All things, according to English scholar, preacher, and wit Thomas Fuller, are difficult before they are easy. Some things, however, just stay difficult, as is the case with many questions on the SAT & ACT. Success lies in knowing where the really hard questions can be found.
Clearly, some test questions pose more of a challenge than others in terms of complexity of content or subtlety of traps. Less evident are degrees or patterns of difficulty. The SAT currently rates test questions on a three-point difficulty scale, though you’ll have to wait until you get your score report to learn each question’s rating. ACT, on the other hand, continues to hide internal classifications of question difficulty from the public. Nonetheless, we can see general difficulty trends on both tests:
READING: Completely random passage and question difficulty on the SAT & ACT.
GRAMMAR: Completely random passage and question difficulty on the SAT & ACT.
MATH: Questions progress in approximate order of difficulty on the SAT & ACT. Question difficulty increases from the first multiple-choice question on an SAT Math section to its last, then restarts from the first Student-Produced Response question to the last.
SCIENCE: Completely random passage and difficulty on ACT Science section, but question difficulty increases from the first question on a passage section to the last, then restarts on the next passage.
Knowing where questions are considered difficult can help test takers make better choices. For example, most students shouldn’t be surprised when they answer the first question on a math section quickly and easily; answering the last math question on that section just as quickly, however, should trigger warning bells for all but the most accomplished test takers. Difficulty awareness helps combat the dangers of overconfidence on the tests. That said, three subtle but important factors complicate the issue of question difficulty on the SAT & ACT:
1. Difficulty is relative.
We all bring certain strengths and weaknesses to the tests that render certain questions easier and others much tougher than than expected. Simply put, one student’s pain is another student’s point. Keep in mind that exceptional preparation and practice makes everything easier.
2. Difficulty is probabilistic.
Question difficulty on these exams is not necessarily based on what or even how content is tested. or example, the same punctuation rule assessed in an easy question could be the cornerstone of the toughest question on the same section. Test makers may set out to write questions of a certain difficulty, but true professionals then assess their assumptions through pretesting. Traditionally, College Board and ACT, Inc. ensured that potential questions were tested in statistically valid ways prior to inclusion on official exams. Pretesting makes sure that questions test what they are meant to test without bias, but also provides valuable insight into actual, as opposed to predicted, difficulty. An easy question would be one that perhaps 60-80% of test takers answer correctly, while a tough one would see that many test takers answering it incorrectly. Neither the SAT nor the ACT includes questions that everyone answers right or gets wrong.
3. Difficulty is inconsistently distributed.
Effective pretesting of questions is essential for any truly valid, reliable, and fair standardized exam. Unfortunately, the mechanisms by which the makers of the SAT & ACT pretest questions don’t appear all that effective. The last iteration of the SAT (and every one prior to it) included the notorious equating or experimental section, through which SAT questions could be fairly pretested before appearing on official exams. The College Board recently jettisoned that extra section, while ACT apparently plans to add one starting in September 2018, though in such a way as to guarantee that students ignore the questions. Thus, we might reasonably observe that internal ratings of question difficulty don’t always meet the highest standard of accuracy or reliability.
Furthermore, the test makers sometimes deviate from their typical difficulty distributions. For example, the Student-Produced Response questions on the No Calculator section of the March 2018 SAT did not progress from easy to difficult as usual, but rather all rated as difficult. Despite the fact that some of them were pretty easy despite their 3-point ratings, these questions serve as solid evidence that difficulty trends should be recognized as guidelines rather than strict rules.
Question difficulty on the SAT & ACT is, itself, a bit difficult to comprehend and master. But as with every other aspect of these exams, skill at recognizing difficulty and using it to your advantage will lead to more points on test day. That’s not so hard to understand, is it?
Mike Bergin