ACM Transactions on Computing Education, Dec 31, 2020
Success in an introductory college computer science (CS) course encourages students to major and ... more Success in an introductory college computer science (CS) course encourages students to major and pursue careers in computer science and many other STEM fields, whereas weak performance is often a powerful deterrent. This article examines the role of high school course taking (AP, regular, or none) in mathematics and in CS as predictors of later success in college introductory computer science courses, measured by students' final grades. Using a sample of 9,418 students from a stratified random sample of 118 U.S. colleges and universities, we found that the observed advantage of taking AP calculus over taking AP CS, seen in an uncontrolled model, was largely confounded by students' background characteristics. After applying multinomial propensity score weighting, we estimated that the effects of taking AP calculus and AP CS on college CS grades were similar. Interestingly, enrollment in both AP calculus and AP CS did not have any additional positive effect, suggesting that both AP calculus and AP CS strengthened similar skills that are important for long-term CS achievement. Taking regular CS did not have a significant effect; taking regular calculus had a positive effect, about half the size of taking AP calculus or AP CS. Thus, the study showed that simply exposing students to any kind of CS course before college does not appear to be sufficient for improving college CS performance; and that advanced CS and advanced calculus in high school may substitute for each other in the preparation of college CS.
Massive open online course (MOOC) studies have shown that precourse skills (such as precomputatio... more Massive open online course (MOOC) studies have shown that precourse skills (such as precomputational thinking) and course engagement measures (such as making multiple submission attempts with assignments when the initial submission is incorrect) predict students' grade performance, yet little is known about whether these factors predict students' course retention. In applying survival analysis to a sample of more than 20,000 participants from one popular computer science MOOC, we found that students' precomputational thinking skills and their perseverance in assignment submission strongly predict their persistence in the MOOC. Moreover, we discovered that precomputational thinking skills, programming experience, and gender, which were previously considered to be constant predictors of students' retention, have effects that attenuate over the course milestones. This finding suggests that MOOC educators should take a growth perspective towards students' persistence: As students overcome the initial hurdles, their resilience grows stronger. 1 | INTRODUCTION The massive open online course (MOOC) was formally introduced to the internet in 2011 (Ng & Widom, 2012). By the year 2017, more than 9,000 MOOCs have come into existence, hosted by more than 800 higher education institutions, serving more than 80 million learners (Shah, 2018). MOOCs have no entry requirements and are easy to access (Kop, 2011; Lee, 2017), have huge numbers of participants (Cohen & Soffer, 2015; Sharples et al., 2012), often partner with prestigious higher educational institutions (Cusumano, 2014), and charge a low or no fee for a wide range of materials, such as lecture videos, online discussion forums, and assessments (Thompson, 2011).
Massive open online courses (MOOCs) show highly irregular participation behaviour among users. In... more Massive open online courses (MOOCs) show highly irregular participation behaviour among users. In this study, using data from Computer Science 50x of HarvardX, we investigated one extreme, yet common strategy to foresee the endgame: taking the final problem set at the beginning of the course. We found such a strategy to be the only dominant trajectory alternative to following the sequence prescribed by the syllabus. Whereas all students who took and passed the final problem set at the beginning of the course subsequently completed the course, those who took and failed the final problem set at the beginning of the course finished the fewest number of milestones, even fewer than those who never attempted the final problem set. Moreover, students with a lower prior programming proficiency were more likely than better prepared students both to take the final problem set early and to fail it. This study revealed the disconcerting phenomenon that many students dropped out of a MOOC because, apparently, their confidence was crushed even before they learned any course content. The study suggests that future MOOC practices and policies should offer informative and constructive syllabi to accommodate students' need for previewing the endgame.
Beginning in 1990, a group of scientists, engineers and educators based at the Harvard-Smithsonia... more Beginning in 1990, a group of scientists, engineers and educators based at the Harvard-Smithsonian Center for Astrophysics (CfA) developed a prototype of a small, inexpensive and fully integrated automated astronomical telescope and image processing system. The MicroObservatory combines the imaging power of a cooled CCD, with a self contained and weatherized reflecting optical telescope and mount. A microcomputer points the telescope and processes the captured images. Software for computer control, pointing, focusing, filter selection as well as pattern recognition have also been developed. The telescope was designed to be used by teachers for classroom instruction, as well as by students for original scientific research projects. Probably in no other area of frontier science is it possible for a broad spectrum of students (not just the gifted) to have access to state-of-the-art technologies that allow for original research projects. The MicroObservatory has also been designed to be...
Project STAR (Science Teaching through its Astronomical Roots) emerged from growing concern about... more Project STAR (Science Teaching through its Astronomical Roots) emerged from growing concern about the decreasing enrollment in science classes, and lack of understanding of science and math concepts (Figure 1). Using astronomy as a focus, Project STAR is trying not simply to increase the enrollment in high school science courses, but also to improve the students’ understanding of science and its role in making sense of the world.The educational approach of Project STAR is based on three principles:• Mastery of a few ideas is more important for students than cursory exposure to many concepts.• Students learn best through hands-on activities.• Students enter the classroom with certain preconceptions, or “naive theories,” about how physical systems work, without a sense of whether their understandings are accurate (often they are not).
Journal of Astronomy & Earth Sciences Education (JAESE), 2016
This article uses U.S. survey data (N=15,847) to characterize the evolution of student interest i... more This article uses U.S. survey data (N=15,847) to characterize the evolution of student interest in an astronomy career in the period between middle school and the beginning of college. We find that middle school students have a relatively high interest in astronomy, which sharply declines with every phase of their education. However, many of the students who leave astronomy - particularly male students - feed heavily into other STEM disciplines. Through statistical modeling, we find that students who spend extracurricular time observing stars, tinkering with mechanical or electrical devices, or reading/watching science or science fiction are significantly more likely than students who do not engage in these activities to hold an interest in pursuing an astronomy career at the end of high school. We also find that females who observe stars during extracurricular time show a greater improvement in their odds of pursuing astronomy than males do. Furthermore, we find that these outside-...
Zenodo (CERN European Organization for Nuclear Research), Sep 3, 2013
To the Teacher vii viii This project has been the beneficiary of many previous development effort... more To the Teacher vii viii This project has been the beneficiary of many previous development efforts, both our own and those of others. We have attempted to credit all those who helped directly, but we also wish to thank those who remain nameless, teachers with a great suggestion at a workshop, a clever idea from an article, a scientist's solution to a difficult problem, and most of all the thousand or so middle school students who contributed their time and passion as we worked out the bugs in our design of DESIGNS.
Zenodo (CERN European Organization for Nuclear Research), Sep 3, 2013
This project has been the beneficiary of many previous development efforts, both our own and thos... more This project has been the beneficiary of many previous development efforts, both our own and those of others. We have attempted to credit all those who helped directly, but we also wish to thank those who remain nameless, teachers with a great suggestion at a workshop, a clever idea from an article, a scientist's solution to a difficult problem, and most of all the thousand or so middle school students who contributed their time and passion as we worked out the bugs in our design of DESIGNS.
ACM Transactions on Computing Education, Dec 31, 2020
Success in an introductory college computer science (CS) course encourages students to major and ... more Success in an introductory college computer science (CS) course encourages students to major and pursue careers in computer science and many other STEM fields, whereas weak performance is often a powerful deterrent. This article examines the role of high school course taking (AP, regular, or none) in mathematics and in CS as predictors of later success in college introductory computer science courses, measured by students' final grades. Using a sample of 9,418 students from a stratified random sample of 118 U.S. colleges and universities, we found that the observed advantage of taking AP calculus over taking AP CS, seen in an uncontrolled model, was largely confounded by students' background characteristics. After applying multinomial propensity score weighting, we estimated that the effects of taking AP calculus and AP CS on college CS grades were similar. Interestingly, enrollment in both AP calculus and AP CS did not have any additional positive effect, suggesting that both AP calculus and AP CS strengthened similar skills that are important for long-term CS achievement. Taking regular CS did not have a significant effect; taking regular calculus had a positive effect, about half the size of taking AP calculus or AP CS. Thus, the study showed that simply exposing students to any kind of CS course before college does not appear to be sufficient for improving college CS performance; and that advanced CS and advanced calculus in high school may substitute for each other in the preparation of college CS.
Massive open online course (MOOC) studies have shown that precourse skills (such as precomputatio... more Massive open online course (MOOC) studies have shown that precourse skills (such as precomputational thinking) and course engagement measures (such as making multiple submission attempts with assignments when the initial submission is incorrect) predict students' grade performance, yet little is known about whether these factors predict students' course retention. In applying survival analysis to a sample of more than 20,000 participants from one popular computer science MOOC, we found that students' precomputational thinking skills and their perseverance in assignment submission strongly predict their persistence in the MOOC. Moreover, we discovered that precomputational thinking skills, programming experience, and gender, which were previously considered to be constant predictors of students' retention, have effects that attenuate over the course milestones. This finding suggests that MOOC educators should take a growth perspective towards students' persistence: As students overcome the initial hurdles, their resilience grows stronger. 1 | INTRODUCTION The massive open online course (MOOC) was formally introduced to the internet in 2011 (Ng & Widom, 2012). By the year 2017, more than 9,000 MOOCs have come into existence, hosted by more than 800 higher education institutions, serving more than 80 million learners (Shah, 2018). MOOCs have no entry requirements and are easy to access (Kop, 2011; Lee, 2017), have huge numbers of participants (Cohen & Soffer, 2015; Sharples et al., 2012), often partner with prestigious higher educational institutions (Cusumano, 2014), and charge a low or no fee for a wide range of materials, such as lecture videos, online discussion forums, and assessments (Thompson, 2011).
Massive open online courses (MOOCs) show highly irregular participation behaviour among users. In... more Massive open online courses (MOOCs) show highly irregular participation behaviour among users. In this study, using data from Computer Science 50x of HarvardX, we investigated one extreme, yet common strategy to foresee the endgame: taking the final problem set at the beginning of the course. We found such a strategy to be the only dominant trajectory alternative to following the sequence prescribed by the syllabus. Whereas all students who took and passed the final problem set at the beginning of the course subsequently completed the course, those who took and failed the final problem set at the beginning of the course finished the fewest number of milestones, even fewer than those who never attempted the final problem set. Moreover, students with a lower prior programming proficiency were more likely than better prepared students both to take the final problem set early and to fail it. This study revealed the disconcerting phenomenon that many students dropped out of a MOOC because, apparently, their confidence was crushed even before they learned any course content. The study suggests that future MOOC practices and policies should offer informative and constructive syllabi to accommodate students' need for previewing the endgame.
Beginning in 1990, a group of scientists, engineers and educators based at the Harvard-Smithsonia... more Beginning in 1990, a group of scientists, engineers and educators based at the Harvard-Smithsonian Center for Astrophysics (CfA) developed a prototype of a small, inexpensive and fully integrated automated astronomical telescope and image processing system. The MicroObservatory combines the imaging power of a cooled CCD, with a self contained and weatherized reflecting optical telescope and mount. A microcomputer points the telescope and processes the captured images. Software for computer control, pointing, focusing, filter selection as well as pattern recognition have also been developed. The telescope was designed to be used by teachers for classroom instruction, as well as by students for original scientific research projects. Probably in no other area of frontier science is it possible for a broad spectrum of students (not just the gifted) to have access to state-of-the-art technologies that allow for original research projects. The MicroObservatory has also been designed to be...
Project STAR (Science Teaching through its Astronomical Roots) emerged from growing concern about... more Project STAR (Science Teaching through its Astronomical Roots) emerged from growing concern about the decreasing enrollment in science classes, and lack of understanding of science and math concepts (Figure 1). Using astronomy as a focus, Project STAR is trying not simply to increase the enrollment in high school science courses, but also to improve the students’ understanding of science and its role in making sense of the world.The educational approach of Project STAR is based on three principles:• Mastery of a few ideas is more important for students than cursory exposure to many concepts.• Students learn best through hands-on activities.• Students enter the classroom with certain preconceptions, or “naive theories,” about how physical systems work, without a sense of whether their understandings are accurate (often they are not).
Journal of Astronomy & Earth Sciences Education (JAESE), 2016
This article uses U.S. survey data (N=15,847) to characterize the evolution of student interest i... more This article uses U.S. survey data (N=15,847) to characterize the evolution of student interest in an astronomy career in the period between middle school and the beginning of college. We find that middle school students have a relatively high interest in astronomy, which sharply declines with every phase of their education. However, many of the students who leave astronomy - particularly male students - feed heavily into other STEM disciplines. Through statistical modeling, we find that students who spend extracurricular time observing stars, tinkering with mechanical or electrical devices, or reading/watching science or science fiction are significantly more likely than students who do not engage in these activities to hold an interest in pursuing an astronomy career at the end of high school. We also find that females who observe stars during extracurricular time show a greater improvement in their odds of pursuing astronomy than males do. Furthermore, we find that these outside-...
Zenodo (CERN European Organization for Nuclear Research), Sep 3, 2013
To the Teacher vii viii This project has been the beneficiary of many previous development effort... more To the Teacher vii viii This project has been the beneficiary of many previous development efforts, both our own and those of others. We have attempted to credit all those who helped directly, but we also wish to thank those who remain nameless, teachers with a great suggestion at a workshop, a clever idea from an article, a scientist's solution to a difficult problem, and most of all the thousand or so middle school students who contributed their time and passion as we worked out the bugs in our design of DESIGNS.
Zenodo (CERN European Organization for Nuclear Research), Sep 3, 2013
This project has been the beneficiary of many previous development efforts, both our own and thos... more This project has been the beneficiary of many previous development efforts, both our own and those of others. We have attempted to credit all those who helped directly, but we also wish to thank those who remain nameless, teachers with a great suggestion at a workshop, a clever idea from an article, a scientist's solution to a difficult problem, and most of all the thousand or so middle school students who contributed their time and passion as we worked out the bugs in our design of DESIGNS.
Uploads
Papers by Philip Sadler