Improving Memory?

Get­ting from Here to There:
Mak­ing Mem­ory Con­sol­i­da­tion Work

By Bill Klemm, Ph. D.

Until con­sol­i­da­tion has occurred, a short-term mem­ory is very vul­ner­a­ble, as all of us have expe­ri­enced from look­ing up a phone num­ber only to have some dis­trac­tion cause us to lose the num­ber before we can get it dialed.

What is “consolidation”?

Brain researchers use the term “con­sol­i­da­tion” for the process whereby short-term mem­ory gets made more permanent.

Here, I would like to dis­cuss some aspects of con­sol­i­da­tion that many peo­ple may not know about: why sleep is so impor­tant, why mem­ory must be prac­ticed, and how test­ing pro­motes consolidation.

1. Over-training: You Can Learn Too Much

Exper­i­ments have shown that human mem­ory per­for­mance unex­pect­edly dete­ri­o­rated if learn­ing ses­sions were increased to four 60-minute ses­sions at reg­u­lar inter­vals on the same day. In other words, the more the sub­jects were trained, the poorer they per­formed. How­ever, this inter­fer­ence did not occur if sub­jects were allowed to nap for 30–60 min­utes between the sec­ond and third sessions.

It is hard to explain why over-training dis­rupts per­for­mance, but I sus­pect that as train­ing tri­als are repeated the infor­ma­tion starts to inter­fere with mem­ory con­sol­i­da­tion, per­haps because of bore­dom or fatigue in the neural cir­cuits that medi­ate the learn­ing. Nap­ping must have a restora­tive func­tion that com­pen­sates for the neg­a­tive effects of over-training. What all this sug­gests is that mem­ory con­sol­i­da­tion would be opti­mized if learn­ing occurred in short ses­sions that are repeated but only with inter­ven­ing naps and on dif­fer­ent days with reg­u­lar night-time sleep. In other words, repeat­ing long study peri­ods in the same day on the same task can be counter-productive. This is yet another rea­son why stu­dents should not cram-study for exams. Learn­ing should be opti­mized by rehears­ing the same learn­ing mate­r­ial on sep­a­rate days where nor­mal sleep occurred each night.

Sources:

- Maquet, P. et al. 2002. Be caught nap­ping: you’re doing more than rest­ing your eyes.Nature Neu­ro­science. 5 (7); 618–619.

- Med­nick, Sara, et al. 2002. The restora­tive effect of naps on per­cep­tual deterioration.Nature Neu­ro­science. 5 (7): 677–681.

2. Los­ing Your Past

Do you remem­ber the names of your elementary-school teach­ers? How about the name of the bully in mid­dle school? Or names of your friends when you were a kid? These are all things you remem­bered well at one time and remem­bered for a long time. But you may well have for­got­ten by now.

A recent study on rats sug­gests what it takes to sus­tain longer term mem­o­ries. Rats in the study learned a “bait shy­ness” task. Rats were given a drink of saccharin-flavored water, and then shortly after­wards injected with lithium, which made them nau­se­ated. This was a typ­i­cal con­di­tioned learn­ing sit­u­a­tion, as with Pavlov’s dogs. In this case, rats typ­i­cally remem­bered to avoid such water for many weeks. This is the basis for “bait shy­ness.” If rats sur­vive a poi­son­ing episode, they will avoid that bait in the future. In this exper­i­ment, one group of rats received an injec­tion directly into the part of the brain that holds taste mem­o­ries. This injec­tion con­tained a drug that blocks a cer­tain enzyme, a pro­tein kinase. These rats lost their learned taste aver­sion. The bad mem­ory was lost irre­spec­tive of when the injec­tion was made dur­ing the 25 days after learn­ing occurred. Giv­ing the enzyme blocker before learn­ing had no effect on learn­ing to avoid the fla­vored water. The pro­tein kinase thus seems to be nec­es­sary for sus­tain­ing a long-term mem­ory. It is pos­si­ble that other long-term mem­o­ries the rats may have had were also wiped out by the enzyme-blocking drug.

So what is the prac­ti­cal impor­tance? I sug­gest that even “long-term” mem­o­ries have to get rehearsed or they may even­tu­ally for­got­ten. Or if you do remem­ber, there is a good chance that the mem­ory is cor­rupted, that is, not totally cor­rect. The con­se­quence is that things that hap­pened long ago may be either for­got­ten, or misremembered.

What sus­tains the enzyme nec­es­sary for long-term mem­ory? I sus­pect it is rehearsal and peri­odic reac­ti­va­tion of the mem­ory. Some sci­en­tists are excited about the pos­si­bil­ity of devel­op­ing a drug to manip­u­late lev­els of the enzyme. The prob­lem with that, how­ever, is that the drug could abol­ish old mem­o­ries that you might not want to for­get (like your name) or may cause you to remem­ber too much that is now irrelevant.

Source: Shema, R., Sack­tor, T. C., and Dudai, Y. 2007. Rapid era­sure of long-term mem­ory asso­ci­a­tions in the cor­tex by an inhibitor of PKM. Sci­ence. 317:951–953.

3. Test­ing Pro­motes Consolidation

Tests do more than just mea­sure learn­ing. Tests are learn­ing events. That is, test­ing forces retrieval of incom­pletely learned mate­r­ial and that very act of retrieval is a rehearsal process that helps to make the learn­ing more per­ma­nent. Test­ing, and not actual study­ing, is the key fac­tor on whether or not learn­ing is con­sol­i­dated into longer term memory.

A recent report from Wash­ing­ton Uni­ver­sity in St. Louis, exam­ined the role that retrieval had on the abil­ity to recall that same mate­r­ial after a delay of a week. In the exper­i­ment, col­lege stu­dents were to learn a list of 40 for­eign lan­guage vocab­u­lary word pairs that were manip­u­lated so that the pairs either remained in the list (were repeat­edly stud­ied) or were dropped from the list once they were recalled. It was like study­ing flash cards: one way is to keep study­ing all the cards over and over again; the other way is to drop out a card from the stack every time you cor­rectly recalled what was on the other side of the card. After a fixed study period, stu­dents were tested over either the entire list or a par­tial list of only the pairs that had not been dropped dur­ing study. Four study and test peri­ods alter­nated back-to-back. Stu­dents were also asked to pre­dict how many pairs they would be able to remem­ber a week later, and their pre­dic­tions were com­pared with actual results on a final test a week later.

The ini­tial learn­ing took about 3–4 tri­als to mas­ter the list, and was not sig­nif­i­cantly affected by the strat­egy used (rehears­ing the entire list or drop­ping items out as they were recalled). On aver­age, the stu­dents pre­dicted that they would be able to remem­ber about half of the list on a test that was to be given a week later. How­ever, actual recall a week later var­ied con­sid­er­ably depend­ing on learn­ing con­di­tions. On the final test, stu­dents remem­bered about 80% of the word pairs if they had been tested on all the word pairs, no mat­ter whether they had been stud­ied mul­ti­ple times with all of them in the list or if they dropped cor­rectly recalled words from the list in later study tri­als. How­ever, recall was only about 30% cor­rect when cor­rectly iden­ti­fied words were dropped from sub­se­quent tests, even though all words were stud­ied repeat­edly. In other words, it was the repeated test­ing, not the study­ing, that was the key fac­tor in suc­cess­ful longer-term memory.

So, what is the prac­ti­cal appli­ca­tion? When using flash cards, for exam­ple, you need to fol­low each study ses­sion (whether or not you drop cards from the stack because you know them), with a for­mal test over all the cards. Then, repeat the process sev­eral times, with study and test epochs back-to-back. Can we extend this prin­ci­ple of fre­quent test­ing to other kinds of learn­ing strate­gies? I would guess so.

Why does forced recall, as dur­ing test­ing, pro­mote con­sol­i­da­tion? It prob­a­bly relates to other recent dis­cov­er­ies show­ing that each time some­thing is recalled the mem­ory is re-consolidated. If the same infor­ma­tion is con­sol­i­dated again and again, the mem­ory is pre­sum­ably reinforced.

This study also showed that the sub­jects could not pre­dict how well they would remem­ber, which is con­sis­tent with my 45 years expe­ri­ence as a pro­fes­sor. Stu­dents are fre­quently sur­prised to dis­cover after an exam­i­na­tion that they did not know the mate­r­ial as well as they thought they did. Tests not only reveal what they know and don’t know, but serve to increase how much they even­tu­ally learn. If I were still teach­ing, I would give more tests. And I would encour­age stu­dents to use self-testing as a rou­tine learn­ing strat­egy, some­thing that one study revealed to be a seldom-used strat­egy. The repeated self-tests should include all the study mate­r­ial and not drop out the mate­r­ial that the stu­dent thinks is already mastered.

Source: Karpicke, Jef­frey D., and Roedinger, Henry L. III. 2008. The crit­i­cal impor­tance of retrieval for learn­ing. Sci­ence. 319: 966–968.

— W. R. (Bill) Klemm, D.V.M., Ph.D. Sci­en­tist, pro­fes­sor, author, speaker As a pro­fes­sor of Neu­ro­science at Texas A&M Uni­ver­sity, Bill has taught about the brain and behav­ior at all lev­els, from fresh­men, to seniors, to grad­u­ate stu­dents to post-docs. His recent books include Thank You Brain For All You Remem­ber and