm might actually have changed the target difficulty
 only a tiny amount—or not at all!

Other reasons the question might FEEL easier include:
* It's a topic the student is stronger in.
* The wording is clearer.
* The student has seen a similar problem before.
* The item is highly discriminating without being obviously difficult.
PEMDAS/Order of Operations questions succumb to this.
* The CAT needed to satisfy a content constraint
for example, selecting a geometry item rather than another algebra item

So the student's conclusion:
* "I must be doing worse."
may simply be [flat out] incorrect.
And this kind of thing becomes a distraction and a diversion of energy et al.
🚨 NOCEBO IN ACTION.

This changes the student's internal dialogue
A PBT student might think:
* "That was a hard question."
An adaptive-test student might think:
* "Why did I get that hard question?"
Those are psychologically different.

-> A student may interpret the variety as feedback about their own performance.
That could lead to thoughts like:
* "I'm moving up."
* "I'm moving down."
* "I must have blown the last one."
* "The computer doesn't think I'm very good."
Those thoughts are about self-evaluation,
 which CAN INFLUENCE CONFIDENCE EVEN IF THE INFERENCES ARE INACCURATE.
-> This moves away from answering tbe question before you.
-> Don't!

And this is one of the biggest differences
To summarize:
Paper SHSAT: Difficulty variation is largely attributed to the test.
Per-question CAT: Difficulty variation is more likely to be attributed to the student's performance.

Whether that attribution is accurate is another matter.
Because CATs:
* adjust gradually
* balance content
* often present questions near a student's ability
THE APPARENT DIFFICULTY OF AN INDIVIDUAL QUESTION IS NOT A RELIABLE INDICATOR OF HOW THE STUDENT IS DOING.
-> Yet many students may still try to read meaning into each question,
-> That interpretation—not the algorithm itself—may be the larger psychological challenge!

Paper SHSAT: "FIXED UNCERTAINTY"
Every student experiences:
* easy questions
* medium questions
* hard questions
in some predetermined order.
The student's internal dialogue was often:
* "The test is getting harder."
or
* "This section is easier than the last."
But there was an IMPLICIT UNDERSTANDING:
* "This is just the test."
The PBT WASN'T REACTING to them.
(Which is different than a student reacting to a PBT, which does happen.)

Per-question CAT: "RESPONSIVE UNCERTAINTY" and "REACTIVE UNCERTAINTY"
Now imagine the student knows:
* "Every answer influences what comes next."
Suddenly every question has two meanings.
It isn't just:
* "Can I solve this?"
It's also:
* "Why did the computer give me this?"
-> That second question simply didn't exist regarding the paper SHSAT.

That/this/there is a subtle but important difference
On paper:
* A question is before you
-> Solve it.
On CAT:
* A question is before you
-> Solve it and
-> Interpret what it says about me.
Even if the student tries not to, it's hard not to wonder.
-> But it's controllable!
-> And you must!

The EMOTIONAL DIFFERENCE between fixed-form and CAT
 may have less to do with the actual adaptive algorithm than with
 THE STUDENT'S MENTAL MODEL OF THE ALGORITHM.
In other words:
Paper SHSAT
* "The test is fixed."
CAT SHSAT
* "THE TEST IS WATCHING ME!"
WHETHER OR NOT THAT'S TECHNICALLY TRUE, that's a very different psychological framing.

Subjectively, however, the impact can still be substantial.
Because that's usually a MISUNDERSTANDING in how a CAT works.
Plus a 13 or 14-year-old may not know any of what we've been discussing.
They may think:
"I just got a really hard question."
-> "Oh no... does that mean I got the last one right?"
Or:
"This one seems easier."
-> "Uh oh... did I just get demoted?"
WHETHER OR NOT EITHER IS TRUE, THE BELIEF ITSELF CAN AFFECT concentration and confidence.
🚨 NOCEBO IN ACTION.

BUT
The biggest psychological issue may not be difficulty changes.
First, the change in difficult will often be minimal and incremental.
Second, while the questions constantly getting harder and easier may be true,
a larger issue is
STUDENTS BELIEVE every question is giving them FEEDBACK about how they're doing.
and so therefore TO THEM EVERY ITEM BECOMES EMOTIONALLY CHARGED.
Especially if they're also thinking, ugh, "it's twice as hard."
🚨 NOCEBO IN ACTION.

It can be self-fulfilling.
Dare I say prophetic.
EVEN IF THE ALGORITHM ACTUALLY MADE ONLY A TINY ADJUSTMENT,
a student might think:
"This question is easy."
-> "I must be doing poorly."
That can lead to:
* rushing
* second-guessing
* overthinking
* panic
* loss of focus
* loopus infinitus
By popular demand:
🚨 NOCEBO IN ACTION.

On a PBT, Question 99 is simply...  Question 99.
ITS difficulty says nothing about YOUR performance.
It was preset, certainly preprinted.
On a CAT:
Students may BELIEVE every question is a SECRET PERSONAL MESSAGE from the COMPUTER.
That changes the psychology of the exam.

BUT
-> "This question feels hard." ≠ "I'm doing badly."
-> This is taking the exam emotionally.
-> What happened to the math or ELA? The question before you?

This overstates things:
The algorithm itself may create dramatic emotional swings
 because the questions change dramatically.
But this is LIKELY A MIXED SIGNAL perception.
-> As the questions likely did not change dramatically.

Instead
The algorithm is LIKELY to ADJUST difficulty+ fairly SMOOTHLY,
but the student's interpretation of those adjustments may create the larger emotional effect.
-> One can't do that

Algorithmic volatility: probably moderate to LOW AFTER the opening portion of the test.
Psychological volatility: potentially moderate to HIGH,
 especially if students know the exam is adaptive
 and start trying to infer their performance from each question.
-> One can't do that
-> Don't try to outsmart the exam.  Ignore it.
-> It's doing a Jedi mind trick.
-> More likely: YOU ARE doing a Jedi mind trick TO YOURSELF.

If the interface gives no indication of whether questions are becoming easier or harder,
and if the difficulty changes are indeed gradual,
many students may not consciously notice much adaptation.

However, students who are aware of the adaptive format
MAY STILL ATTRIBUTE MEANING TO EVERY QUESTION,
even when the algorithm has barely changed its target difficulty.
THAT EXPECTATION ITSELF CAN BECOME PART OF THE TESTING EXPERIENCE.
🚨 NOCEBO IN ACTION.

But this is ABLE TO BE CONTROLLED.
If a student knowns the exam is adaptive, THEY CAN ALSO KNOW THEY CAN CONTROL.
A nocebo effect occurs when negative expectations themselves produce negative outcomes.
Something occurs because you expected it.
-> So don't expect it.
You didn't with the PBT even though it adapted in its own way.
So don't here.

A similar mechanism could occur on a per-question adaptive exam:
* The student knows the exam is adaptive.
* They believe an "easy" question means they're doing poorly.
* They become anxious.
* The anxiety leads to rushed work, second-guessing, or careless errors.
* Those new errors actually lower their performance.
And similarly if they believe it's a "hard" question.
-> The key point is that THE (THEIR) BELIEF CONTRIBUTES TO THE OUTCOME,
not necessarily the adaptive algorithm itself.
-> Ironically, A MISTAKE MAY HAVE A LARGER EFFECT
 than a tiny 0.x-point difficulty adjustment that triggered the worry!

-> Feeling "a certain type of way" is testing by emotions
-> What happened to the math and ELA?

The psychological process might look like this:

A student thinks a question is easier (or contrarily harder)
-> "I must be doing badly."
-> Anxiety increases
-> Concentration decreases
-> More mistakes occur
-> Confidence drops further
That's very similar to a nocebo-type feedback loop.
🚨 NOCEBO IN ACTION.

This ties in with DISTRACTIONS such as hypervigilance and active performance monitoring.
Instead of SIMPLY SOLVING the problem, the student is instead constantly asking:
* "Why did the computer give me this question?"
* "Was that easier?"
* "Did I just move down?"
* "Am I still above the cutoff?"
-> These are questions you usually cannot answer.
Nor need to.
This is similar to always over-dwelling on your score on a PBT.
-> THESE CONSUME working memory, time, and effort, and become distractions.
For a 13 or 14-year-old, who already has limited working memory under stress,
this EXTRA LAYER of self-monitoring may be more DAMAGING than actual adaptive adjustments.
-> Never try to infer how you're doing from the APPARENT difficulty of a question.
-> Don't try to outsmart an adaptive exam
-> Try to ignore it.
-> But: "There is not try, only do."
-> Clear your mind.  Take a breath. Focus on the question before you.
-> Doing this is part of your prep journey.
-> Must.  Always.  Never not.

This may come about because:
* Question difficulty, up or down, isn't always obvious.
You get anxious about being anxious.  A loop.
* Content balancing means the test can't simply present the hardest available item next.
So it doesn't.
* Two questions can feel very different yet be calibrated to similar difficulty.
Don't jump at shadows, no less artificial shadows.
* The CAT is designed to keep questions near your ability,
  so many students will perceive a fairly steady level of challenge.
So yes, a nocebo-like expectation effect is a plausible concern.

However
the LARGER RISK may not be the adaptive algorithm itself,
but the STUDENT'S BELIEF that every question is revealing their standing,
which can DIVERT ATTENTION+ AWAY from solving the problem in front of them.

And don't forget, this is the SHSAT after all, not a picnic.
Don't take it personally.
-> The SHSAT has always been a challenging exam with challenging questions.
-> Focus on that, and not fabricated emotional baggage such as "what it's doing to you."

-> Coming to fruition yielding an interesting paradox
A WELL-DESIGNED CAT ACTUALLY TRIES TO MAKE THE EXPERIENCE FEEL FAIRLY CONSTANT
 BY KEEPING QUESTIONS NEAR YOUR ABILITY.
And builds up in SMALL INCREMENTS,
 even if a student gets every question correct.
Ironically, students may "experience" the opposite and PROJECT that:
* "Every question is telling me something."
when, from the algorithm's perspective, it's simply saying:
* "I'M STILL JUST TRYING TO ESTIMATE YOUR ABILITY."
-> "Pay no attention to that man behind the curtain"
Toto pulls back a curtain to reveal an ORDINARY man operating levers and
creating ILLUSIONS OF A GIANT, TERRIFYING HEAD.
The whole yellow brick road was a scam, at face value.
Things are often not as they appear and are often what we make of them, justified or not.

This has implications for student preparation
A most valuable piece of advice for students taking a per-question adaptive SHSAT would be:
* TREAT EVERY QUESTION EXACTLY AS YOU DID ON THE PAPER SHSAT.
* NEVER TRY TO INFER HOW YOU'RE DOING FROM HOW DIFFICULT A QUESTION FEELS.
* Reset.  Clear your mind.  Breathe.  Do your best.

That's easier said than done, but it addresses the psychological trap directly.
The (projected) responsiveness itself becomes part of the testing experience,
and it can impact performance.
The paper SHSAT had a predetermined mix of difficulties that students largely accepted as
"the test."
A CAT may present a similarly varied experience,
but because students know the sequence depends on their responses,
they're more likely to interpret each question as feedback about themselves,
EVEN WHEN THAT INTERPRETATION ISN'T WARRANTED.
That distinction is subtle, but one of the most interesting psychological differences between a PBT and a CAT.

-> An adaptive exam, practice or otherwise, needs to model this.
-> Adaptation should not be too noticeable if CAT done right.
-> The notion of getting all questions correct may not be every student's reality.
-> Continue to treat the SHSAT as the exam you know and love.
The Adaptive SHSAT
MORE COMING SOON
PART X
July, 2026

The Adaptive SHSAT builds upon the Digital SHSAT

Details on, and resources for, the digital SHSAT from 2025 and beyond follows:

April 16, 2025 NOTE: The DOE finally released some online digital practice exams. They can be found at the NYC SHSAT Portal located at https://nycshsat.myassessmentsupport.com Click on that, scroll to "Practice Tests" and click on "Learn More". Once there scroll to "SHSAT Practice Tests" and follow the directions. Of course, read everything as you go through all these links, reading the info they provide, checking out the SRT, etc., and all the things I've explained below.

My advice would be to hold off on actually doing the online exams for now. First, they are mostly multiple-choice repeats of past exams. Second, they will be the only official online practice exams being made available (2 full practice exams for grade 8, and a smaller math section for grade 9), and so they are a limited resources that you probably should not consider using until toward the very end of your prep; just skim them for now but don't do the exams yet.

On this note, be cautious of 3rd party digital exams that may be made available, as getting questions right is not as easy as it seems, and you want to be sure you're working with quality questions on these new question types, and even of the previous PBT (paper based test) questions.

I will be posting videos with an unboxing and rip down of the exam shortly.

Any questions feel free to message me.

The Digital SHSAT, Let's talk, AMA, PART 1
January 7, 2025
The Digital SHSAT, What we know, What's misinformation, PART 2
January 8, 2025
The Digital SHSAT, A few more details, PART 3
January 28, 2025

A few additional details/expectations:

The Digital SHSAT, Surmising NEW Question Types, ELA, Part 4
January 29, 2025
The Digital SHSAT, Surmising NEW Question Types, MATH, Part 5
January 31, 2025
The Digital SHSAT, Surmising NEW Question Types, MATH, Part 6
January 31, 2025
The Digital SHSAT, AMA (Ask Me Anything) Q&A, Part 7
February 3, 2025

Digital SHSAT AMA Q&A

The Digital SHSAT, AMA (Ask Me Anything) Q&A, Part 8
February 13, 2025

Digital SHSAT AMA Q&A

The Digital SHSAT, DOE UPDATE!, Part 9
February 18, 2025 7pm

Digital SHSAT, DOE UPDATE

The Digital SHSAT, Intro to "The SRT", Part 10
February 21, 2025
The Digital SHSAT, The SRT, NAVIGATION TOOLS, Part 11
February 21, 2025
The Digital SHSAT, The SRT, "MY TOOLS" TOOLS, Part 12
February 21, 2025
The Digital SHSAT, The SRT, "TEST SUPPORTS" TOOLS, Part 13
February 21, 2025
The Digital SHSAT, The SRT, "EQUATION EDITOR" TOOL, Part 14
February 21, 2025
The Digital SHSAT, The SRT, Digital Question Types, Part 15
February 21, 2025
Self-Prep for the 2025 Digital SHSAT Webinar
March 20, 2025
Self-Study for the 2025 Digital SHSAT — GregsTutoringNYC@gmail.com
  1. Welcome and introduction
  2. How to structure an effective self-study PLAN for the digital SHSAT
  3. The most important CONCEPTS and question types to focus on
  4. Time management STRATEGIES to maximize efficiency on test day
  5. Common mistakes students make and how to avoid them
  6. RESOURCES and TOOLS that can aid in preparation
  7. Live Q&A with Kenny Tan and Greg for personalized GUIDANCE
Referenced links: