Thursday, March 26, 2015

Impressions on the new practice PSAT

A few days ago the College Board posted a new PSAT practice test.  This is the practice test for the redesigned PSAT to be given in October 2015.  It is also a glimpse of what the redesigned SAT might look like. 

There has been a great deal of speculation about the redesigned test.   A number of people have opined that the new test would be an “ACT clone.”  I, myself, have speculated that the new test would be designed as more of a high school exit exam than a college entrance exam.  Quite a few pundits have pointed out that the SAT was losing market share to the ACT and thought that the redesign might be an effort to gain that share back.

Upon looking over the test, I am now prepared to make the following statements:

The writing section IS barely distinguishable from the writing portion of the ACT.  Otherwise, this test is in no way an ACT clone.  On the other hand, I do feel that this test represents a fundamental shift in purpose.

There are a number of possible purposes for giving a standardized exam.  (This would be as opposed to a teacher-made assessment.)  Here are a few:

  • To document student learning (or lack thereof) of particular skills and concepts
  • To distinguish among (or rank) students
  • To drive the curriculum*


Prior to 2015, the PSAT’s fundamental purpose has been a balance of the first two items.  As the National Merit Scholarship Qualifying Test it has been used to distinguish top students from the pack, and it has documented whether or not students have a grasp of particular skills and concepts.  It has NOT been used to drive the curriculum...until now.  In my opinion, several of the changes were specifically designed to have an impact on the nature and content of classroom instruction.

One of the elements of Common Core language arts is a focus on having students read critically.  In an effort to have the students respond to what the author actually said – as opposed to how they feel about what they think the author said – students are being asked to point to pieces of the text as “evidence.”  In the PSAT critical reading portion, students are asked to choose the best evidence for almost every question.  Now I’m not a language arts teacher, and I don’t have any special critical reading or psychometric expertise.  However, it seems to me that this is about the same as asking the same question twice.  In other words, if you correctly answer the original question, then the answer to the “evidence” follow up is trivial.  If you missed the original question it would be impossible.  Will there be some kind of scoring mechanism that uses this question-pairing to determine when the correct answer was obtained by guessing?  Maybe.  It is more likely an attempt to make sure teachers require their students to give evidence in class.

Here is another example from the math section.  Months ago, when the sample questions were released, I noticed that some of the questions were much longer and more involved.  This represents a distinct shift.  Up until now I have told my top students, “If you are more than 3 steps into an algebra process, you probably missed something.”  One of my main criticisms of some of the test prep books out there has been that too many of the math questions can ONLY be solved with the application of tedious algebraic steps and are thus not representative of the real thing.  However, this has changed with the PSAT practice test.  It’s interesting.  If I’m a professor of mathematics or engineering I’m interested in multiple different aspects of their math skills.  I want them to have a solid grasp of the concepts, good number sense, AND I want them to be able to keep track of what they are doing through a long problem that requires many steps and sub-steps to solve.

Up until now, the major college entrance exams did a decent job of testing the first two. (The SAT was better than the ACT in my opinion.) They really haven’t attempted to do the third.  This is largely because a multiple choice or single final answer exam format is a TERRIBLE way to assess that third skill.  If the student’s answer is incorrect, you don’t know if he really can’t negotiate the process or if he just made a silly error in the middle.  Graded homework and teacher-made assessments where partial credit is given are much better means of determining whether or not the student can handle a long problem.  Both of those would be reflected in the students’ grades.  By trying to test it in this format, you add no new information.  The only reason I can come up with to include problems like this is to encourage math teachers to have students practice longer, more complex problems.

When the College Board first announced that the SAT would be re-designed, an admissions officer at a small, selective school wondered in an online forum, “Will the rSAT do a better job of distinguishing among students at the top end of the spectrum?”  This was what she was hoping for.  Other users of the forum predicted that it would not.  To do so would require a test with a wider standard deviation, and a large contingent of the (math-challenged) public believes that a wide standard deviation is inherently “unfair.”  In fact, the speculation was that the new test would do a worse job of highlighting differences among students.  Given that the redesigned test appears to have abandoned that goal altogether in favor of driving the curriculum, I’d say that admissions officers at selective schools will be plumb out of luck.



* You may be wondering what it means to have a test “drive the curriculum.”  Let’s look at an extreme example.  Throughout the 1990’s and early 2000’s North Carolina had a state writing assessment.  In fourth, eighth and tenth grades students had to write a timed essay and send it off to be scored by specially trained “experts.”  After several years of dismal results, some statisticians called “foul.”  They pointed out that the scoring methods were severely flawed, and thus that the scores were ultimately meaningless.  (By the way, the scoring methods used to score the ACT and SAT essays have some of the same issues.)  The state’s surprising response?  “Yes, we know.”  Students (and by extension their teachers and parents) suffered through this test for YEARS.  Why?  Because if there’s a writing test – even a flawed one - teachers will spend time teaching writing.  The average amount of classroom time spent on writing instruction quadrupled.  Prior to the test, some teachers had spent ZERO time teaching writing.


To pre-order the practice book for the redesigned SAT:

Sunday, March 15, 2015

Something to think about if your kid has received a college rejection letter

I was sitting in a metal folding chair in a hotel ballroom. The speaker had just said something that hit me so hard that I couldn’t listen to what she said next. I needed to stop and process that statement. I was thinking, “Holy crap!” and “Well, now that she’s said it, it seems so obvious. Why didn’t I figure this out before?”

The speaker was head of admissions at an elite private university. She was talking to a group of elite high school students and their parents, and what she said was this, “It is not my job to make the perfect admissions decision on your kid. No one is paying me to do that. It is my job to build the next freshman class.” She went on to point out something we already knew: The school could only accept a small fraction of the fully qualified applicants.

As high school students, or parents of high school students, or maybe just members of the public with an opinion, we tend to view the process from the outside in. And it’s like looking into a fish bowl; the view is distorted. We tend to walk around with an impression of the process, and we think that it works something like this: The school takes all of the applications and ranks them from first to last. Then it takes enough applications off the top of the list to fill the class, and boom! Done! And because we subconsciously realize that it would be impossible to rank the kids from first to last without considering numerical measures like GPA’s and test scores, we assume these must play a very large role in the process. If my kid doesn’t get in, that means he didn’t rank high enough on the list. The fact is, it doesn’t work that way. A lot of factors come into play. At the end of the admissions period, the incoming class needs to have all kinds of kids: Some who play on the football team and some who play in the band, some who join the Young Democrats and some who join the Young Republicans, some who build robots and some who work on the school newspaper. All of them need to have a shot at doing the course work.

Lots of factors come into play, and it is certain that the school did not make the exact right admissions decision for every single applicant. They may not have made the exact right decision on your kid. I’ve been there, and you have my sympathy. It doesn’t mean they’ve made a value judgment about your kid, and it doesn’t mean he wouldn’t have done well there. It just means when they selected the components of the class, they didn’t select your kid. They may have made a mistake. Or maybe not. Either way, the system isn’t broken. Because making the perfect decision on your kid wasn’t one of their goals in the first place.