Update: Bass, Carpenter and Fritz have recently presented this effective practice at the following conferences:

• 10/27/21 Educause Poster | presentation file (UMBC login req'd)
• 10/12/21 UMBC Faculty Development Center workshop
• 10/1/21 USM "Silver Linings" Faculty Showcase
• 5/18/21 Educause Learning Initiative (ELI) annual meeting
• 2/4/21 International Center for Academic Integrity (ICAI) Mid-Atlantic conference.
• FYI: A very similar approach at the University of Michigan

In a recent Washington Post story about UMBC’s pandemic response to online learning, reporter Nick Anderson mentioned a video prepared by chemistry senior lecturer Tara Carpenter to address a key issue: academic integrity of online exams.

[Carpenter] has tinkered with the pacing and format of quizzes and tests to prevent students from getting an unfair advantage through the Internet. In a video she shared with The Post, Carpenter told students she detected some cheating on the first exam. She warned them to avoid the temptation to use online help forums as an illicit shortcut to the answer. "Posting questions to have someone else answer them is 100 percent cheating," she said. "I'm watching for it, and I will report it."

The video -- a 25-min. Panopto screencast that Carpenter has agreed to answer questions about in a UMBC VoiceThread -- was prepared for her College of Natural and Mathematical Sciences (CNMS) colleagues, as part of the summer Planning Instructional Variety in Online Teaching (PIVOT) program. 

Carpenter developed her approach with chemistry colleague, lecturer Sarah Bass, and leverages three key functions available in all Blackboard Original courses (the bolded first part of each step links to a related Blackboard FAQ and/or instructional technology effective practice):

A large pool or bank of questions that can automatically populate any exam (see up to timecode 4:53 in the screencast, which covers the first three steps in Carpenter’s accompanying spreadsheet). Carpenter and Bass have developed nearly 1,500 questions that can be used in four, 20-25 question “Learning Checkpoint” (LCP) assessments in this semester’s CHEM 101 and CHEM 102 “Introduction to Chemistry” courses that collectively have more than 900 enrolled students. In steps four through six (up to time code 10:12), Carpenter creates four parts to an LCP assessment that are populated with questions of comparable rigor or complexity, which she’s identified and organized in the overall pool or bank.
Create groups (invisible to students), randomly populated for the exam that correspond to an LCP’s four parts, but no one group gets the same sequence of parts (this is step 7 and goes up to time code 11:43). Along with variations of questions from pools, the variation of exam parts they comprise means the chances that any two students get the exact same exam questions or sequence are very low (below is an example of the schema from Carpenter’s screencast):

Adaptive release of content, which is used to require that students “sign” an honor pledge before they can access the rest of the CHEM LCP assessments. This begins steps 8 through 10, which “bring it all together” in terms of how pools, groups and the honor pledge “pre-requisite” students must meet to access and complete an LCP exam. In addition to reminding students of UMBC’s commitment to academic integrity, which the Faculty Development Center has encouraged for years, Carpenter and Bass allow students to use their notes and textbook. But given the complexity, variety and length of the LCP assessments, they both encourage students not to rely on this strategy alone. “They’ll simply run out of time,” says Bass. “Students seem to get that.”

In addition to Chemistry, three Physics instructors, Eric Anderson, Lili Cui and Cody Goolsby-Cole, have already adopted Carpenter’s and Bass’ approach for assessments in their high-enrollment introductory Physics courses this term. “It has been working out well,” says Cui, who, along with her colleagues, had already adjusted her approach last spring, by spreading out fewer, monthly high-stakes exams into more frequent, weekly Friday quizzes that cumulatively have the same weight as prior terms. This also could facilitate a testing effect (link provided by FDC Director Linda Hodges), in which assessments can be opportunities for students to learn in their own right, perhaps especially if they occur sooner and more frequently after study.

“At first glance, Carpenter’s and Bass’ approach may appear to be complex, but it’s actually a pretty straight-forward use of current Blackboard functionality, especially for high-enrollment courses,” says John Fritz, associate vice president of Instructional Technology. “Also, it focuses less on trying to catch students cheating and more on appealing to -- maybe even rewarding -- their own commitment to effective study habits and academic integrity.”

By John Fritz

View the recording ▶️

View the recording ▶️

View the recording ▶️

View the recording ▶️

View the recording ▶️