iiinova — where your digital self keeps evolving

David Webb

@djphysicswebb.bsky.social

212 followers 588 following 66 posts

Thinks a lot about teaching and learning. Retired physicist. Hunts for demographic inequity within the system. #EduSky 🍎 #Science 🧪

Posts Replies Media

David Webb @djphysicswebb.bsky.social · 30/04/2025

Re-posting and including picture showing no gender gap on low stakes in-term exams and significant negative gender gap on medium stakes in-term exams. Gender gap on the high stakes final exam is 50% larger than on the medium stakes exams. #iTeachPhysics 🎢 #Physics ⚛️ #EduSky 🍎 #academia #HigherEd

As explained in the text, each in-term exam was z-scored on a class-by-class basis and the gender gap for each group of classes was determined with HLM while controlling for each student's incoming GPA. This in-term exam score gender gap is plotted as a function of time through the 10-week term at the times the exams and retakes were taken. On the right, after the end of the term, are the overall averages for those first-attempt exams and for the recorded scores after retakes were completed. One sees that the female students generally had higher exam scores (i.e. a positive gender gap) after the first attempt and that, overall, retakes favored the male students. For reference the (clearly negative) gender gaps for 179 non-retake classes are also shown on a week-by-week basis through their terms.

0129

David Webb @djphysicswebb.bsky.social · 17/11/2024

4) Organization of material: Organize material so that students learn the concepts early in the term, then use those conceptual ideas in more complex (and computational) situations later in the term. Historically marginalized students grade gaps completely removed. tinyurl.com/yf27f5kn 6/8

Graph with final exam grade gaps between students from historically marginalized racial/ethnic groups and their peers. Two types of class are compared: 1) traditional physics courses where topics were introduced chapter-by-chapter as usual and 2) a class where all of the concepts from the entire quarter are introduced in the first 60% of the term and then computations are practiced (using those concepts) in the final 40% of the term. On the same final exam (graded at the same time) the concepts first class had no demographic grade gap while the usual classes had grade a grade gap of about 0.7 standard deviations.

100

David Webb @djphysicswebb.bsky.social · 17/11/2024

3) Exams: This may refer mainly to STEM classes. Offer retake exams so students can learn from a poor result and show you that they have learned the material. Gender gap in physics course was completely removed. tinyurl.com/yf27f5kn 5/8

Graph showing a 0.2 grade point gender gap in physics courses where students were not offered retake exams and no gender gap in the same physics course but with retake exams offered to all students.

100

David Webb @djphysicswebb.bsky.social · 17/11/2024

2) Class time: Add class-time for your students to make their own sense of the ideas of the course, through discussion with their peers (Active Learning). Fewer students drop, more students pass, more students graduate AND demographic gaps are reduced or eliminated. tinyurl.com/58duj66f 4/8

Graph showing that active learning physics courses at San Jose State Univ. and at UC Davis had roughly 60% to 80% fewer students dropping the course than dropped the lecture-based courses that they supplanted. Also showing that students from historically marginalized racial/ethnic groups dropped the lecture-based course at higher levels than their peers but dropped the active-learning courses at about the same levels.

Graph showing that active-learning physics courses at San Jose State Univ. and at UC Davis graduated in STEM majors at higher rates than the lecture-based courses that they supplanted. Also showing that the graduation levels for students from historically marginalized racial/ethnic groups and their peers are nearer to each other for students coming from the active-learning classes.

100

David Webb @djphysicswebb.bsky.social · 17/11/2024

1) Grading: Don’t use the percent scale when grading. 70% of a percent grade scale is wasted by being used for fine scale judgements of poor work (D’s and F’s). Demographic gaps were reduced when using a GPA-type (4-point) grade scale. tinyurl.com/yjcwrcw3 3/8

A graph comparing the grade gaps between physics students from historically marginalized racial/ethnic groups from classes using percent scale grade and classes using GPA-like (4-point) scale grading. The grade gaps are nonzero even after controlling for physics understanding so they represent a grading bias. Importantly, the percent scale grade gap is more than three times larger than the 4-point scale grade gap showing that the percent scale amplifies an underlying grade bias against historically marginalized students.

100

David Webb @djphysicswebb.bsky.social · 12/11/2024

When I read that angry GenX men voted 60% Trump, I can't help thinking of these two pieces of data. They had a lot of lead in their blood as toddlers and they were part of the largest violent crime generation in their late-teens/early-20s. Of course also COVID.

Graph showing early life blood levels in US children - showing that if you were 38-50 in 2015 then you had the highest blood levels in your early life. From: https://www.pnas.org/doi/10.1073/pnas.2118631119

Graph showing reported violent crime rate over time - violent crime peaked in early 1990s. From https://www.statista.com/statistics/191129/reported-violent-crime-in-the-us-since-1990/

110