LTM Retrieval: Getting information to LTM from STM


 Like real banks, it is easier to make deposits into our memory banks than it is to make withdrawals (Much information is forgotten, but not gone). Below are 3 examples of retrieval--but not storage-- failures. In technical terminology, the following 3 examples illustrate that information available (stored) in memory is not always accessible (retrievable).

  1. Tip-Of-the-Tongue (TOT) phenomenon (words are not actually on the tip of your tongue--they are in memory, but can't be retrieved)
  2. Recognition is generally easier than recall (e.g., it is easier for you to recognize the names of the 7 dwarfs than it is for you to retrieve all 7).
  3. Savings scores: Relearning is faster than learning (to revise an old saying, "You can quickly re-teach an old dog old tricks"). For example, suppose you needed 20 minutes to learn the 7 dwarfs the first time, but now you can't remember a single one. The good news is that it might take you only 5 minutes to learn them them all the second time-- a savings of  15 minutes (which, using the formula for savings scores as time saved/original time spent, results in a savings score of 15/20 which is 75%). Similarly, even if it seemed that you "forgot" everything you learned in a course, you  could  relearn that information much more easily than when you learned it the first time. Indeed, when taking advanced courses that have introductory courses as prerequisites, you will probably quickly relearn what you learned--but seemed to have totally forgotten--from those introductory courses. (So, although you can question whether a tree falling in a forest makes a sound if nobody is around to hear it, an encoded memory does leave a trace [detectable by recognition and savings] even when it cannot be retrieved.)

Because retrieval is such a big problem (and because tests ask you to take information out of memory rather than put information into memory), much of your study time should focus on retrieving--not merely recognizing!--information. Specifically, you should

What causes retrieval failures?

Not simply the passing of time (Despite recent attempts to revive decay theory, the idea that memories decay from LTM is pretty much dead.)

Evidence that retrieval failures are not due to time alone:

  1. Grandparents can have vivid and accurate memories for long ago events while having hazy and inaccurate memories for recent events.
  2. Hypermnesia: Under certain conditions, people can have better recall for information a week or longer after they learned the information than they had shortly after they learned the information. Usually, those "certain conditions" involve being repeatedly tested on the same information without ever seeing the information again. In a typical study, a participant might be given a list of 24 words to remember. A few minutes later, when asked to recall the words, the participant may recall 18 of the words. A week later, participants return to the lab and are again asked to remember the words. On the first attempt, the participant may recall only 1 or 2 words. But, with more attempts, the participant will recall more words. Indeed, by the 6th time the participant tries to recall the words, the participant may recall all 24 words--even though the participant could only recall 18 words when tested right after being given the list! Hypermnesia is possible because some of the "forgotten" information doesn't go away; it is just hard to find. As a result, when participants keep looking, they find much of that "forgotten" information.
(To see a diagram of a typical hypermnesia experiment, mouse over this link.)
Typical Hypermnesia Experimental Procedure and Results
Typical Procedure Study List of 24 Words. Recall words Experimenter: "Experiment is over, but would you like to come in next
week for a different memory experiment?"
"Recall words from last week."
Recall 1
Recall 2 Recall 3 Recall 4 Recall 5 Recall 6
Typical Number of Words Recalled   18   3 6 11 14 18 24


So, on the one hand, time, by itself, does not cause retrieval failures.

On the other hand, however,  retrieval failures are often linked to time as Ebbinghaus' forgetting curve illustrates (see the graph  below).


* If you just focused on the rapid drop during the first part of Ebbinghaus' forgetting curve, forgot about savings (that relearning is much faster than the original learning), and did not realize that meaningful information is retained much longer than the nonsense syllables that Ebbinghaus used in his forgetting studies, you might think that this comedian's idea for a 5-minute university was a great idea rather than just a great comedy routine.

After noting that the forgetting curve starts off as a sharply falling line, but then becomes a curve as forgetting levels off, answer the following three questions. Then, check your answers by clicking on the buttons below. 

  1. What bad news about memory does the forgetting curve reveal?

  2. Answer: You can forget a great deal of information in just an hour, and you can forget about 2/3 in less than a week. This is why you must continue to test yourself over the material even after you know it. So, you need to review your class notes regularly, and you can't assume that the information you knew for the quiz is information that you will remember for the final.

  3. What good news about memory does the forgetting curve reveal?
    Answer: After a week, for most practical purposes, you stop forgetting.

  4. Does the forgetting curve support or go against the idea that we should have year-round schools to prevent the forgetting that occurs over the long summer vacation?
    Answer: Since almost all forgetting occurs in the first week, it really doesn't matter whether summer break is 2 weeks or 12 weeks.

But if time doesn't cause retrieval failures (as suggested by hypermnesia and people having good memories for long ago events), why are retrieval failures often linked to time--as you know from the forgetting curve and your own experience? To answer that question, let's first ask: "Why do retrieval failures happen at all?"

3 proposed reasons:

  1. Interference: getting sidetracked by bumping into similar, but wrong information. This skit from SNL dramatizes how information that is perceived as similar to what you want to remember can interfere with retrieval of the desired information. 
  2. Cue-related forgetting (lacking the right cues): not knowing where to look for the information (as you'll see, being without cues--being "cue less"-- may result in being clueless when it comes to recalling information).

  3. Repression (unconscious motivated forgetting): unconsciously not wanting to remember the information.

We will now look at each of these three explanations for retrieval failures in more depth.

#1. Interference: If you just throw stuff into your memory like you would throw stuff into a garbage dump, you may not be able to find what you are looking for because other stuff  is in the way. (Sherlock Holmes explains it this way) If some of the other stuff resembles what you are looking for, it may be that you have two or more memories competing to be retrieved--and the loser may be the one you wanted to retrieve.

Really a problem when information is perceived as similar.

So, when studying information, you should try to make the new information different from what you already know before you try to memorize it. Similarly, if you are using imagery to memorize something, you might try to make your image unusual in some way, such as making it much bigger than such normal objects really are.

2 types of interference:

Proactive interference: Old (Previously learned) information hurts retrieval of new information.

Classic experimental set up for  demonstrating proactive interference:


Group 1Learns
List A
List B
Tested on
List B
Group 2Learns
List B
Tested on
List B

Results: Group 1 does worse than Group 2 because proactive interference from List A acts to interfere with Group 1's recall of List B. How much worse? That will depend on how similar the two lists are-- the more similar, the worse Group 1's recall.

Think of other examples of proactive interference. Hints:

Retroactive interference: Newly (Recently) learned information acts to hurt memory for old information ("retro" means "backwards").

Classic experimental set up for  demonstrating retroactive interference:

Group 1Learns
List A
List B
Tested on
List A
Group 2Learns
List A
Tested on
List A

Results: Group 1 does worse than Group 2 because retroactive interference from the recently learned List B acts to interfere with Group 1's recall of List A. Will the act of learning about retroactive interference recently act to interfere with your memory of proactive interference? Have your more recent phone numbers, addresses, and passwords interfered with access to your old ones?

Animation to help you understand the difference between proactive and retroactive interference

Practice distinguishing proactive interference from retroactive interference

A phenomenon that shows both types of interference and also shows how passing of time can't account for forgetting--the serial position curve:

Serial position
Graph Courtesy of Creative Commons License 3.0 via Wikimedia Commons

Questions to think about when looking at the serial position curve

Given that recall is good for the beginning and for the end, but poor for the middle (e.g., we can easily remember the first U.S. President [Washington] and, despite what repression would predict, the last former President [Trump], but may have trouble remembering middle presidents like Chester Arthur), what does this mean in terms of

Your knowledge of interference can help you refute lies. The problem with trying to refute a lie is that to refute it, you usually repeat it--and repeating it may actually make people remember the lie. The solution is a "truth sandwich" in which you state the facts, refute the lie, and then state the facts again. That way, the lie is subjected to proactive interference from your first statement of the fact and retroactive interference from your final restatement of the fact. In short, just like with the serial position curve, people will remember the beginning and the end of what you said (the fact) rather than what you said in the middle (the lie).

Short (less than 1 minute) video to help you understand interference and the serial position curve.

Short (one minute) animation showing the implications of interference for how you should study.

Look at some terms that you might have trouble remembering because of interference

#2 Cue-Dependent Forgetting: Inadequate cues as a cause of retrieval failure

        Cues trigger memories. In a sense, the cues you have for retrieving the information are like hooks that help you fish for information: The more hooks, the more likely it is you will catch the information. Given the importance of cues, it is not surprising that much forgetting is due to not having the right cues. Not having cues, like not having the address of a person you want to visit or not having the file name for the computer file you want to access, makes it unlikely that you will find what you need.


(See additional possible explanations for childhood amnesia)

On to solving memory problems by using mnemonics

Back to Solving Encoding Problems

Back to Memory Menu

Back to Lecture Menu

Copyright 2020-2023 Mark L. Mitchell