This page was last updated November 14, 2026.
This page describes how research on the teaching and learning of physics (physics education research, or PER) has shaped this book: what the research says about how students learn upper-division electricity and magnetism, and about learning in general, and what the book does in response. Each finding gives its source, and the full references are at the bottom of the page. The page is updated as the book changes. Some of the research is the author's own: since 2019 he and colleagues at BYU and the University of Central Arkansas have studied BYU's first-semester upper-division course in electricity and magnetism, for which the first-semester chapters of this book were written. That work has its own section below.
A caution about the evidence: most of this research documents what students find difficult, and much less of it tests what a textbook can do about a difficulty. The controlled studies cited here (on practice testing, interleaved practice, pretests, refutation text, and the placement of figures) are mostly from introductory courses or from other subjects. The book's responses are informed judgment, guided by the research rather than proven by it.
The research in brief
- The hardest part is setting up the mathematics, not doing it. Upper-division students usually carry out the calculus well. Their difficulties come in choosing a method, building an integral for a particular physical situation, and connecting the mathematics to the physics (Pepper et al. 2012; Wilcox et al. 2013b).
- Choosing a method is the most persistent difficulty. Asked on an exam for the potential of a charged disk, 27% of 172 students first found the electric field, and 11% tried Gauss's law, which the disk's symmetry does not allow (Wilcox et al. 2013b). Recognizing that a problem calls for an integral is a difficulty that persists from middle-division mechanics into upper-division magnetostatics (Wilcox and Corsiglia 2019).
- Students rarely check their answers unless asked to. Only 8% of those 172 exam solutions included an explicit check (Wilcox et al. 2013b). In think-aloud interviews at BYU, students began every problem by drawing the situation but checked their answers in only 4 of 12 problems (Hecht et al. 2025).
- Pictures of fields are often misread. Students take divergence to mean that field lines spread apart and curl to mean that they bend (Baily et al. 2015; Bollen et al. 2016; Stuckey and Quan 2024). Asked about the steady current in a wire that narrows, 51.2% of 284 students said that the divergence of the current density is not zero there, because the current density grows or its lines crowd together (Baily et al. 2017). Shown the standard picture of a plane electromagnetic wave, about 60% of introductory students said that the field is zero at a point outside the drawn curves (Ambrose et al. 1999).
- Research-based teaching helps at this level too. On a conceptual test of upper-division electrodynamics given at 9 institutions, students in courses with interactive teaching (concept questions, tutorials) averaged 59.3%, and students in standard lecture courses 41.5% (Baily et al. 2017).
- In the author's own course, students who reworked missed exam problems for partial credit did better on the same problems on the final exam, largely regardless of how they had done the first time (Mason et al. 2023); students valued tutorials built on the steps of a problem-solving framework, though some found the steps too rigid (Mason et al. 2024; Hecht et al. 2025); and students rated capacitance, multipoles, and the vector potential as the hardest topics of the semester, with the fields in matter not far behind (Mason et al. 2024).
- From cognitive science: practice testing and practice spread out over time are the best-supported study techniques (Dunlosky et al. 2013); homework that mixes problem types improved later problem solving in an introductory physics course, although the students believed they learned less from it (Samani and Pan 2021); and text that states a misconception and shows why it fails changes it more often than ordinary explanation (Tippett 2010).
Research on this course at BYU
Five papers by the author and his colleagues on BYU's first-semester course bear on the book so far. They are small studies in one course, and the papers call their results preliminary.
Learning from mistakes
In three offerings of the course taught by the author (2019–2021; 75 consenting students), students could rework missed problems on each unit exam for half of the points they recovered, and three of those problems reappeared word for word on the final exam, without notice. Students declined the offer on 50 of 99 eligible occasions. Those who reworked a problem did better on it on the final exam, largely regardless of how they had done on the unit exam, while the final-exam scores of those who declined followed their unit-exam scores more closely. The benefit was larger for applying principles correctly than for choosing the right principles in the first place. Because the students chose whether to rework, the comparison is not randomized. Interviews found novice habits such as searching a formula sheet for the right formula and memorizing worked examples whole; the formula searching, on a multipole problem, persisted even in some students who reworked it (Mason and Colton 2022; Mason et al. 2023).
In the book: every worked example begins with a sketch of what is given and builds its result from a principle; Section 8.3.1 says what the multipole expansion means before any algebra, and Section 8.4 tests the dipole approximation in a worked example.
Tutorials built on a problem-solving framework
Since 2023 the course's tutorials have been written around a problem-solving framework developed for introductory physics: draw or interpret a sketch of the situation, describe what is given and what is wanted, name the principles that apply and say why, set up the solution, carry out the mathematics, and check the result with limiting cases or units, followed by three questions about how confident the student is. In Fall 2023 (ten tutorials, 20 to 50 minutes each; 14 of 33 students in the study), students valued the step-by-step structure, but some found it too linear, with too little room to set up a problem their own way; in the previous year's survey, after two pilot tutorials, 6 students preferred more structure and 4 less. Tutorial scores correlated with the rest of the course grade (R = 0.60 for the 9 students who turned in all ten tutorials; Mason et al. 2024).
In Fall 2024, half of a class of 36 used each tutorial with the framework's steps written in and half without, the two halves alternating from one tutorial to the next. On exam problems about the tutorial topics, the two versions did not differ significantly (the framework version scored 4.9% higher on choosing the right concepts, p = 0.36). No student interviewed at the end of the semester had noticed the difference, but shown the two versions side by side, 5 of 10 survey respondents preferred the framework version and 1 the other. Questions about the framework itself, such as what about a situation shows that Laplace's equation applies, left some students unsure what kind of answer was expected. In think-aloud interviews, every student began by drawing the situation and nearly all chose a correct principle, but students checked their answers in only 4 of 12 problems, with little relation between how confident they were and whether they checked (Hecht et al. 2025).
In the book: the six tutorials of the book's first half (Tutorials 3.1, 8.1, 9.1, 11.1, 13.1, and 16.1) are the author's revisions of the tutorials from these studies. Tutorial 3.1's ring of charge, for example, walks through every step of the framework, from a sketch and a list of what is given to a closing question on whether the answer is reasonable, and Tutorials 11.1 and 16.1 end with a “Wrap-up” that asks the same. The worked examples also begin with a sketch, and in Chapter 4 (the pilot chapter) and Chapters 1, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 18, 19, and 20 each one has a check written into its solution, marked “Reality check:”. Each box of warm-up questions closes by asking what the student found hardest to understand.
Which topics students find hardest
In end-of-semester surveys in Fall 2022 and Fall 2023, 17 students rated 29 topics of the course from 1 (struggled a lot) to 5 (little or no difficulty). The lowest ratings went to finding capacitance (2.71); electric quadrupole moments and potentials, finding the vector potential from current densities, and magnetic dipoles (2.76 each); and electric dipoles, boundary conditions for all fields, and force, torque, and potential energy (2.94 each). The highest went to electric field lines (4.35), three-dimensional integrals and derivatives (3.94), and Gauss's law for E and Coulomb's law for charge densities (3.71 each). In written comments, students named the separation vector (the “script-r method”) as especially hard, and the fields D, H, and A and how they are related to the others (“H and M are still a mystery to me”). After the Fall 2023 tutorials, students voiced the least confidence about the concepts of the tutorials on dielectrics and magnetic matter, and about the mathematics of those on Gauss's law and the quadrupole field (Mason et al. 2024). The concept inventory described next also had its lowest scores on the questions about fields in materials (Rundquist et al. 2026).
| Topic (average rating) | What the book does |
|---|---|
| Finding capacitance (2.71) | Section 5.8 explains why the charge on a capacitor is proportional to its voltage and gives the steps: put +Q and −Q on the conductors, find the field between them, integrate it from the negative conductor to the positive one (which avoids the most common sign error), and divide; the charge should cancel, a useful check. Examples 5.10 and 5.11 work the parallel-plate and spherical capacitors, the second with its narrow-gap and isolated-sphere limits. Section 5.9.4 gives a capacitor's energy, and Section 9.7.1 fills capacitors with a dielectric (Examples 9.8 and 9.9). Three warm-up questions on Section 5.8 ask whether a capacitor with twice the charge has twice the capacitance, what the charge Q of a capacitor is when its total charge is zero, and what a thin metal sheet slid into the gap does to the capacitance (new, November 2026). |
| Electric quadrupole moments and potentials (2.76) | Section 8.3.1 says what the multipole expansion means before any algebra, and Section 8.3.2 where it converges, with a figure; Section 8.4 tests the dipole approximation in a worked example; Section 8.5 builds the quadrupole moment tensor, writing two of its components out in full (Section 8.5.1); Tutorial 8.1; warm-up questions on what the component Qxx measures and on the two conventions for the quadrupole moment in use (new, November 2026). |
| Finding the vector potential from current densities (2.76) | Section 12.3 applies the script-r method to A, with worked examples (Section 12.3.1) and tips for Mathematica and Python (Section 12.3.2); Section 12.4 finds A from B by the flux method where the symmetry allows. Planned: a tutorial on the flux method in Chapter 12. |
| Dipoles, magnetic (2.76) and electric (2.94); force, torque, and potential energy (2.94) | Electric dipoles in Sections 8.3.4 (the first two terms of the expansion), 8.4, and 9.1 (torque, energy, and force on a dipole; polar molecules; induced dipoles); magnetic dipoles in Sections 12.6 (the dipole term of A, the field of a dipole, and the field near one) and 13.1 (torque, energy, and force; where atomic moments come from). Warm-up questions in Chapter 10 ask why the field of a current loop falls off as 1/z3 although each piece of the loop gives an inverse-square field, and how the field at the center of a loop differs from the field at the center of an electric dipole (new, November 2026). Chapter 12's ask what plays the role of the total charge in the expansion of A, why a loop's dipole moment depends on its vector area rather than its area, and how fast the vector potential of a figure eight of current falls off (new, November 2026). Forces on dielectrics and on magnetic materials: Sections 9.10.1 and 13.11.2. |
| Boundary conditions for all fields (2.94) | The conditions are derived where each field is introduced: E⊥ (Section 4.4), E∥ (Section 5.2), dielectric interfaces (Section 9.8), B (Section 11.3), and magnetic interfaces (Section 13.9). Section 7.1.1 restates the electric ones for boundary-value problems, and Chapter 14 puts the electric and magnetic conditions side by side, in vacuum and in materials (Sections 14.1.1 and 14.1.2), with a figure of the pillbox and the loop that give them. |
| The separation vector (named in written comments) | See “Building a Coulomb's-law integral” under Mathematics below: the script-r method's step list, a sketch of r, r′, and the separation vector in every such example, Tutorial 3.1, and two warm-up questions in Chapter 3. |
| The fields D, H, and A, and how they are related to P, M, and B (named in written comments; the least confidence in the concepts of the tutorials on dielectrics and magnetic matter) | Sections 9.5–9.6 and 13.5–13.6, including “D is not determined by the free charge alone” (Section 9.6.1) and its magnetic counterpart (Section 13.6.2); Tutorial 13.1 compares B and H in a bar magnet, and Tutorial 9.1 relates polarization to polarizability; Chapter 14 sets the electric and magnetic equations side by side, in vacuum and in materials (Sections 14.1.1 and 14.1.2); the triangle diagram of J, B, and A (Section 12.8); warm-up questions in Chapter 9 on what D is, on bound charge, and on the boundary conditions at dielectric interfaces, and in Chapter 13 on free and bound current, on the poles of a magnet as sources of H, on the minus sign in the definition of H, and on which of B and H is continuous at the surface of a magnetized slab (new, November 2026). |
A concept inventory: E-MICE
Existing tests of upper-division electricity and magnetism cover only part of the subject, are free-response and slow to score (CU Boulder's electrostatics and electrodynamics tests), or were written for engineering students. The Electro-Magnetism Inventory for Conceptual Evaluation (E-MICE) is a multiple-choice test of the concepts of the whole first semester. Its wrong answers come from documented student difficulties, from students' written answers to open-ended conceptual questions in BYU's course, and from the experience of four faculty members and of student researchers. About 60 candidate questions were cut to 33, and that version was given to all 20 students of one section at the end of Fall 2025. The mean score was 72.4% (individual scores ranged from 39% to 97%), and 20 of the 33 questions separated stronger from weaker students well (a Kelley discrimination index of at least 0.40). The three questions on materials (P, D, M, and H) had the lowest average, 57% correct (Rundquist et al. 2026). A revised version of 35 questions, which takes in reviews by professors at other universities, has been prepared, and the authors plan to submit it to PhysPort.
In the book: so that E-MICE can be used with students who learned from this book, the book keeps its own questions different from the inventory's items.
Student difficulties by topic
Each row gives a finding, its source, and what the book does. Section, example, and exercise numbers refer to the current version of the book. “Planned” marks a change that has been decided on but is not yet in the posted version.
Mathematics
| Finding | Source | What the book does |
|---|---|---|
| Divergence and curl. Students calculate divergence and curl well but read them from pictures poorly: divergence as field lines spreading apart, curl as lines bending. Many also read the differential laws as non-local, as if a charge anywhere made the divergence of E nonzero everywhere. Worksheets that tie pictures of fields to the calculations improved students' understanding. | Bollen et al. 2015, 2016, 2018; Baily et al. 2015; Stuckey and Quan 2024 | Section 1.4.2 says that the divergence is not nonzero wherever the arrows spread apart, and works through the field of a point charge, whose arrows spread but whose divergence is zero away from the charge. Section 1.4.3 does the same for curl and curving lines, and Section 1.4.4 explains locality. Figures of six fields; animations of a shrinking box and of paddle wheels. Exercises P1.12, P1.13, and P1.20 ask for a sketch and a prediction (from the flux out of a small box, or from a paddle wheel) before the calculation (Exercises P1.12 and P1.20 new, November 2026). |
| The divergence theorem. Students translate the theorem into words more easily than they interpret it physically. | Huffman et al. 2020 | Section 1.5.3 explains the theorem twice in physical terms: as sources inside a volume and the flow out through its surface, and as small cubes whose shared faces cancel, with a figure of those cubes (new, November 2026). |
| The Dirac delta function. Students struggle to recognize when a delta function is the right tool, to write a described charge distribution with delta functions, to see that a delta function has units, and to integrate delta functions in three dimensions and in curvilinear coordinates. | Wilcox and Pollock 2015a | Section 1.6 gives the units of the delta function and the densities of a point, a line, and a sheet of charge, with a figure; Exercise P1.18 asks for densities in curvilinear coordinates. Section 1.6.2 also writes the three-dimensional delta function in Cartesian, spherical, and cylindrical coordinates, shows each one-dimensional delta function as a surface and their product as the point where three surfaces meet, and integrates the density of a cylindrical shell of charge back to its total charge, with a figure (new, November 2026). |
| Building a Coulomb's-law integral. Of 166 exam solutions that set up the integral, 42% had trouble with the charge element dq, most often by integrating over a region with no charge, by using an element with the wrong units, or by using the total charge in place of the density. Of 148 that wrote the distance from the source to the field point, 47% got it wrong. Two thirds of these students drew the position vectors, but only half of those used the drawing in their solution. Only 3 of 8 students interviewed drew the source and field-point position vectors on their own; the authors suggest that the usual shorthand for the separation vector hides them. At BYU, students named the separation vector (the “script-r method”) in end-of-semester surveys as one of the hardest parts of the course. In upper-division courses at the University of Washington, only about half of the students set limits on a charge integral that described where the charge was (some described its width but not its location), and only about 10% wrote the length of a small step around the origin as r dθ; 55% wrote dθ alone. | Wilcox et al. 2013a, 2013b; Mason et al. 2024; Xue 2021 | The “script-r method” (Sections 3.5, 5.5, 10.6, and 12.3): draw r (origin to field point), r′ (origin to a piece of the source), and the separation vector between them; write each on a line of its own, building the separation vector as r − r′, then its magnitude; then write the element and the density in the source variables; then check the direction, units, and limits of the result. The step lists also say what the element is: the size of one piece of the source, so that λ dℓ′ is the charge on that piece, with a reminder that an angle is not a length (a step dθ′ moves the source point a distance r′ dθ′). The step for the density says to use the density, not the total charge, and to set the limits by integrating over the source, counting each piece of it once (new, November 2026). Every worked example of this kind is solved this way, beginning with a sketch of the three vectors. Two of Chapter 3's warm-up questions have students build the separation vector themselves and catch two common slips: subtracting the unit vectors of the two points, and drawing the piece of charge at a special point instead of an arbitrary one (new, November 2026). Two of Chapter 10's catch slips particular to the Biot–Savart law: a cross product of unit vectors taken at two different points, and a reversed current counted twice, once in the line element and once in the limits (new, November 2026). Two of Chapter 12's carry the method over to the vector potential: which precaution still applies without the cross product, and what a semicircle of current contributes at its center, where every piece is the same distance away (new, November 2026). Tutorial 3.1 has students draw the three vectors for an arbitrary piece of a ring of charge and write each one out before integrating. |
| Current densities. Expressing the current density of a given situation is a new difficulty in magnetostatics. | Wilcox and Corsiglia 2019 | Section 10.4 (current from a nonuniform current density, with figures). In the step list of Section 10.6, the step for the current density includes a check that its integral over a cross section is the total current (new, November 2026). |
| Separation of variables. Students struggle to justify the product form of the solution and to see why an infinite sum is needed; they do not think of the method unprompted, miss implicit boundary conditions (such as a potential that stays finite at the origin), and rarely reflect on the solution. | Wilcox and Pollock 2015b; Pollock and Wilcox 2015 | Section 7.4 justifies the method by uniqueness, argues from the boundary conditions which variable gets sines and which gets exponentials, lists the implicit conditions (“V is finite at r = 0”) along with the others, and checks its answers. One of Chapter 6's warm-up questions asks why no finite number of sine terms can add up to a square wave on its flat part, and Chapter 7's ask how far the potential of Example 7.1 reaches between its plates, why a product solution cannot decay exponentially in both directions, and what the condition far away decides (new, November 2026). Planned: tutorials on Fourier's trick (Chapter 6) and on separation of variables (Chapter 7). |
| Complex exponentials. Faculty name complex exponentials, and getting measurable quantities from them, among the main difficulties of upper-division electrodynamics; students handling waves at a boundary make errors with them. | Baily et al. 2014; Ryan et al. 2018 | Chapter 18, a mathematics review chapter on complex numbers before the chapters on waves, with both sign conventions for the phase and time averages of products; every complex quantity carries a tilde. Chapter 18's warm-up questions ask, for example, what multiplying by 1 + i does to a point, which way the wave Re[ei(3z+4t)] travels, and what must be known about a phase before saying which of two oscillations peaks first (new, November 2026). |
| Choosing a method. See “The research in brief” above. | Wilcox et al. 2013b; Wilcox and Corsiglia 2019 | Each method has its own section, “Solving for ... when given ...”, with a numbered list of steps; Section 5.5.2 says when to find the potential through the field and when to integrate for it directly; triangle diagrams (Sections 5.11, 12.8, and 24.7) show how sources, fields, and potentials are related. “Which method?” flowcharts are in Chapter 14, which reviews electrostatics and magnetostatics together (electrostatics, Figure 14.9; magnetostatics, Figure 14.10), and close Chapter 16 (fields that change in time, Figure 16.17) (new, November 2026). Planned: questions that ask only how a problem would be solved (see Assessment formats below). |
| Checking a result. Only 8% of 172 exam solutions included an explicit check, and half of the students interviewed made no attempt to check their answers. In think-aloud interviews at BYU, students checked their answers in only 4 of 12 problems, with little relation between how confident they were and whether they checked. | Wilcox et al. 2013b; Hecht et al. 2025 | The step lists end with a check of direction, units, and limits, and the examples check their results (limiting cases; reflection plus transmission equal to one). In Chapter 4, the pilot chapter, and in Chapters 1, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 18, and 19, each worked example now has its check written into the solution, on a line of its own marked “Reality check:” (new, November 2026). Tutorial 3.1 ends with a step on whether the answer is reasonable, and Tutorials 11.1 and 16.1 with a “Wrap-up” that asks the same. Three of Chapter 5's warm-up questions and one of Chapter 6's practice checking: three ask for reasons why a classmate's result cannot be right, without redoing the calculation, and one asks what to conclude from a computed potential that does not satisfy Laplace's equation (new, November 2026). Planned: “is this answer reasonable?” questions. |
Physics
| Finding | Source | What the book does |
|---|---|---|
| Gauss's law and symmetry. Even strong juniors struggle to build a complete symmetry argument, to go from a field to the charge that produces it, and to recognize when the symmetry is not enough for Gauss's law to give the field. Introductory and engineering students often reason from surface features instead. A symmetry argument answers two separate questions: which components of the field are zero, and which coordinates it can depend on. Students who mix the two choose surfaces less well: on one University of Washington exam about a uniformly charged sphere, the students whose argument about what the field depends on mixed in its direction judged correctly whether a proposed Gaussian surface would work 11% of the time, against 56% for those who kept the two apart (93 students). | Pepper et al. 2010; Singh 2006; Campos et al. 2023; Xue 2021 | Example 4.1 goes from a field to the charge that produces it, and Section 4.2.3 says that the two forms of Gauss's law answer opposite questions; Section 4.3.2: the three kinds of symmetry, each stated as two separate answers, which components of the field are zero and which coordinates it can depend on, with the kinds of reason that settle each and both arguments made for a sphere (the two answers kept apart, new, November 2026), and a charge density that is not symmetric enough, with a pointer to a figure of such a field; Example 4.3 makes the symmetry argument component by component, with a figure (new, November 2026); Exercise P4.9 asks students to sketch the field lines near a cube, a finite line, and a disk, and to use the sketch to explain why Gauss's law cannot give the field (the sketch step new, November 2026); Colton's Second Commandment: always draw a Gaussian surface. Planned: a tutorial on making the symmetry argument. |
| Zero enclosed current is not zero field. Students take a zero enclosed current to mean a zero magnetic field. In interviews, students who wrote the circulation as a sum over the sides of their loop avoided the mistake, and others dropped it when reminded that the integral is a sum. After a tutorial on Ampère's law, the most common wrong answer, about which loops are useful, chose the loops that enclose current. | Wallace and Chasteen 2010 | Section 4.2.3 says that a zero flux does not mean a zero field. Section 11.1.3 gives the same caution for the circulation of B, with a figure of B along a loop beside a wire, where B·dℓ is positive on one part of the loop and negative on the other (new, November 2026). Exercise P11.2(d) asks for a sketch marking where B·dℓ is positive and where negative on a square that encloses no current (new, November 2026), and Exercise P11.13(c) asks it as a true/false question. |
| Ampère's law asks for many skills at once: finding the field's direction and what it depends on, choosing a loop, taking B out of the integral, curvilinear coordinates, and current densities, which no single earlier course teaches together. On an exam about the magnetic field of a slab of current, students mixed the two parts of the symmetry argument in 28% of their answers on what the field depends on, and in 11% of their answers on which of its components are zero (131 students). | Manogue et al. 2006; Wallace and Chasteen 2010; Xue 2021 | Section 11.2: the conditions for taking B out of the integral, a step list, and the four standard geometries, each stated, as in Chapter 4, as two separate answers, which components of the field are zero and which coordinates it can depend on, and argued that way in Examples 11.1–11.4 (new, November 2026); Example 11.1 argues away each unwanted component of the field; Colton's Third Commandment: always draw an Amperian loop; Tutorial 11.1, whose Part 2(b) asks the same two questions separately (new, November 2026). Warm-up questions ask what a circular loop tells you when the field is not parallel to it, why the closed-cylinder argument that removes the radial part of B cannot remove that of E, and what a path that winds twice around a wire encloses (new, November 2026). |
| Current in a wire that narrows. 51.2% of 284 students said that the divergence of the current density is not zero where a wire with a steady current narrows, because the current density grows there or its lines crowd together. | Baily et al. 2017 | Section 10.4.1 explains why the current density grows where a wire narrows while its divergence stays zero, with a computed figure of the lines of J through a narrowing strip and a small box that as many lines leave as enter (new, November 2026). |
| Conservation laws and energy flow. Faculty count the physical meaning of the conservation laws among the main goals of upper-division electrodynamics. On a conceptual test, students in several courses scored consistently lower on the questions about Faraday's law, the continuity equation, and field energy than on the others. | Baily et al. 2014, 2017 | Section 20.1 states the continuity equation in words; Section 20.2.3 shows energy flowing into a wire from the field around it. |
| Boundary conditions in electrodynamics. Students do not think of using boundary conditions outside wave problems (for example, to argue that the electric field just outside a wire carrying a steady current is not zero). With waves at a boundary, they drop the exponentials from the fields, get the sign of the reflected wave vector wrong, and mishandle k·r; at a boundary at z = d, they cancel the exponentials as if the boundary were at z = 0. | Ryan et al. 2015, 2018 | Section 20.2.3 finds the field just outside a current-carrying wire from the continuity of its parallel component; Sections 21.6 and 21.7 write all three waves with their exponentials and match them “at every point of the boundary and at every instant”; Section 21.4 explains k·r. Exercise P21.12 puts the boundary at z = d and begins with a sketch of the three waves (new, November 2026). Warm-up questions in Chapter 9 use the boundary conditions to find the field in thin cavities cut in a dielectric, and to explain why field lines just outside a dielectric with a very large dielectric constant meet it at right angles (new, November 2026). |
| Pictures of plane waves. Shown the standard picture of a plane wave (sinusoidal E and B along one axis), only about 10% of some 1400 introductory students ranked the fields at four points of a plane perpendicular to the axis as equal; about 60% said that the field is zero at a point outside the curves. | Ambrose et al. 1999 | Section 21.4 says that the phase is the same everywhere on a plane perpendicular to the direction of travel, with a figure of the wave fronts as planes. The caption of the standard picture (Figure 21.12) says that its curves show the fields only at points on the axis, and that at each instant the fields are the same at every point of a plane perpendicular to it (new, November 2026). |
| The Ampère–Maxwell law. Students find it hard to see when a displacement current appears, think that the law can be used only with a symmetric loop, and some think that the changing electric field must cross the whole surface bounded by the loop. Many read the law as cause and effect. | Suárez et al. 2024, 2025 | Section 16.1.1 uses two surfaces bounded by the same loop at a charging capacitor; Section 16.3 explains how the displacement current does, and does not, produce magnetic fields; Tutorial 16.1. Example 16.1 has a sentence and a figure on a loop larger than the capacitor's plates, whose flat surface the changing field crosses only in part (new, November 2026). |
| Faraday's law and EMF. Motional EMF with extended conductors or moving contacts is troublesome; teaching EMF as one concept across batteries, transient currents, and induction helps. | Zuza et al. 2012, 2014, 2020 | Section 15.1.1 defines EMF as the circulation of the force per charge, for any force; Section 15.1.2 finds motional EMF from the magnetic force on the charges; the flux rule for moving circuits (Section 15.1.3) is kept distinct from Faraday's law for changing fields (Section 15.2); Exercises P15.3 and P15.9 (the Faraday disk and Feynman's disk). |
| Simultaneity in relativity. After instruction, students at all levels have serious difficulty with the relativity of simultaneity and the role of observers; the beliefs are deeply held. | Scherr et al. 2001, 2002; Alstein et al. 2021 | Section 25.2 builds each observer's frame from a lattice of rulers with a synchronized clock at every intersection; Section 25.3.1 treats simultaneity with a figure; spacetime diagrams throughout; Exercise P25.32 includes a true/false statement on simultaneity. |
| Introductory concepts. On every question of a standard introductory test that fewer than half of introductory students answered correctly after instruction, fewer than two thirds of upper-level students did. | Maries et al. 2022 | True/false questions in 20 chapters; a first multiple-choice question, which also asks students to explain their choice, at the end of Chapter 7 (Exercise P7.6; new, November 2026). Planned: multiple-choice conceptual questions in nearly every chapter. |
A check against CU Boulder's lists of student difficulties
The University of Colorado Boulder's materials for its junior-level electrostatics and magnetostatics course include short lists of the difficulties their instructors saw, topic by topic. These are course materials rather than research papers, but they make a useful checklist. We grouped their entries into 13 themes and looked for each one in the book.
All 13 are addressed in the book. Eleven are addressed where the topic is taught. The other two are about choosing a method, which the research above finds to be the most persistent difficulty; the book addresses them in one paragraph, one exercise, and the electrostatics flowchart at the end of Chapter 9 (new, November 2026), and the planned method questions will give students practice with them.
| Difficulty in CU's lists | Where the book addresses it |
|---|---|
| Area elements written with the wrong units, and never checked | Section 1.3.3 builds each area element from the lengths of its edges and checks one against the area of a sphere. |
| δ(x − 2) read as a point charge; products of delta functions and their dimensions | Section 1.6.2 (a line of charge λδ(x)δ(y) and a sheet σδ(z), with a figure); Section 1.6.2 also says that a single delta function such as δ(x − 2) is a whole plane, not a point, and shows the delta functions of curvilinear coordinates as surfaces, with a figure (new, November 2026); Exercise P1.18. |
| A smeared-out surface charge treated like a point charge (an infinite field near a disk) | Section 3.5.2: a uniformly charged disk has a finite field close to its center. |
| Zero flux read as zero field; factors of two with pillboxes | Section 4.2.3, with a figure of a box beside a point charge; Sections 4.3.4 and 4.5; a warm-up question in Section 9.7 on two common mistakes with a pillbox at a capacitor plate, one of them a factor of two (new, November 2026). |
| Choosing a reference point for the potential; a zero circulation of E not connected with the potential | Section 5.3.2 (the reference point, with the infinite line of charge); Section 5.3.1 (four equivalent statements of a conservative field). |
| Laplace's equation not seen as Poisson's equation with no charge; images and uniqueness; where an image solution holds | Section 7.1; Section 7.2.1 (“The image charge does not exist below the plane; it is only a device for finding the potential above it.”); warm-up questions on a second solution between two grounded plates that only the condition far away rules out, and on the work needed to pull a charge away from a grounded plane, which is half the work needed to separate the charge from its image (new, November 2026) |
| The multipole expansion thought valid only far away, or only when the monopole term vanishes | Section 8.3.2: the expansion is exact wherever the field point is farther from the origin than any of the charge; Section 8.3.3: far away, the first term that does not vanish is enough. The figure of the expansion's geometry shows the sphere outside which the series converges (new, November 2026). A warm-up question asks why the series needs the field point to be farther from the origin than the charge (new, November 2026). |
| What D is, and when the free charge alone determines it | Section 9.6.1, “D is not determined by the free charge alone.” Warm-up questions ask what D is, and what E and D are in a slab polarized parallel to its faces, where D is not zero although there is no free charge (new, November 2026). |
| Current, volume current density, and surface current density confused | Section 10.4, with a nonuniform current density; a warm-up question on a current squeezed into a thin layer at the surface of a wire (new, November 2026). |
| Boundary conditions on B not derived from a pillbox and a loop | Section 11.3 (with a figure of the pillbox and the loop) and Section 13.9; warm-up questions on why the jump in B must be parallel to the surface and perpendicular to the current, and on what the two conditions allow (new, November 2026). |
| No electric field outside a solenoid whose current changes, because B is zero there | Section 15.3.1: outside the solenoid there is no magnetic field at all, but there is an electric field (with a plot). |
| Finding the potential by first finding the field, when integrating for the potential directly is easier | Section 5.5.2, “Choosing a route”; the electrostatics flowchart (Figure 14.9; new, November 2026); a warm-up question in Chapter 5 that asks which route to take for three charge distributions (new, November 2026). Planned: method questions. |
| Gauss's law used where the symmetry is not enough and Coulomb's law is needed | Section 4.3.2; Exercise P4.9 (including the uniformly charged disk); Section 5.5.2; the electrostatics flowchart (Figure 14.9; new, November 2026); the same warm-up question in Chapter 5, two of whose three distributions lack the symmetry Gauss's law needs (new, November 2026). Planned: method questions. |
How people learn: findings from cognitive science
| Finding | Source | In the book |
|---|---|---|
| Practice testing. Of ten common study techniques, practice testing and practice spread out over time have the most support. Being tested beats restudying on tests days later, even though restudying wins on an immediate test and feels more effective. | Dunlosky et al. 2013; Roediger and Karpicke 2006 | True/false questions in 20 chapters; warm-up questions in Chapters 1–13 and 15–20, each box ending with a request for a summary written without looking back (new, November 2026). Planned: an end-of-chapter review in nearly every chapter, with conceptual questions and problems from earlier chapters mixed in. |
| Interleaved practice. In two lecture sections of introductory physics, homework that mixed topics, instead of practicing one topic at a time, raised median scores on two later surprise tests of new and harder problems by 50% and 125%. The students rated the mixed homework as harder and believed, wrongly, that they had learned less from it. | Samani and Pan 2021 | Planned: in the end-of-chapter review, problems in random order whose section does not say which method to use, with a short note telling students why the set is there. |
| Pretests. Answering questions before studying, even unsuccessfully, improves later learning of the tested content. | Richland et al. 2009 | Under consideration: a short pretest for each tutorial, as in the tutorials of the University of Washington and CU Boulder. |
| Worked examples. Worked examples help most early in learning; leaving more and more steps to the learner eases the move to solving problems alone; and support that helps beginners can hinder more experienced learners. | Renkl and Atkinson 2003; Kalyuga et al. 2003 | 135 fully worked examples. The tutorials and the multi-part exercises lead from explicit steps to open ones. |
| Self-explanation. Good students explain worked examples to themselves and tie each step to a principle; weaker students do not. In an eye-tracking study, students spent about 40% of their time on the words of worked examples yet recalled little of their conceptual content. | Chi et al. 1989; Smith et al. 2010 | Key equations are followed by what they mean (“With words, ...”), and the examples say why each step is taken. In Chapter 4, the pilot chapter, and in Chapters 1, 3, 5, 6, 7, 9, 10, 11, 12, 13, 15, 16, 18, 19, and 20, the worked examples ask short “why?” questions at key steps, each a sentence of its own, with the answers in footnotes, so that students explain the step to themselves before they look (new, November 2026). |
| Refutation text. Text that states a misconception and shows why it fails changes the misconception more often than ordinary explanation. | Tippett 2010 | Cautions in the margins and in the text that name a tempting idea and show why it fails: divergence is not where arrows spread apart; a zero flux is not a zero field; a zero circulation is not a zero field; the current density can grow without having a divergence. |
| Words beside pictures. Words placed next to the parts of a figure they describe help learning, mainly for complex material. | Ginns 2006, as summarized by Beege et al. 2019 | Nearly all figures are in the margin, on the page that discusses them; the start sketches of the examples sit beside their explanations; notes on equations sit beside the equations. |
| Seductive details. Interesting but irrelevant additions lower learning by a small to moderate amount, mainly when learners take them to be part of what they must learn. | Kienitz et al. 2023 | The portraits and short biographies of about 80 scientists are set off in the margin, visibly separate from the physics. |
Tutorials and course design
At CU Boulder, a junior-level electricity and magnetism course redesigned around faculty learning goals, concept questions with peer discussion, and tutorials produced larger conceptual gains than traditional courses (Chasteen et al. 2012); concept questions with peer instruction work in upper-division courses (Pollock et al. 2010). On a conceptual question about the magnetic field of a cylinder of current, students in such courses scored between 36% and 69%, and students in traditionally taught courses 28% (Wallace and Chasteen 2010). Attendance at the tutorials predicted students' conceptual scores, though not their exam scores, after accounting for their earlier grades (Chasteen 2012). In writing tutorials for advanced students, CU found that questions about signs, and about whether a quantity is zero, work better than long calculations; that the scaffolding should start explicit and become more open; that a page takes about ten minutes; and that a challenge step keeps faster groups busy (Baily et al. 2014).
The book has 15 tutorials inside its chapters, most of them written from the author's course handouts. The six in the first half (Tutorials 3.1, 8.1, 9.1, 11.1, 13.1, and 16.1) are the author's revisions of the tutorials studied in his course, built on the steps of a problem-solving framework (see “Research on this course at BYU” above). Each has a printable version for class and an answer key in the instructor's manual. The goal is about one tutorial per chapter, and eight more are planned (Chapters 4, 5, 6, 7, 12, 15, 24, and 27), designed with CU's lessons and the BYU findings in mind. Whether each tutorial gets a pretest will be decided when the first new one is written.
Warm-up questions (new, November 2026). In Just-in-Time Teaching, students answer short “WarmUp” questions on the reading a few hours before class, for credit on effort rather than on correctness, and the instructor reads the answers to shape the class (Novak et al. 1999). In the author's course, each class began with a short conceptual quiz, not for credit, that reviewed earlier topics and introduced the day's (Mason et al. 2023). The book puts a box of two or three such questions (more for a long section) at the head of each section's exercises: a concept check with a reason, an idea put in the student's own words, a classmate's claim to judge, or the effect of changing one assumption. Each box ends by asking for a short summary of the section written without looking back, a form of practice testing (Roediger and Karpicke 2006), and for what the student found hardest to understand. Chapters 1–13 and 15–19 have them now (257 questions in 108 boxes); the other chapters will follow, all but the short Chapter 14, which sets the equations of electrostatics and magnetostatics side by side.
Visualization
In a course that consistently emphasized free-body diagrams, most students drew them on exams without being asked, and students who drew them correctly were much more likely to solve the problem; the strongest students also used their diagrams to check their work (Rosengrant et al. 2009). In upper-division electrostatics, two thirds of 148 students drew the position vectors on an exam, but only half of those used the drawing in their solution (Wilcox et al. 2013b). In think-aloud interviews at BYU, every student began by drawing the situation (Hecht et al. 2025). A meta-analysis of 26 studies found a medium-sized advantage of animations over static pictures, larger when the animation shows the change that is to be learned (Höffler and Leutner 2007). Students misread standard pictures of fields and waves (Ambrose et al. 1999; Bollen et al. 2017; Campos et al. 2019).
- Start sketches. Every worked example but one begins with a hand-drawn sketch of what is given and what is asked, beside a sentence explaining that drawing it is the first step. In problems that use the separation vector, the sketch shows r, r′, and the separation vector, and the solution then uses them line by line. The keys to the tutorials begin the same way.
- Figures beside the text. About 670 figures, nearly all in the margin on the page that discusses them. Most exercises with a geometry of their own have a figure of what is given.
- Figures computed from the physics. Field lines are traced numerically, and some figures are computed exactly (for example, the lines of current density through a narrowing strip), so that what the picture shows is what the equations say.
- Animations. 27 animations, each linked from the figure it extends.
- Sketch, predict, compute. Exercises that ask for a sketch and a prediction before a calculation (Exercises P1.12, P1.13, and P1.20), or for a sketch that explains a result (Exercises P4.9, P11.2, P11.9, and P21.12).
Assessment formats
- Now in the book: true/false questions (“True or false? Explain your reasoning.”) at the end of 20 chapters, a first multiple-choice question that asks for the reason behind the choice (Exercise P7.6; new, November 2026), and an AI programming exercise at the end of nearly every chapter, in which students use AI tools to explore a concept computationally.
- Now in the book: the new front-matter section “Suggestions for Instructors” (new, November 2026) recommends letting students earn back part of the lost credit by turning in corrected exam solutions, with the keys released only after the corrections are in (Brown et al. 2016).
- Planned: an end-of-chapter review in nearly every chapter, after a pilot in Chapter 4: (a) true/false and multiple-choice conceptual questions (Dunlosky et al. 2013; Roediger and Karpicke 2006; Maries et al. 2022; a coupled multiple-response format has been validated at this level, Wilcox and Pollock 2014); (b) “how would you solve this problem?” questions, in which students name the method, the section, and the reason, without solving; (c) problems in random order, including some from earlier chapters, that students must first classify and then solve (Samani and Pan 2021); and (d) the AI programming exercise. In Chapter 4, the true/false exercise now sits at the end of the chapter, as part (a) of its review (new, November 2026).
- Why (b): the author has used such questions for exam review and on exams. In an introductory course in which the instructors modeled written strategies (the principle to use, why it applies, and how to apply it), students chose the right principle on five end-of-course questions 70% of the time, against 48% in a later class of the same course taught without them; this was not a randomized comparison (Leonard, Dufresne, and Mestre 1996).
- “Which method?” flowcharts for electrostatics, magnetostatics, and time-dependent fields: questions down one side, where “yes” leads to a method (with its section) and “no” to the next question. The charts for electrostatics and magnetostatics are in Chapter 14, which reviews the two subjects together (Figures 14.9 and 14.10), and the chart for time-dependent fields closes Chapter 16 (Figure 16.17; new, November 2026). The text that introduces the two charts, in Section 14.3, says that students often have trouble identifying which method to use when a problem does not announce it, and cites the research (Wilcox et al. 2013b; new, November 2026).
- A concept inventory for the first semester, E-MICE, is being developed by the author's group (see “Research on this course at BYU” above). The book keeps its own questions different from its items.
Changes made because of this research
| Date | Change | Research |
|---|---|---|
| November 2026 | Section 10.4.1: why the current density grows where a wire narrows while its divergence stays zero, with a computed figure. | Baily et al. 2017 |
| November 2026 | Section 11.1.3: a zero enclosed current means a zero circulation, not a zero field, with a figure of B along a loop beside a wire. | Wallace and Chasteen 2010 |
| November 2026 | Exercises P1.12 and P1.20: sketch a field, predict its divergence from the flux out of a small box, then calculate. | Baily et al. 2015; Bollen et al. 2016, 2018; Stuckey and Quan 2024 |
| November 2026 | Example 4.3: the symmetry argument for a line of charge, component by component, with a figure. | Pepper et al. 2010; Singh 2006; Campos et al. 2023 |
| November 2026 | Sections 3.5 and 5.5: the step for the charge density says to use the density, not the total charge, and to integrate only over the region that holds the charge. | Wilcox et al. 2013a, 2013b |
| November 2026 | Section 10.6: the step for the current density includes a check that its integral over a cross section is the total current. | Wilcox and Corsiglia 2019 |
| November 2026 | Sketch steps in Exercises P4.9 (field lines where the symmetry is not enough), P11.2 (where B·dℓ is positive and negative on a loop that encloses no current), and P11.9; Sections 4.2.3 and 4.3.2 point to figures of a zero-flux box in a strong field and of a field that is not radial. | Pepper et al. 2010; Campos et al. 2023; Wallace and Chasteen 2010 |
| November 2026 | Exercise P21.12: reflection and transmission at a boundary at z = d, beginning with a sketch of the three waves. | Ryan et al. 2015, 2018 |
| November 2026 | Figure 21.12's caption: the curves of the standard plane-wave picture show the fields only along the axis. | Ambrose et al. 1999 |
| November 2026 | The three-dimensional delta function in three coordinate systems, with a figure (Section 1.6.2); a figure of the divergence theorem's small cubes (Section 1.5.3); the sphere of convergence in the multipole expansion's figure (Section 8.3.2); a loop larger than a capacitor's plates in Example 16.1, with a figure. | Wilcox and Pollock 2015a; Huffman et al. 2020; CU Boulder's course materials; Suárez et al. 2025 |
| November 2026 | A “which method?” flowchart for electrostatics at the end of Chapter 9 (now Figure 14.9, in Chapter 14): from the symmetry of the charges to Gauss's law, boundary-value methods, the multipole expansion, the script-r method, and the numerical methods, each with its section. | Wilcox et al. 2013b; Wilcox and Corsiglia 2019 |
| November 2026 | Chapter 4's worked examples (the pilot): a “Reality check:” line in each solution, and short “why?” questions at key steps, answered in footnotes. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Chapter 4's true/false exercise moved to the end of the chapter, as part (a) of its end-of-chapter review. | Dunlosky et al. 2013; Roediger and Karpicke 2006 |
| November 2026 | Warm-up questions at the head of each section's exercises in Chapter 4 (twelve questions in five boxes), each box ending with a request for a summary written without looking back and for the hardest point. | Novak et al. 1999; Roediger and Karpicke 2006 |
| November 2026 | Warm-up questions in Chapters 1 and 2, in the same form (26 questions in eleven boxes). | Novak et al. 1999; Roediger and Karpicke 2006 |
| November 2026 | “Which method?” flowcharts for magnetostatics at the end of Chapter 13 (now Figure 14.10, in Chapter 14) and for fields that change in time at the end of Chapter 16 (Figure 16.17), in the form of the electrostatics chart: from the symmetry of the currents to Ampère's law, boundary-value problems with magnetic materials, the multipole expansion, the Biot–Savart law, and the numerical methods; and from the check that the fields change slowly to the flux rule, the Faraday-current and displacement-current methods, and the fields from the currents at each instant. | Wilcox et al. 2013b; Wilcox and Corsiglia 2019 |
| November 2026 | Warm-up questions in Chapter 3, in the same form (13 questions in six boxes), two of them on building the separation vector. | Novak et al. 1999; Roediger and Karpicke 2006; Wilcox et al. 2013a, 2013b |
| November 2026 | The book's conceptual questions checked against the items of the E-MICE concept inventory and kept distinct from them, so that the inventory can be used with students who learned from this book. | Rundquist et al. 2026 |
| November 2026 | Warm-up questions in Chapter 5, in the same form (24 questions in ten boxes): three on capacitance, one on choosing a route to the potential, and three on checking a result. | Novak et al. 1999; Roediger and Karpicke 2006; Mason et al. 2024; Wilcox et al. 2013b; Hecht et al. 2025 |
| November 2026 | Warm-up questions in Chapter 6, in the same form (eight questions in three boxes): one on why a Fourier series needs infinitely many terms to reproduce a jump, and one on checking a result. | Novak et al. 1999; Roediger and Karpicke 2006; Wilcox and Pollock 2015b; Wilcox et al. 2013b; Hecht et al. 2025 |
| November 2026 | Warm-up questions in Chapter 7, in the same form (13 questions in four boxes, six of them for the long section on separation of variables): two on what the condition far away decides, one on the energy of the image problem, two on how the relaxation method converges and where its grid is least accurate, and one on reading a series solution for how far the potential reaches. | Novak et al. 1999; Roediger and Karpicke 2006; Wilcox and Pollock 2015b; Wilcox et al. 2013b |
| November 2026 | Warm-up questions in Chapter 8, in the same form (ten questions in five boxes): one on why a single number need not be a scalar, one on where the multipole expansion converges, one on the dipole moment of a single charge placed off the origin, one on a potential that is the same at a point while the field there is not, and two on the quadrupole moment. In Chapter 7, the first multiple-choice question in an end-of-chapter review, made from a warm-up question that was too long for its box. | Novak et al. 1999; Roediger and Karpicke 2006; Mason et al. 2024; Maries et al. 2022 |
| November 2026 | Warm-up questions in Chapter 9, in the same form (23 questions in ten boxes): two on what D is and when it is not zero although there is no free charge, one on whether bound charge is real, one on two common mistakes with a pillbox, and three on the boundary conditions at dielectric interfaces. | Novak et al. 1999; Roediger and Karpicke 2006; Mason et al. 2024; Ryan et al. 2015; Rundquist et al. 2026; CU Boulder's course materials |
| November 2026 | Warm-up questions in Chapters 10 and 11, in the same form (25 questions in ten boxes): two on slips in setting up the Biot–Savart integral, three on reading Ampère's law, one on surface current density, two on the boundary conditions on B, two on magnetic dipoles, and one on testing a program against a case whose answer is known. | Novak et al. 1999; Roediger and Karpicke 2006; Wilcox et al. 2013a, 2013b; Manogue et al. 2006; Wallace and Chasteen 2010; Mason et al. 2024; CU Boulder's course materials |
| November 2026 | Warm-up questions in Chapter 12, in the same form (18 questions in seven boxes): two on the script-r method for the vector potential, one on a vector potential that is zero where the field is not, and three on magnetic dipoles. | Novak et al. 1999; Roediger and Karpicke 2006; Mason et al. 2024 |
| November 2026 | Chapter 12's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the ten solutions, and short “why?” questions, answered in footnotes, in seven of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Warm-up questions in Chapter 13, in the same form (24 questions in eleven boxes), eight of them on the field H and how it differs from B: among them the poles of a magnet as sources of H, the sign in its definition, why a material can change it, and which of the two is continuous at a surface. | Novak et al. 1999; Roediger and Karpicke 2006; Mason et al. 2024; Rundquist et al. 2026 |
| November 2026 | The hardest warm-up questions of Chapters 3–13, reviewed one by one by the author: hints added to five, easier versions of three, three rewritten at his direction, one split in two, and one moved to the exercises (Exercise P11.11, on why the field of a lone segment of current does not obey Ampère's law), with an easier question in its place. Example 12.7 gained a “why?” question on why thin wires at the centers of two bus bars give too large a field, and Section 12.5 a short qualitative example of the reason, with a figure. | Novak et al. 1999; Chi et al. 1989 |
| November 2026 | Chapter 13's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the thirteen solutions, and short “why?” questions, answered in footnotes, in eight of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Warm-up questions in Chapter 15, in the same form (13 questions in five boxes): among them what an open circuit does to the EMF of a sliding bar, the sign bookkeeping of the flux rule done the other way around, which way a galvanometer jumps when the current is reversed or doubled, and which part of the electric field drives the current in a circuit that moves through a static field. | Novak et al. 1999; Roediger and Karpicke 2006; Rundquist et al. 2026 |
| November 2026 | Chapter 15's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the seven solutions, and short “why?” questions, answered in footnotes, in three of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Warm-up questions in Chapter 16, in the same form (11 questions in five boxes): among them whether the displacement current is a flow of charge, what changes when the gap of a charging capacitor conducts slightly, how the field between its plates can be found both from the displacement current and from the real currents alone, and whether the quasistatic approximation depends on the size of the current. | Novak et al. 1999; Roediger and Karpicke 2006; Rundquist et al. 2026 |
| November 2026 | Chapter 16's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the three solutions, and short “why?” questions, answered in footnotes, in two of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Warm-up questions in Chapter 17, in the same form (9 questions in four boxes): among them what becomes of ε0 if the coulomb is redefined, whether a field of one statvolt per centimeter pushes on a charge as hard as a field of one gauss, and why no choice of units can remove the 4π from both Coulomb's law and Gauss's law. | Novak et al. 1999; Roediger and Karpicke 2006; Rundquist et al. 2026 |
| November 2026 | Chapter 1's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the eight solutions, and short “why?” questions, answered in footnotes, in five of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Chapter 3's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the five solutions, and short “why?” questions, answered in footnotes, in four of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Chapter 5's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the twelve solutions, and short “why?” questions, answered in footnotes, in six of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | The worked examples of Chapters 6, 7, and 8, in the form of the Chapter 4 pilot: a “Reality check:” in each of the eight solutions, and short “why?” questions, answered in footnotes, in three of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Chapter 9's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the twelve solutions, and short “why?” questions, answered in footnotes, in two of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Chapters 10 and 11's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the twelve solutions, and short “why?” questions, answered in footnotes, in two of them. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | The step lists of the script-r method (Sections 3.5, 5.5, 10.6, and 12.3) now say what the element is (the size of one piece of the source, so that λ dℓ′ is its charge; a step in an angle needs its scale factor, r′ dθ′ and not dθ′; for a current, I dℓ′ is the moving charge in the piece times its velocity) and how to set the limits: integrate over the source, counting each piece of it once. | Xue 2021 |
| November 2026 | Symmetry arguments in two separate steps: the step lists for Gauss's law and Ampère's law (Sections 4.3 and 11.2), the lists of symmetries that follow them (Sections 4.3.2 and 11.2.2), Examples 11.1–11.4, and the two chapter summaries now ask which components of the field are zero, and why, separately from which coordinates it can depend on, and why. Section 4.3.2 names the kinds of reason for each (no piece of the source contributes to a component, pairs of pieces cancel it, or a rotation that leaves the source unchanged would reverse it; sliding or rotating the field point leaves the source looking the same) and makes both arguments for a sphere. Part 2(b) of Tutorial 11.1 asks the two questions separately as well. | Xue 2021 |
| November 2026 | Warm-up questions in Chapter 18, in the same form (12 questions in five boxes): among them what multiplying by 1 + i does to a point, why zz* is always positive although z² is real only for some z, why the cube roots of a complex number differ for m = 0 and m = 1, what must be known before saying that a current with a phase of 40° peaks before the voltage, which way a wave travels, how a complex wave number changes a wave, and, in the other sign convention, whether the mass of a driven oscillator leads the force. | Novak et al. 1999; Roediger and Karpicke 2006 |
| November 2026 | Chapter 18's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the two solutions, and a short “why?” question, answered in a footnote, in one of them (why the mass lags the force by exactly 90° at resonance). This is the first chapter of the second semester to have the worked-example features. | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Section 14.3 now cites the research behind its statement that students often have trouble identifying which method to use when a problem does not announce it (as when problems are not grouped by chapter, as on a final exam), where the “which method?” flowcharts are introduced. | Wilcox et al. 2013b |
| November 2026 | Warm-up questions in Chapter 19, in the same form (17 questions in seven boxes): among them whether charge collects where a copper wire carrying a steady current joins an aluminum one, what must differ about the conduction electrons of silver if it conducts best although aluminum has more of them per unit volume, which end of an inductor is at the higher potential while the current through it decreases, and why the impedance of a circuit of resistors, capacitors and inductors can never have a negative real part. | Novak et al. 1999; Roediger and Karpicke 2006 |
| November 2026 | Chapter 19's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the five solutions, and a short “why?” question, answered in a footnote, in one of them (why the voltages across the inductor and the capacitor of a series circuit are always opposite in phase). | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| November 2026 | Warm-up questions in Chapter 20, in the same form (13 questions in five boxes): among them whether the magnetic field does part of the work of heating a wire because the Poynting vector contains it, which sign the stress tensor gives to a face of an imaginary box in a uniform field, what is wrong with concluding that charges outside a Maxwellian surface cannot affect the force on the charges inside it, and whether fields that lose their angular momentum pass it on when the shells that carry them are held fixed. | Novak et al. 1999; Roediger and Karpicke 2006 |
| November 2026 | Chapter 20's worked examples, in the form of the Chapter 4 pilot: a “Reality check:” in each of the five solutions, and short “why?” questions, answered in footnotes, in three of them (Examples 20.1, 20.3, and 20.4: why no energy flows along the inside of the wire, why the hemisphere far away can be dropped, and why a Maxwellian surface may cut through the ball of charge). | Wilcox et al. 2013b; Chi et al. 1989; Smith et al. 2010 |
| Planned | Warm-up questions in the other chapters; the rest of the end-of-chapter review, first in Chapter 4 (its “how would you solve this problem?” questions next); eight new tutorials. | See the sections above |
Sources
All are peer-reviewed except one Ph.D. dissertation (Xue 2021) and the CU Boulder course materials at the end.
- P. Alstein, K. Krijtenburg-Lewerissa, and W. R. van Joolingen, “Teaching and learning special relativity theory in secondary and lower undergraduate education: A literature review,” Phys. Rev. Phys. Educ. Res. 17, 023101 (2021). https://doi.org/10.1103/PhysRevPhysEducRes.17.023101
- B. S. Ambrose, P. R. L. Heron, S. Vokos, and L. C. McDermott, “Student understanding of light as an electromagnetic wave: Relating the formalism to physical phenomena,” Am. J. Phys. 67, 891–898 (1999). https://doi.org/10.1119/1.19144
- C. Baily, L. Bollen, A. Pattie, P. van Kampen, and M. De Cock, “Student thinking about the divergence and curl in mathematics and physics contexts,” 2015 PERC Proceedings, 51–54 (2015). https://doi.org/10.1119/perc.2015.pr.008
- C. Baily, M. Dubson, and S. J. Pollock, “Developing tutorials for advanced physics students: processes and lessons learned,” 2013 PERC Proceedings, 61–64 (2014). https://doi.org/10.1119/perc.2013.pr.003
- C. Baily, Q. X. Ryan, C. Astolfi, and S. J. Pollock, “Conceptual assessment tool for advanced undergraduate electrodynamics,” Phys. Rev. Phys. Educ. Res. 13, 020113 (2017). https://doi.org/10.1103/PhysRevPhysEducRes.13.020113
- M. Beege, M. Wirzberger, S. Nebel, S. Schneider, N. Schmidt, and G. D. Rey, “Spatial continuity effect vs. spatial contiguity failure. Revising the effects of spatial proximity between related and unrelated representations,” Front. Educ. 4, 86 (2019). https://doi.org/10.3389/feduc.2019.00086
- L. Bollen, P. van Kampen, and M. De Cock, “Students' difficulties with vector calculus in electrodynamics,” Phys. Rev. ST Phys. Educ. Res. 11, 020129 (2015). https://doi.org/10.1103/PhysRevSTPER.11.020129
- L. Bollen, P. van Kampen, C. Baily, and M. De Cock, “Qualitative investigation into students' use of divergence and curl in electromagnetism,” Phys. Rev. Phys. Educ. Res. 12, 020134 (2016). https://doi.org/10.1103/PhysRevPhysEducRes.12.020134
- L. Bollen, P. van Kampen, C. Baily, M. Kelly, and M. De Cock, “Student difficulties regarding symbolic and graphical representations of vector fields,” Phys. Rev. Phys. Educ. Res. 13, 020109 (2017). https://doi.org/10.1103/PhysRevPhysEducRes.13.020109
- L. Bollen, P. van Kampen, and M. De Cock, “Development, implementation, and assessment of a guided-inquiry teaching-learning sequence on vector calculus in electrodynamics,” Phys. Rev. Phys. Educ. Res. 14, 020115 (2018). https://doi.org/10.1103/PhysRevPhysEducRes.14.020115
- B. R. Brown, A. Mason, and C. Singh, “Improving performance in quantum mechanics with explicit incentives to correct mistakes,” Phys. Rev. Phys. Educ. Res. 12, 010121 (2016). https://doi.org/10.1103/PhysRevPhysEducRes.12.010121
- E. Campos, E. Hernandez, P. Barniol, and G. Zavala, “Analysis and comparison of students' conceptual understanding of symmetry arguments in Gauss's and Ampere's laws,” Phys. Rev. Phys. Educ. Res. 19, 010103 (2023). https://doi.org/10.1103/PhysRevPhysEducRes.19.010103
- E. Campos, G. Zavala, K. Zuza, and J. Guisasola, “Electric field lines: The implications of students' interpretation on their understanding of the concept of electric field and of the superposition principle,” Am. J. Phys. 87, 660–667 (2019). https://doi.org/10.1119/1.5100588
- S. V. Chasteen, “Teasing out the effect of tutorials via multiple regression,” 2011 PERC Proceedings, AIP Conf. Proc. 1413, 143–146 (2012). https://doi.org/10.1063/1.3680015
- S. V. Chasteen, S. J. Pollock, R. E. Pepper, and K. K. Perkins, “Transforming the junior level: Outcomes from instruction and research in E&M,” Phys. Rev. ST Phys. Educ. Res. 8, 020107 (2012). https://doi.org/10.1103/PhysRevSTPER.8.020107
- M. T. H. Chi, M. Bassok, M. W. Lewis, P. Reimann, and R. Glaser, “Self-explanations: How students study and use examples in learning to solve problems,” Cogn. Sci. 13, 145–182 (1989). https://doi.org/10.1016/0364-0213(89)90002-5
- J. Dunlosky, K. A. Rawson, E. J. Marsh, M. J. Nathan, and D. T. Willingham, “Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology,” Psychol. Sci. Public Interest 14, 4–58 (2013). https://doi.org/10.1177/1529100612453266
- P. Ginns, “Integrating information: A meta-analysis of the spatial contiguity and temporal contiguity effects,” Learn. Instr. 16, 511–525 (2006). https://doi.org/10.1016/j.learninstruc.2006.10.001
- J. Hecht, A. J. Mason, M. Rundquist, S. Read, D. Neilsen, and J. S. Colton, “Problem-Solving Frameworks for Tutorials in Upper-Division Electricity and Magnetism,” 2025 PERC Proceedings, 190–195 (2025). https://doi.org/10.1119/perc.2025.pr.Hecht
- T. N. Höffler and D. Leutner, “Instructional animation versus static pictures: A meta-analysis,” Learn. Instr. 17, 722–738 (2007). https://doi.org/10.1016/j.learninstruc.2007.09.013
- J. A. Huffman, M. Vignal, and B. R. Wilcox, “Investigating upper-division students' interpretations of the divergence theorem,” 2020 PERC Proceedings, 222–227 (2020). https://doi.org/10.1119/perc.2020.pr.Huffman
- S. Kalyuga, P. Ayres, P. Chandler, and J. Sweller, “The expertise reversal effect,” Educ. Psychol. 38, 23–31 (2003). https://doi.org/10.1207/S15326985EP3801_4
- A. Kienitz, M.-C. Krebs, and A. Eitel, “Seductive details hamper learning even when they do not disrupt,” Instr. Sci. 51, 595–616 (2023). https://doi.org/10.1007/s11251-023-09632-w
- W. J. Leonard, R. J. Dufresne, and J. P. Mestre, “Using qualitative problem-solving strategies to highlight the role of conceptual knowledge in solving problems,” Am. J. Phys. 64, 1495–1503 (1996). https://doi.org/10.1119/1.18409
- C. A. Manogue, K. Browne, T. Dray, and B. Edwards, “Why is Ampère's law so hard? A look at middle-division physics,” Am. J. Phys. 74, 344–350 (2006). https://doi.org/10.1119/1.2181179
- A. Maries, M. J. Brundage, and C. Singh, “Using the Conceptual Survey of Electricity and Magnetism to investigate progression in student understanding from introductory to advanced levels,” Phys. Rev. Phys. Educ. Res. 18, 020114 (2022). https://doi.org/10.1103/PhysRevPhysEducRes.18.020114
- A. J. Mason and J. S. Colton, “Reworking exam problems to incentivize improved performance in upper-division electrodynamics,” 2022 PERC Proceedings, 298–303 (2022). https://doi.org/10.1119/perc.2022.pr.Mason
- A. J. Mason, J. M. McCardell, P. A. White, and J. S. Colton, “Improving performance in upper-division electricity and magnetism with explicit incentives to correct mistakes,” Phys. Rev. Phys. Educ. Res. 19, 020104 (2023). https://doi.org/10.1103/PhysRevPhysEducRes.19.020104
- A. J. Mason, J. Hecht, D. Baldwin-Bott, D. Neilsen, and J. S. Colton, “A Preliminary Study of Upper-Division Tutorials Explicitly Built on a Problem-Solving Framework,” 2024 PERC Proceedings, 268–273 (2024). https://doi.org/10.1119/perc.2024.pr.Mason
- G. M. Novak, E. T. Patterson, A. D. Gavrin, and W. Christian, Just-in-Time Teaching: Blending Active Learning with Web Technology (Prentice Hall, 1999). https://aapt.org/Conferences/newfaculty/upload/jittbook.pdf
- R. E. Pepper, S. V. Chasteen, S. J. Pollock, and K. K. Perkins, “Our best juniors still struggle with Gauss's Law: Characterizing their difficulties,” 2010 PERC Proceedings, AIP Conf. Proc. 1289, 245–248 (2010). https://doi.org/10.1063/1.3515212
- R. E. Pepper, S. V. Chasteen, S. J. Pollock, and K. K. Perkins, “Observations on student difficulties with mathematics in upper-division electricity and magnetism,” Phys. Rev. ST Phys. Educ. Res. 8, 010111 (2012). https://doi.org/10.1103/PhysRevSTPER.8.010111
- S. J. Pollock, S. V. Chasteen, M. Dubson, and K. K. Perkins, “The use of concept tests and peer instruction in upper-division physics,” 2010 PERC Proceedings, AIP Conf. Proc. 1289, 261–264 (2010). https://doi.org/10.1063/1.3515218
- S. J. Pollock and B. R. Wilcox, “Upper-division students' use of separation of variables,” 2015 PERC Proceedings, 259–262 (2015). https://doi.org/10.1119/perc.2015.pr.060
- A. Renkl and R. K. Atkinson, “Structuring the transition from example study to problem solving in cognitive skill acquisition: A cognitive load perspective,” Educ. Psychol. 38, 15–22 (2003). https://doi.org/10.1207/S15326985EP3801_3
- L. E. Richland, N. Kornell, and L. S. Kao, “The pretesting effect: Do unsuccessful retrieval attempts enhance learning?,” J. Exp. Psychol. Appl. 15, 243–257 (2009). https://doi.org/10.1037/a0016496
- H. L. Roediger and J. D. Karpicke, “Test-enhanced learning: Taking memory tests improves long-term retention,” Psychol. Sci. 17, 249–255 (2006). https://doi.org/10.1111/j.1467-9280.2006.01693.x
- D. Rosengrant, A. Van Heuvelen, and E. Etkina, “Do students use and understand free-body diagrams?,” Phys. Rev. ST Phys. Educ. Res. 5, 010108 (2009). https://doi.org/10.1103/PhysRevSTPER.5.010108
- M. Rundquist, S. Read, J. Hecht, J. Colton, and A. Mason, “The electro-magnetism inventory for conceptual evaluation: preliminary results and refinement,” accepted for the 2026 PERC Proceedings.
- Q. X. Ryan, B. R. Wilcox, and S. J. Pollock, “Student difficulties with boundary conditions in electrodynamics,” 2015 PERC Proceedings, 283–286 (2015). https://doi.org/10.1119/perc.2015.pr.066
- Q. X. Ryan, B. R. Wilcox, and S. J. Pollock, “Student difficulties with boundary conditions in the context of electromagnetic waves,” Phys. Rev. Phys. Educ. Res. 14, 020126 (2018). https://doi.org/10.1103/PhysRevPhysEducRes.14.020126
- J. Samani and S. C. Pan, “Interleaved practice enhances memory and problem-solving ability in undergraduate physics,” npj Sci. Learn. 6, 32 (2021). https://doi.org/10.1038/s41539-021-00110-x
- R. E. Scherr, P. S. Shaffer, and S. Vokos, “Student understanding of time in special relativity: Simultaneity and reference frames,” Am. J. Phys. 69, S24–S35 (2001). https://doi.org/10.1119/1.1371254
- R. E. Scherr, P. S. Shaffer, and S. Vokos, “The challenge of changing deeply held student beliefs about the relativity of simultaneity,” Am. J. Phys. 70, 1238–1248 (2002). https://doi.org/10.1119/1.1509420
- C. Singh, “Student understanding of symmetry and Gauss's law of electricity,” Am. J. Phys. 74, 923–936 (2006). https://doi.org/10.1119/1.2238883
- A. D. Smith, J. P. Mestre, and B. H. Ross, “Eye-gaze patterns as students study worked-out examples in mechanics,” Phys. Rev. ST Phys. Educ. Res. 6, 020118 (2010). https://doi.org/10.1103/PhysRevSTPER.6.020118
- J. T. Stuckey and G. M. Quan, “Student understanding of divergence and curl in upper-division electricity and magnetism,” 2024 PERC Proceedings, 414–419 (2024). https://doi.org/10.1119/perc.2024.pr.Stuckey
- Á. Suárez, A. C. Martí, K. Zuza, and J. Guisasola, “Learning difficulties among students when applying Ampère-Maxwell's law and its implications for teaching,” Phys. Rev. Phys. Educ. Res. 20, 010143 (2024). https://doi.org/10.1103/PhysRevPhysEducRes.20.010143
- Á. Suárez, A. C. Martí, K. Zuza, and J. Guisasola, “Understanding student challenges with circulation and Ampère–Maxwell law,” Eur. J. Phys. 46, 025701 (2025). https://doi.org/10.1088/1361-6404/ada1b1
- C. D. Tippett, “Refutation text in science education: A review of two decades of research,” Int. J. Sci. Math. Educ. 8, 951–970 (2010). https://doi.org/10.1007/s10763-010-9203-x
- C. S. Wallace and S. V. Chasteen, “Upper-division students' difficulties with Ampère's law,” Phys. Rev. ST Phys. Educ. Res. 6, 020115 (2010). https://doi.org/10.1103/PhysRevSTPER.6.020115
- B. R. Wilcox, M. D. Caballero, R. E. Pepper, and S. J. Pollock, “Upper-division student understanding of Coulomb's law: Difficulties with continuous charge distributions,” 2012 PERC Proceedings, AIP Conf. Proc. 1513, 418–421 (2013) [2013a]. https://doi.org/10.1063/1.4789741
- B. R. Wilcox, M. D. Caballero, D. A. Rehn, and S. J. Pollock, “Analytic framework for students' use of mathematics in upper-division physics,” Phys. Rev. ST Phys. Educ. Res. 9, 020119 (2013) [2013b]. https://doi.org/10.1103/PhysRevSTPER.9.020119
- B. R. Wilcox and G. Corsiglia, “Cross-context look at upper-division student difficulties with integration,” Phys. Rev. Phys. Educ. Res. 15, 020136 (2019). https://doi.org/10.1103/PhysRevPhysEducRes.15.020136
- B. R. Wilcox and S. J. Pollock, “Coupled multiple-response versus free-response conceptual assessment: An example from upper-division physics,” Phys. Rev. ST Phys. Educ. Res. 10, 020124 (2014). https://doi.org/10.1103/PhysRevSTPER.10.020124
- B. R. Wilcox and S. J. Pollock, “Upper-division student difficulties with the Dirac delta function,” Phys. Rev. ST Phys. Educ. Res. 11, 010108 (2015) [2015a]. https://doi.org/10.1103/PhysRevSTPER.11.010108
- B. R. Wilcox and S. J. Pollock, “Upper-division student difficulties with separation of variables,” Phys. Rev. ST Phys. Educ. Res. 11, 020131 (2015) [2015b]. https://doi.org/10.1103/PhysRevSTPER.11.020131
- B. C. Xue, “Identifying and Addressing Student Difficulties with the Source-Field Relationship using Tutorials in Upper-Division Electromagnetism,” Ph.D. dissertation, University of Washington (2021). https://digital.lib.washington.edu/researchworks/handle/1773/47123
- K. Zuza, J. Guisasola, M. Michelini, and L. Santi, “Rethinking Faraday's law for teaching motional electromotive force,” Eur. J. Phys. 33, 397–406 (2012). https://doi.org/10.1088/0143-0807/33/2/397
- K. Zuza, J.-M. Almudí, A. Leniz, and J. Guisasola, “Addressing students' difficulties with Faraday's law: A guided problem solving approach,” Phys. Rev. ST Phys. Educ. Res. 10, 010122 (2014). https://doi.org/10.1103/PhysRevSTPER.10.010122
- K. Zuza, M. De Cock, P. van Kampen, T. Kelly, and J. Guisasola, “Guiding students towards an understanding of the electromotive force concept in electromagnetic phenomena through a teaching-learning sequence,” Phys. Rev. Phys. Educ. Res. 16, 020110 (2020). https://doi.org/10.1103/PhysRevPhysEducRes.16.020110
- Course materials (not peer-reviewed): University of Colorado Boulder, Physics Education Research group, “Resources – Course Materials,” https://www.colorado.edu/per/resources/course-materials (the junior-level electrostatics and magnetostatics materials, with their lists of student difficulties).