An Unusual Distribution: Part Deux

An Unusual Distribution: Part Deux

In 2014-2015, a study was conducted of exam grades in AP Physics 1 classes at Little Rock Central High. This study was intended to inform recommendations to take this class, as this was the first year it was offered. These classes had nearly equal numbers of students across each of the 9-12 grade levels, which is unusual for a high school elective. Investigations were completed on whether grade level had any statistically significant effect upon performance in those exams. It did not.

In the 2016-2017 school year, this experiment was repeated with a new group of students. These classes also had an unusual distribution of grade levels, with approximately 37% Freshmen, 14% Sophomores, 20% Juniors and 29% Seniors. Again, little to no difference in performance between grade levels was found. However, additional investigations into gender and ethnicity for this year turned up some interesting results!

AP Physics 1 is not an easy class. While intended by the College Board to be an entry-level physics class, the depth of understanding to do well is significant. Higher-level application of algebraic principles is necessary. So, it’s important to understand whether grade level, gender and ethnicity should be factored in to recommending this course for students. About one section of freshmen students, usually with the strongest math skills in their class, takes AP Physics 1. They take it alongside upperclassmen, many of whom have had more experience with math and science. Classes have similar numbers of females and males, and have nearly equal numbers of Asian and White students. Black students make up a sizeable minority, with only a few students of other ethnicities.

The study uses raw exam scores on nine major unit tests throughout the year. Each grade level is compared via a bar graph with standard error bars. T-Tests are conducted for groups that appear to have significance from the graphs, with a p-value of 0.05 or below considered significant and highlighted in the data.

Students were grouped according to their listed grade levels, genders and ethnicities according to district-provided information. Personally identifiable information was removed from the raw data. Grade levels were limited to 9, 10, 11, and 12. Genders were limited to male and female. Though more were available, only three ethnicities were considered for this study due to low sample size of some populations. For this study, White, Black and Asian ethnicities were compared (1, 2 and 4 in the data).

Grade level had three statistically significant events, all involving juniors. They scored lower than freshmen on one occasion, sophomores on another, and were significantly lower than the overall average on a third. Criteria was set for overall significance when a specific event occurred for one third or more of the exams. Since each of these events were different, this was not enough to describe Juniors as lower performing than others. It casts some doubt on the null hypothesis which states that grade level makes no difference, but does not refute it.

Gender, on the other hand, had three identical events occur. Females scored higher on tests than males on each of these three events. Females make up 43% of the AP Physics 1 students for this year. This meets the criteria for significance, so it refutes the null hypothesis that gender makes no difference on exam grades. More studies are needed to confirm this result, but recommendations should take into account that females may be more ready to take this class than males.

Ethnicity experienced no significant events. Sample size for whites and Asians were similar, but fewer black students took the class. This led to higher standard deviation for that segment of the population. Nevertheless, the data supports the null hypothesis: ethnicity has no bearing on performance in AP Physics 1.

When looking at this analysis, considerations should be made about the individual students who take the course. Students with greater passion, experience and ability in science will likely be recommended to take this course earlier rather than later. At Little Rock Central High, two courses of Calculus-based physics are also available, so many upperclassmen take those instead or after they take AP Physics 1. This study is intended to guide changes in the current recommendation scheme, not inform counselors on an entirely new one.

In the end, this study concludes that students of any grade level or ethnicity, given the right background and motivation, can be successful in AP Physics 1. Females are likely to have better performance in the class, and further studies into the causal links between their gender and performance should be done.

Thanks for reading this research summary! Please share it where you can. You can contact me at foleyphysics@gmail.com if you have any questions.

Patrick Foley

Little Rock Central High

Physics Teacher

Copyright 5/12/2017