what works what doesn't what's promising

what works what doesn't scientific american mind and what works what doesn't
Dr.PeterCena Profile Pic
Published Date:02-07-2017
Your Website URL(Optional)
46 scientific american mind s eptember/October 2013 miq513Dunl3p.indd 46 7/1/13 6:04 PMH HOW OW WE WE LE LEAR ARN N WHAT WORKS, WHAT DOESN’T Some study techniques accelerate learning, whereas — others are just a waste of time but which ones are which? An unprecedented review maps out the best pathways to knowledge B Y JOHN DUNLOSKY, KATHERINE A. RAWSON, ELIZABETH J. MARSH, MITCHELL J. NATHAN AND DANIEL T. WILLINGHAM ILLUSTRATIONS BY CELIA JOHNSON Mind.ScientificAmerican.com SCIENTIFIC AMERICAN MIND 47 miq513Dunl3p.indd 47 7/1/13 6:04 PM ANDREW LILLEY iStockphoto (blocks)ducation generally focuses on what you study, such as algebra, the elements of the periodic table or how to conjugate verbs. But learning how to study can be just as important, with lifelong bene ts. It can teach you to pick up knowledge faster and E more ef ciently and allow you to retain information for years rather than days. Cognitive and educational psychologists have developed ported by experimental evidence, and students are not being and evaluated numerous techniques, ranging from rereading taught how to use the ones that work well. In fact, the two to summarizing to self-testing, for more than 100 years. Some study aids that students rely on the most are not effective. One common strategies markedly improve student achievement, of them may even undermine success. whereas others are time-consuming and ineffective. Yet this One potential reason is that the huge amount of research is information is not making its way into the classroom. Teach- overwhelming, making it dif cult for educators and students ers today are not being told which learning techniques are sup- to identify the most practical and advantageous ways to study. To meet this challenge, we reviewed more than 700 scienti c articles on 10 commonly used learning techniques. We focused FAST FACTS on strategies that seem to be easy to use and broadly effective. Rating the Best Ways to Study We also took a closer look at a couple of methods that are very popular with students. Some study methods work in many different To receive our recommendation, a technique must be 1 situations and across topics, boosting test useful in a range of learning conditions, such as whether a stu- performance and long-term retention. Learning how dent works alone or in a group. It must assist learners of vari- to learn can have lifelong bene ts. ous ages, abilities and levels of prior knowledge—and it must have been tested in a classroom or other real-world situation. Self-testing and spreading out study ses- Learners should be able to use the method to master a variety 2 sions—so-called distributed practice—are of subjects, and their performance should bene t no matter excellent ways to improve learning. They are ef cient, what kind of test is used to measure it. The best approaches easy to use and effective. also result in long-lasting improvements in knowledge and comprehension. Underlining and rereading, two methods that Using these criteria, we identi ed two clear winners. They 3 many students use, are ineffective and can produced robust, durable results and were relevant in many sit- be time-consuming. uations. Three more are recommended with reservations, and  ve—including two popular learning aids—are not advised, ei- Other learning techniques need further test- ther because they are useful only in limited circumstances or 4 ing and evaluation. In the meantime, stu- because not enough evidence supports a higher rating. We en- dents and teachers can put proved study methods to courage researchers to further explore some of the untested use in classrooms and at home. techniques, but students and teachers should be cautious about relying on them. 48 SCIENTIFIC AMERICAN MIND September/October 2013 miq513Dunl3p.indd 48 7/1/13 5:54 PMTHE GOLD STAR WINNERS T TH HE G E GO OL LD S D ST TA AR W R WIIN NN NE ER RS S 1. SELF-TESTING Quizzing Yourself Gets High Marks asked to memorize word pairs, half of tention for people with Alzheimer’s dis- which were then included on a recall ease. Short, frequent exams are most ef- test. One week later the students remem- fective, especially when test takers re- bered 35 percent of the word pairs they ceive feedback on the correct answers. had been tested on, compared with only Practice testing works even when its 4 percent of those they had not. In an- format is different from that of the real other demonstration, undergraduates test. The bene cial effects may last for were presented with Swahili-English months to years—great news, given that word pairs, followed by either practice durable learning is so important. testing or review. Recall for items they had been repeatedly tested on was 80 IS IT PRACTICAL? Yes. It requires modest percent, compared with only 36 percent amounts of time and little to no training. for items they had restudied. One theory is that practice testing triggers a mental HOW CAN I DO IT? Students can self-test search of long-term memory that acti- with  ash cards or by using the Cornell HOW IT WORKS: Unlike a test that evalu- vates related information, forming mul- system: during in-class note taking, ates knowledge, practice tests are done tiple memory pathways that make the make a column on one edge of the page by students on their own, outside of information easier to access. where you enter key terms or questions. class. Methods might include using  ash You can test yourself later by covering cards (physical or digital) to test recall or WHEN DOES IT WORK? Anyone from pre- the notes and answering the questions answering the sample questions at the schoolers to fourth-year medical stu- (or explaining the keywords) on the end of a textbook chapter. Although dents to middle-age adults can bene t other side. most students prefer to take as few tests from practice testing. It can be used for as possible, hundreds of experiments all kinds of factual information, includ- RATING: High utility. Practice testing show that self-testing improves learning ing learning words in foreign languages, works across an impressive range of for- and retention. making spelling lists and memorizing mats, content, learner ages and retention In one study, undergraduates were the parts of  owers. It even improves re- intervals. 2. DISTRIBUTED PRACTICE WE REVIEWED For Best Results, Spread Your Study over Time MORE THAN 700 HOW IT WORKS: Students often “mass” SCIENTIFIC ARTI CLES their study—in other words, they cram. ON 10 COM MON But distributing learning over time is much more effective. In one classic exper- LEARNING iment, students learned the English equiv- TECHNIQUES TO alents of Spanish words, then reviewed the material in six sessions. One group did IDENTIFY THE MOST the review sessions back to back, another ADVANTAGEOUS had them one day apart and a third did the reviews 30 days apart. The students in WAYS TO STUDY. the 30-day group remembered the trans- lations the best. In an analysis of 254 Mind.ScientificAmerican.com SCIENTIFIC AMERICAN MIND 49 miq513Dunl3p.indd 49 7/1/13 5:55 PM ISTOCKPHOTO (star)studies involving more than 14,000 par- your own. You will have to plan ahead should be spaced six to 12 months apart. ticipants, students recalled more after and overcome the common student ten- Although it may not seem like it, you ac- spaced study (scoring 47 percent overall) dency to procrastinate. tually do retain information even during than after massed study (37 percent). these long intervals, and you quickly re- HOW CAN I DO IT? Longer intervals are learn what you have forgotten. Long de- WHEN DOES IT WORK? Children as young generally more effective. In one study, lays between study periods are ideal to as age three bene t, as do undergraduates 30-day delays improved performance retain fundamental concepts that form and older adults. Distributed practice is more than lags of just one day. In an In- the basis for advanced knowledge. effective for learning foreign vocabulary, ternet-based study of trivia learning, word de nitions, and even skills such as peak performance came when sessions RATING: High utility. Distributed prac- mathematics, music and surgery. were spaced at about 10 to 20 percent of tice is effective for learners of different the retention interval. To remember ages studying a wide variety of materials IS IT PRACTICAL? Yes. Although text- something for one week, learning epi- and over long delays. It is easy to do and books usually group problems together sodes should be 12 to 24 hours apart; to has been used successfully in a number by topic, you can intersperse them on remember something for  ve years, they of real-world classroom studies. THE RUNNERS-UP THE RUNNERS-UP Despite their promise, the following learning techniques fall short, in many cases because not enough evidence has been amassed to support their use. Some techniques, such as elaborative in- terrogation and self-explanation, have not been evaluated suf ciently in real-world educational contexts. Another emerging method called interleaved practice has just begun to be systematical- ly explored. Nevertheless, these techniques show enough potential for us to recommend their use in the situations described brie y here. 3. ELABORATIVE INTERROGATION Channel Your Inner Four-Year-Old tive interrogation, learners produce ex- the subject. Its power increases with prior planations for facts, such as “Why does knowledge; German students bene tted it make sense that…?” or “Why is this from elaborative interrogation more true?” In one experiment, for example, when they were learning about German students read sentences such as “the hun- states than about Canadian provinces, for gry man got into the car.” Participants in example. It may be that prior knowledge an elaborative interrogation group were permits students to generate more appro- asked to explain why, whereas others priate explanations for why a fact is true. were provided with an explanation, such as “the hungry man got into the car to go to the restaurant.” A third group simply PROMPTING STU- read each sentence. When asked to recall which man performed what action DENTS TO ANSWER (“Who got in the car?”), the elaborative- “WHY?” QUESTIONS, HOW IT WORKS: Inquisitive by nature, we interrogation group answered about 72 are always looking for explanations for percent correctly, compared with about CALLED ELABORA- the world around us. A sizable body of 37 percent for the others. TIVE INTERROGATION, evidence suggests that prompting stu- dents to answer “Why?” questions also WHEN SHOULD I USE IT? When you are ALSO FACILITATES facilitates learning. learning factual information—particular- LEARNING. With this technique, called elabora- ly if you already know something about 50 SCIENTIFIC AMERICAN MIND September/October 2013 miq513Dunl3p.indd 50 7/1/13 7:07 PMThe effects of this technique appear to IS IT PRACTICAL? Yes. It requires minimal works for a broad range of topics but be robust across ages, from fourth grad- training and makes reasonable time de- may not be useful for material more ers through undergraduates. Elaborative mands. In one study, an elaborative- complex than a factual list. Bene ts for interrogation clearly improves memory interrogation group required 32 minutes learners without prior knowledge may for facts, but whether it also might en- to do a task that took 28 minutes for a be limited. More research will be needed hance comprehension is less certain, and reading-only group. to establish whether elaborative interro- there is no conclusive information about gation generalizes to various situations how long the gains in learning persist. RATING: Moderate utility. The technique and different types of information. 4. SELF-EXPLANATION How Do I Know? sentence provide for you?” and “How whether the technique is more lasting in does it relate to what you already know?” people of high or low knowledge. Similar to elaborative interrogation, self- explanation may help integrate new in- IS IT PRACTICAL? Unclear. On the one formation with prior knowledge. hand, most students need minimal in- struction and little to no practice, although WHEN SHOULD I USE IT? It bene ts kinder- one test of ninth graders showed that stu- gartners to college students and helps in dents without training tended to para- solving math problems and logical rea- phrase rather than generate explanations. soning puzzles, learning from narrative On the other, a few studies report that this texts and even mastering endgame strat- technique is time-consuming, increasing egies in chess. In younger children, self- time demands by 30 to 100 percent. explanation can help with basic ideas such as learning numbers or patterns. RATING: Moderate utility. Self-explana- The technique improves memory, com- tion works across different subjects and HOW IT WORKS: Students generate expla- prehension and problem solving—an im- an impressive age range. Further research nations of what they learn, reviewing pressive range of outcomes. Most studies, must establish whether these effects are their mental processing with questions however, have measured effects within durable and whether the time demands such as “What new information does the only a few minutes, and it is not known make it worthwhile. 5. INTERLEAVED PRACTICE (The Authors) Mixing Apples and Oranges JOHN DUNLOSKY is professor of psy- HOW IT WORKS: Students tend to study in chology at Kent State University. blocks,  nishing one topic or type of KATHERINE A. RAWSON is associate problem before moving on to the next. professor of psychology at Kent But recent research has shown bene ts State. ELIZABETH J. MARSH is asso- for interleaved practice, in which stu- ciate professor of psychology and dents alternate a variety of types of in- neuroscience at Duke University. formation or problems. In one study, for MITCHELL J. NATHAN is professor of example, college students learned to psychology, educational psychology, compute the volumes of four different and curriculum and instruction at geometric shapes. In a so-called blocked- the University of Wisconsin–Madi- practice condition, they  nished all the son. DANIEL T. WILLINGHAM is pro- problems for one shape before moving fessor of psychology at the Universi- on to the next. In interleaved practice, ty of Virginia. the problems were intermixed. When tested one week later, the interleaved Mind.ScientificAmerican.com SCIENTIFIC AMERICAN MIND 51 miq513Dunl3p.indd 51 7/1/13 5:55 PMWhat Doesn’t Work hese techniques were rated as low ability is also woefully underexplored. utility because they are inef cient, Most of the bene t of rereading appears T ineffective or bene cial only for cer- to accrue from the second reading, with tain types of learning and for short periods diminishing returns from additional rep- of retention. Most students report reread- etitions. No experimental research has ing and highlighting, yet these techniques assessed it using materials from actual do not consistently boost performance, courses—ironic, given that this strategy and they distract students from more pro- is the one most commonly reported by ductive strategies. Other methods men- students. tioned below are just too time-consuming. WHAT YOU SHOULD DO INSTEAD: Don’t HIGHLIGHTING waste your time—in head-to-head com- ■ Students commonly report under- parisons, rereading fares poorly against lining, highlighting or otherwise marking more active strategies such as elabora- material. It is simple and quick—but it does little to improve per- tive interrogation, self-explanation and practice testing. formance. In controlled studies, highlighting has failed to help U.S. Air Force basic trainees, children and remedial students, as Three less commonly used study techniques also fared +  well as typical undergraduates. Underlining was ineffective re- ■ poorly in our assessment. “Imagery for text learning” gardless of text length and topic, whether it was aerodynamics, needs more evidence before it can be recommended, whereas ancient Greek schools or Tanzania. “summarization” and “keyword mnemonic” appear to be ineffec- In fact, it may actually hurt performance on some higher-level tive and time-consuming. tasks. One study of education majors found that underlining re- In summarization, students identify a text’s main points, ex- duced their ability to draw inferences from a history textbook. It cluding unimportant material. Whether it works is dif cult to an- may be that underlining draws attention to individual items rather swer, as it has been implemented in many different ways. It is than to connections across items. unknown whether summarizing small pieces of a text or large chunks of it works better or whether the length, readability or or- WHAT YOU SHOULD DO INSTEAD: Highlighting or underlining can ganization of the material matters. be useful if it is the beginning of a journey—if the marked informa- With keyword mnemonics, imagery is used to enhance mem- tion is then turned into  ash cards or self-tests. Given that stu- ory; for example, a student learning the French word la dent dents are very likely to continue to use this popular technique, fu- (“tooth”) might use the similar-sounding English word “dentist” to ture research should be aimed at teaching students how to high- form a mental image of a dentist holding a large molar. Mnemon- light more effectively—which likely means doing it more judiciously ics do seem to help with foreign-language vocabulary, word de ni- (most undergraduates overmark texts) and putting that informa- tions and medical terminology, but the effects have not been tion to work with a more useful learning technique. shown to endure, and in the end the effort involved in generating keywords may not be an ef cient use of time. REREADING Another technique that uses mental pictures is imagery for ■ In one survey of undergraduates at an elite university, 84 text learning, in which students are told to create images for every percent said they reread textbooks or notes during study. It re- paragraph they read. Research has revealed a patchwork of incon- quires no training, makes modest demands on time, and has sistent results that have not been shown to last over the long shown some bene ts on recall and  ll-in-the-blank-style tests. term. Teachers may consider instructing students to attempt Yet the evidence is muddy that rereading strengthens compre- using this technique with image-friendly texts, but further dem- hension, and whether its effects depend on knowledge level or onstrations of its usefulness are necessary. See the Psychological Science in the Public Interest article “Improving Students’ Learning with Effective Learning Techniques: Promising Directions from Cognitive and Educational Psychology,” on which this story for Scienti c American Mind is based,   at the Association for Psychological Science’s Web site: www.psychologicalscience.org 52 SCIENTIFIC AMERICAN MIND September/October 2013 miq513Dunl3p.indd 52 7/1/13 5:55 PMpractice group was 43 percent more ac- ing without any instruction. Teachers STUDENTS ARE NOT curate. Interleaving allows students to could also use the technique in the class- practice selecting the correct method room: After one kind of problem (or BEING TAUGHT THE and encourages them to compare differ- topic) is introduced, practice  rst focus- BEST STRATEGIES, ent kinds of problems. es on that problem. Once the next kind of problem is introduced, it is mixed in PERHAPS BECAUSE WHEN SHOULD I USE IT? When the types with examples of earlier subjects. It may TEACHERS THEM- of problems are similar, perhaps be- take a little more time than blocking cause juxtaposing them makes it easier practice, but such slowing most likely is SELVES ARE NOT to see what is different about them. worthwhile, re ecting cognitive pro- SCHOOLED IN THEM. Blocked practice—doing all the items cesses that boost performance. from one category in a row—may be more effective when the examples are RATING: Moderate utility. Interleaved not very much alike because it high- nose cardiac disorders. Yet two studies practice improves learning and reten- lights what they have in common. of foreign-vocabulary learning showed tion of mathematical knowledge and It is possible that interleaved prac- no effect for interleaved practice. Nev- boosts other cognitive skills. The litera- tice bene ts only those who are already ertheless, given how much difficulty ture on interleaved practice is small, reasonably competent. Outcomes are many students have in mathematics, it however, and includes enough negative also mixed for different types of con- may still be a worthwhile strategy for results to raise concern. It may be that tent. It improves performance on alge- that subject. the technique does not consistently bra problems and was effective in a work well, or perhaps it is not always study that trained medical students to IS IT PRACTICAL? It seems to be. A moti- used appropriately—topics for future interpret electrical recordings to diag- vated student could easily use interleav- research. What We Have Learned trained or reminded, still require further (Further Reading) Why don’t students use more effec- research. But even now teachers can in- tive study techniques? It seems they are corporate the most successful approach- ◆ Ten Bene ts of Testing and Their not being taught the best strategies, per- es into lesson plans so that students could Applications to Educational Practice. H. L. Roediger III, A. L. haps because teachers themselves are not adopt them on their own. For instance, Putnam and M. A. Smith in Psychol- schooled in them. In our survey of six ed- when moving to a new section, a teacher ogy of Learning and Motivation, Vol. ucational-psychology textbooks, only can start by asking students to do a prac- 55: Cognition in Education. Edited one technique—“keyword mnemon- tice test that covers important ideas from by Jose P. Mestre and Brian H. ics”—was covered in every book. None the previous section and providing im- Ross. Academic Press, 2011. ◆ Interleaving Helps Students offered much guidance on the use, effec- mediate feedback. Students can inter- Distinguish among Similar tiveness or limitations of different ways leave new problems with related ones Concepts. D. Rohrer in Education- of studying. from preceding units. Teachers can har- al Psychology Review, Vol. 24, A second problem may be that in the ness distributed practice by reintroduc- No. 3, pages 355–367; educational system, the emphasis is on ing major concepts during the course of September 2012. ◆ Using Spacing to Enhance teaching students critical-thinking skills several classes. They can engage students Diverse Forms of Learning: and content. Less time is spent on teach- in explanatory questioning by prompting Review of Recent Research and ing them how to learn. The result can be them to consider how the information is Implications for Instruction. S. K. that students who do well in their early new to them or why it might be true. Carpenter, N. J. Cepeda, D. Rohrer, years, when learning is closely super- These learning techniques are no S.H.K. Kang and H. Pashler, ibid., pages 369–378. vised, may struggle once they are expect- panacea. They bene t only those who ◆ When Is Practice Testing Most ed to regulate their own learning in high are motivated and capable of using them. Effective for Improving the Dura- school or college. Nevertheless, we expect that students bility and Ef ciency of Student Some questions, such as the best age will make meaningful gains in class- Learning? K. A. Rawson and J. for students to start using a technique room performance, on achievement tests Dunlosky, ibid., pages 419–435. and how often they will need to be re- and during their lifetime. M Mind.ScientificAmerican.com SCIENTIFIC AMERICAN MIND 53 miq513Dunl3p.indd 53 7/1/13 5:55 PM