

AP World History Statistics: How Are U.S. Students Scoring?Bill StricklandEast Grand Rapids High School 

1. % Correct
This shows the percentage of students who answered each question on the 2002 AP World History Exam correctly. (e.g. 90% of students answered Question #9 correctly, while only 9% answered Question #64 correctly). Note: Question #29 was thrown out (presumably because it did not meet College Board /ETS standards^{1}). There are a few variations of the basic chart: A. % Correct: The basic line graph.^{2} B. Previous 10: This graph shows how well students performed on the most recent 10 questions. It seeks to answer students' question, "How am I doing lately?" Note how the Previous 10 average falls throughout the exam. C. Cumulative Average: This shows the overall running average of all questions. Note how it stays roughly constant for the first half of the exam, then inexorably drops for the rest of the exam. D. By Tens: This is a variation on the Previous 10, except that rather than the most recent 10 questions, it shows how students performed on Questions #110, then on #1120, etc. Note how students seem to hit the "Wall" around Question #40. E. Data Sheet: This is the raw data that I've used to create the graphs. 
1 
I have no way to quantitatively prove this, but
my anecdotal experience in the classroom tells me that students lose their
mental focus after ~35 questions. Better students may recover, but I think
this chart shows how important it is to have a demanding semester exam (with
~70 APquality questions) that forces students to keep concentrating for
a solid hour of sustained mental effort. When students finish the Multiple
Choice section of the Exam, they should not expect to feel "fresh" or "invigorated."
I tell my students that they should expect to feel like "mental JellO."
It is normal to feel overwhelmed by the difficulty of the questions!
Keep a stiff upper lip and prepare for the Essay section! 
2 
2. AP World History Essay Scores Stats 
3 
National Mean Scores: This bar chart just shows the national mean scores on the various essay questions. I created this chart simply because I was curious what a "good" score was. (Is a '5' on the DBQ "good"? Is a '3' on the C&C "bad?") Obviously, it varies tremendously from year to year, and question to question.  4 
Essay Score Distribution:. Note the trend from DBQ through COT to
C&C. The average score drops, the frequency of No Response essays increases.
Students might lose track of time during the exam. Students might become
tired and less able to write a quality essay as they struggle through the
two hour of the writing portion. 
5 
How to use this in the classroom: When your students write their first DBQ of the year and "only" score a '3' these graphs can encourage them that they are not as substandard as they may fear. When students are writing their essays on the exam and are saying to themselves, "I don't know what I'm doing. I have nothing to say!" This chart can reinforce for them the importance of every single essay point. ("Think of the difference between getting a '2' and a '3' on each essay.")  6 
Teachers should realize that there are 27 possible points on the Essay section (9 points x 3 essays). The median score of the entire essay section has averaged only 10.37! Getting your students to write three 4point essays will put them ABOVE the national average.  7 
3. APWH MC Score Distribution  
This is the "big" file, and has several graphs that can be tricky to interpret. Combination of Multiple Choice & Essay Scores:^{3} Students often make the mistake of thinking that doing well on one half of the exam guarantees them a high score, while doing poorly on one half ruins their chances. This graph is designed to dispel both of these misconceptions. A student who gets 100% of the Multiple Choice questions correct (and then walked out of the exam) would score a '3'. Same with a student who didn't answer any Multiple Choice questions correctly, but wrote three perfect 9point essays. Note: the lines for this graph are approximate only. The precise levels are determined by the Chief Reader and ETS immediately after the essay grading is completed.^{4} 
8 
Correlation Between M/C Score and Final Score My students always ask me, "I got x% of the Multiple Choice questions correct on the semester exam. What AP score would that get me?" This graph is designed to answer that question as nearly as possible. Note: there is no absolute answer, only a probability. Remember that the national average score on the multiple choice section of the exam is approximately 50%. In the center of this graph you'll see a vertical line labeled "50%". Students who answered 50% of the multiple choice section questions correctly were most likely to earn a '3', but some who wrote better essays earned a '4', while those with weaker essays earned a '2.' Note: a small percentage of students earned a '5' or '1'. Obviously the higher one's multiple choice section score, the farther right on the graph and the higher the likely Final Score. It is important to note, though, that it is possible for a student to score 70% on the multiple choice section and earn a final score of '3', while another student might score only 50% on the multiple choice section and earn a final score of '5'. The lesson is that there are NO GUARANTEES! It is POSSIBLE (though not likely) to answer 30% of the multiple choice questions correctly and still earn a '3', but only by writing outstanding essays. It is also POSSIBLE to answer 70% of the multiple choice questions correctly and still earn a '3', but only by writing poor essays. Teachers: use this chart to impress the importance of wellroundedness on your students. They can't be good at only one part of the exam!  9 
Hypothetical Student Scores There is a score computation worksheet in the 2002 AP World History Released Exam, (p. 87) but it has the userfriendliness of an IRS form. This chart is designed to show how various realistic student scores would be scored by the AP Exam. Note: probably the most important fact is the note that each Essay point is weighted the same as 2.07 Multiple Choice questions.^{5}  10 
Student A: This student correctly answered only 25 out of 70 multiple choice questions, incorrectly answered 30 questions, and skipped 15 questions. In addition, this student earned only 5 points on their three essays.^{6} Student A's Composite Score was 26, just shy of the Composite Score of 27 needed for a Final Score of '2'. Student B: This student did a little better on the multiple choice section (30 correct, 30 incorrect, and 10 skipped) and also wrote slightly better essays. Their Composite Score was 37, squarely in the middle of the '2' range. Students C, D, & E: I structured three hypothetical students together to highlight a few principles: All three of these students score the same Final Score ('3') and virtually identical Composite Scores (56, 57, 58). The point is that Student C is a wellrounded student, evenly able in both the multiple choice and essay sections. Student D is the student who does poorly on the multiple choice section, but is a better writer than his/her multiple choice section score would predict. Student E is a gifted multiple choice question answerer, but is weak on the writing portion of the exam. ALL THREE STUDENTS SCORE THE SAME. Student F: This student illustrates what is necessary to earn a '5'. Perfect: Just in case students start to complain, I remind them that someone who answered all multiple choice and essays perfectly would earn a Composite Score of 120. Getting a Final Score of '5' requires a Composite Score of 'only' 78 (65%). 
11 
Detailed Hypothetical Student Scores: This is just a variation that shows the essays broken down individually. All the data is identical to the previous graph.  12 
Should I Guess? Occasionally a student will be so afraid of failure that they won't answer a question unless they're absolutely sure of the answer. This is too cautious for AP World History. This graph shows the effect that guessing would have on a student's score. The first bar shows how a 'perfect' student would score on the multiple choice section of the exam. The second bar shows a student's score if the student narrowed the answers down to two, then guessed. (Remember that there is a 1/4 point 'penalty' for incorrect answers, which neutralizes the effect of random guessing). The point of this chart is that guessing between two answers is NOT random, and will in fact help a student's score. Guessing between 3 possible answers will also help, though only 16.25% as much as a consistent right answer. How should Teachers use this graph? Talk with your students. Reassure them that guessing between 2 or 3 answers WILL help their overall score. Leaving a question blank is almost as bad as answering it incorrectly. Note: While randomly guessing among 4 answers will officially help students' scores (+6.25%), one has to factor in the TIME it takes students to read and answer each question. Many students have difficulty finishing the exam, and may not want to spend valuable time guessing between 4 answers on a question when they still have several more questions to answer.  13 
Weighted MC Score: and % Correlation: This is the graph that started it all. On page 88 in the 2002 Released Exam there is a table that shows how a certain score on the Multiple Choice section correlates to students' Final Score. Unfortunately, it's a table, not a graph. I've simply attempted to convert the numerical info in the Released Exam to an easy to read graph. Note: the graph assumes that students have answered ALL the multiple choice questions.  14 
4. APWH Nat'l Score Distrib by Grade  
Student Distribution: This graph simply shows what percent of AP World History students are at each grade level. One common question that teachers, administrators, and parents ask is, "Can sophomores handle the demands of an AP course?" While these graphs can't give specific advice on this question, they can show that the overwhelming majority of APWH students are in fact sophomores.  15 
Grade Level Age vs. Exam Score: How do students of different ages perform on the APWH Exam? This graph addresses this question. Note how there is significant improvement in students' performance between Freshman and Sophomore, and Sophomore and Junior years, but there is only a slight difference between Juniors' and Seniors' performance.  16 
Age vs. Passing Rate, and Passing Rate vs. Mean Score: This data is NOT specific to World History exclusively, but compares how different age students perform on ALL AP subjects. Can Freshmen take an AP course? Well, that depends on how one interprets this graph's data. Certainly there is a (very) select group of Freshmen in the country who can take AP courses and pass AP exams, but their number is so small, that they really fall into the "exception" category. Freshmen passing rate and mean scores are the highest of any age group, but there are so few Freshmen compared to Juniors and Seniors that it is difficult to make accurate generalizations from this graph alone.  17 
5. APWH Nat'l Score Distrib History  
APWH % Score Distrib: How hard is it to get a '5'? Has the % of students earning a '2' changed since World History started? This graph just shows the simple Final Score distributions since the beginning of the AP World History exam. Note how the % of lower scores (12) has increased slightly in the past 2 years. I don't know exactly why this is, but I suspect it's because the number of students taking the World History exam has grown so quickly, that many schools and teachers are now offering the course with less than adequate teacher training and student preparation. I have no hard factual data to support this conclusion, but it seems to fit the data apparent in this graph.  18 
APWH # Score Distribution: This is just a variation on the above, with the raw number of students listed rather than the percentage. I don't think there's anything terribly significant about these numbers, but I didn't know what I might see until I created the graph.  19 
APWH Mean Grade, Mean vs. Standard Deviation, Growth Rate: Each of these charts are fairly selfexplanatory, but I'm not enough of a statistician to know if the figures are significant. As AP World History has grown, the average grade has dropped slightly.  20 
6. Why Take AP  
Why Take AP: I use this graph at Open House/Back to School night. While I think the benefits of AP in helping students earn college admissions are well documented, parents are ultimately concerned with seeing their students through college, not just into college. This graph addresses the question of how AP influences students' ability to graduate college. When I show this graph to parents, I talk about how the AP exams are a good indicator of future college success, but even just the experience of taking the AP course is "half the battle." The information for these graphs comes from Tom Luce and Lee Thompson, Do What Works: How Proven Practices Can Improve America's Public Schools (Accent Education Press, 2005).  21 
Biographical Note: Bill Strickland teaches AP World History at East Grand Rapids High School in East Grand Rapids, Michigan. Notes ^{1} Each exam question is analyzed to ensure that performance on any given question does not greatly vary between set populations, such as males, females, whites, African Americans, and Latinos. On the rare occasions when such analysis shows that the wording of an exam question might have contributed to inequitable performance by one set population, the question is omitted from the scoring. (Excerpted from http://apcentral.collegeboard.com/exam/0,,152000,00.html). ^{2} Source: 2002 AP World History Released Exam, p. 43. ^{3} The format for this chart was suggested by one of my students, Alex Constantelos, '08. ^{4} The complete description of the Scoring
Process is described in detail at ^{5} This is a rather tricky calculation. Depending on how one defines "M/C Question" this ratio could change. Technically the 27 possible points on the Essay section are worth the same as the 70 M /C Questions, which would make each Essay Point worth 2.59 M/C questions. However, when answering a M/C Question, most students choose between answering the question correctly and answering it incorrectly, not between answering correctly and skipping the question entirely. Thus a single M/C Question is really worth 1.25 points on the M/C section, taking into account the 0.25 penalty for incorrect answers. Thus the formula I used when calculating this ratio is Essay Question = (2.2222/(1.25 x 0.8571)) = 2.07. This is all labeled as clearly as I can make it on the Data Sheet in the M S Excel file. ^{6} ^{ }It doesn't matter which essay earned which score. A point on any essay (DBQ, COT, or C&C) is worth the same as a point on any other essay.


Home  List Journal Issues  Table of Contents 
© 2008 by the Board of Trustees of the University of Illinois 
