Improve grades, reduce failure—undergrads should tell profs 'Don't lecture me'

May 12, 2014
Freeman heads into the auditorium to help as students huddle trying to reason out an answer to a genetics question. Credit: U of Washington

A significantly greater number of students fail science, engineering and math courses that are taught lecture-style than fail in classes incorporating so-called active learning that expects them to participate in discussions and problem-solving beyond what they've memorized.

Active learning also improves exam performance – in some cases enough to change grades by half a letter or more so a B-plus, for example, becomes an A-minus.

Those findings are from the largest and most comprehensive analysis ever published of studies comparing lecturing to in undergraduate education, said Scott Freeman, a University of Washington principal lecturer in biology. He's lead author of a paper in the Proceedings of the National Academy of Sciences the week of May 12.

Freeman and his co-authors based their findings on 225 studies of undergraduate education across all of the "STEM" areas: science, technology, engineering and mathematics. They found that 55 percent more students fail lecture-based courses than classes with at least some active learning. Two previous studies looked only at subsets of the STEM areas and none before considered failure rates.

On average across all the studies, a little more than one-third of students in traditional lecture classes failed – that is, they either withdrew or got Fs or Ds, which generally means they were ineligible to take more advanced courses. On average with active learning, a little more than one-fifth of students failed.

"If you have a course with 100 students signed up, about 34 fail if they get lectured to but only 22 fail if they do active learning according to our analysis," Freeman said. "There are hundreds of thousands of students taking STEM courses in U.S. colleges every year, so we're talking about tens of thousands of students who could stay in STEM majors instead of flunking out – every year."

This could go a long way toward meeting national calls like the one from the President's Council of Advisors on Science and Technology saying the U.S. needs a million more STEM majors in the future, Freeman said.

Attempts by college faculty to use active learning, long popular in K-12 classrooms, started taking off in the mid-1990s, Freeman said, though lecturing still dominates.

Scott Freeman proposes a question, has students consider answers individually or in groups and then calls on students at random to explain their reasoning -- an approach he and colleagues call "Ask, don't tell." Credit: U of Washington

"We've got to stop killing student performance and interest in science by lecturing and instead help them think like scientists," he said.

In introduction to biology courses, Freeman's largest UW class had 700 students, he expects students to read their $200 textbooks and arrive in class knowing the material for the day. Quizzes on the readings the night before keep their feet to the fire.

"These students got into college by being ferocious memorizers so we don't need to spend class time going over what they've already read," Freeman says. "A reading assignment on how sperm and eggs form might then lead me to ask the class how male contraceptives might work. After giving them time to come up with their own ideas and rationale, I might give them a couple more minutes to discuss it with each other, and then I call on students randomly to start the discussion."

Knowing they could get called on at any time encourages students to stay focused.

Having students use clickers ¬– hand-held wireless devices – to answer multiple-choice questions in class is another example of how active learning keeps students engaged.

"We characterize it as, 'Ask, don't tell,'" said Mary Wenderoth, a UW principal lecturer and a co-author on the paper.

For the paper, more than 640 studies comparing lecturing with some kind of active learning were examined by Freeman, Wenderoth and their other co-authors, Sarah Eddy, Miles McDonough, Nnadozie Okoroafor and Hannah Jordt, all with the UW biology department, and Michelle Smith with the University of Maine. The studies, conducted at four-year and community colleges mainly in the U.S., appeared in STEM education journals, databases, dissertations and conference proceedings.

Some 225 of those studies met the standards to be included in the analysis including: assurances the groups of students being compared were equally qualified and able; that instructors or groups of instructors were the same; and that exams given to measure performance were either exactly alike or used questions pulled from the same pool of questions each time.

The data were considered using meta-analysis, an approach long used in fields such as biomedicine to determine the effectiveness of a treatment based on studies with a variety of patient groups, providers and ways of administering the therapy or drugs.

About grade improvement, the findings showed improvements on exams increased an average of 6 percent. Using grading typical in UW's introductory biology, physics and chemistry courses, a gain of 6 percent would have raised students half a grade turning a C-plus into a B-minus, for example, or a B-plus into an A-minus.

If the failure rates of 34 percent for lecturing and 22 percent in classes with some active learning were applied to the 7 million U.S. undergraduates who say they want to pursue STEM majors, some 2.38 million students would fail lecture-style courses vs. 1.54 million with active learning. That's 840,000 additional students failing under lecturing, a difference of 55 percent compared to the failure rate of active learning.

"That 840,000 is a large portion of the million additional STEM majors the president's council called for," Freeman said.

Community colleges and universities could help faculty incorporate effective active learning by providing guidance – the UW, for instance, has a Center for Teaching and Learning to share expertise – as well as rewards, Freeman said.

Explore further: Observation protocol documents college instruction and STEM learning

More information: Active learning increases student performance in science, engineering, and mathematics, PNAS,

add to favorites email to friend print save as pdf

Related Stories

Recommended for you

Precarious work schedules common among younger workers

Aug 29, 2014

One wish many workers may have this Labor Day is for more control and predictability of their work schedules. A new report finds that unpredictability is widespread in many workers' schedules—one reason ...

Girls got game

Aug 29, 2014

Debi Taylor has worked in everything from construction development to IT, and is well and truly socialised into male-dominated workplaces. So when she found herself the only female in her game development ...

Computer games give a boost to English

Aug 28, 2014

If you want to make a mark in the world of computer games you had better have a good English vocabulary. It has now also been scientifically proven that someone who is good at computer games has a larger ...

Saddam Hussein—a sincere dictator?

Aug 28, 2014

Are political speeches manipulative and strategic? They could be – when politicians say one thing in public, and privately believe something else, political scientists say. Saddam Hussein's legacy of recording private discussions ...

User comments : 5

Adjust slider to filter visible comments by rank

Display comments: newest first

5 / 5 (3) May 12, 2014
Courses geared towards students majoring in the subject often include lab sessions - several hours per week more per credit hour than a lecture. Students lacking the interest or aptitude gravitate to the survey lecture-only courses.
The two sample populations have been pre-selected with a strong bias, so the samples are not comparable.
3 / 5 (2) May 12, 2014
Universities used to be geared toward the top 10% of students. Colleges, the top 30% and trade schools, the balance. The old way is superior.
5 / 5 (1) May 13, 2014
"the President's Council of Advisors on Science and Technology saying the U.S. needs a million more STEM majors in the future" -- don't believe it. Tech companies just want cheap labor.
not rated yet May 13, 2014
Courses geared towards students majoring in the subject often include lab sessions - several hours per week more per credit hour than a lecture. Students lacking the interest or aptitude gravitate to the survey lecture-only courses.
The two sample populations have been pre-selected with a strong bias, so the samples are not comparable.

If college were nothing more than some sort of higher tier trade school, I'd agree. In a decent undergrad curriculum however, all students do have to take courses outside their primary focus. Many of those classes are lecture style. At any given instant, how many students are devoting time to such classes? Does it make sense to seek to use that time to greater benefit?
1 / 5 (2) May 13, 2014
The old system gave us Shakespeare, Michelangelo, Newton, Boyle, Byron, Yeats, Maxwell, Twain, Sargent, Pasteur. Today, they provide individuals who only script and produce video games. From the steam engine to electricity, they produced technologies never seen before. Today, everything is entrenched so much that to suggest a physics beyond that seen is to open yourself to derision. The ability to accept the presence of God is in line with the philosophy that admits it has no means of denying God's presence. "Science" says it cannot prove God is not present but denies it anyway. As for getting rid of lectures, how many average students using average methods will be able to reproduce even just some of all centuries of mathematics in a few semesters, algebra, geometry, calculus, trigonometry, vector functions, complex analysis? How many can derive just the binomial formula on their own?