Sometimes it seems easier, at least for me, to remember an answer, when there are more than 2 options to choose from.

For example, I easily remember my pin (14 bit), but hardly the lighter switch (up or down- 1 bit) for "on".

Why is that? Or is it just a wrong impression?

  • $\begingroup$ Could this get a more descriptive (read, searchable) title? I get what you're getting at, but I think "why I remember long/complex stuff while I forget simple stuff" is closer to how most would think of the problem $\endgroup$ – Ben Brocka Jun 9 '12 at 19:45
  • $\begingroup$ I have dyslexia and I have similar issues with binary problems; I'm always confusing extremes, especially with something symmetric like a light switch. $\endgroup$ – Keegan Keplinger Jun 9 '12 at 20:52
  • $\begingroup$ As a side not, computational analogues (esp. number of bits) can be very misleading when it comes to human mind. $\endgroup$ – Piotr Migdal Jun 9 '12 at 23:21
  • $\begingroup$ @StevenJeuris I don't think this title really captures the problem.. $\endgroup$ – Ben Brocka Jun 10 '12 at 1:14
  • $\begingroup$ It is a wrong impression. You remember your pin best because you actively learned it. You did that because you will have a problem if you forget it. The on or off position of your light switch isn't important because it takes you only 1 sec to try it out. $\endgroup$ – danijar Jun 28 '12 at 22:06

There are at least three phenomena, which may play the role:

A too short information can easily interfere. So if it is on/off, it may interfere with its state a few days before, or with if the iron is on or off.

Moreover, things are remembered with redundancy. So, for example, one can remember not only PIN as numbers visually, but also as words, as a 'snapshot' of numbers written on a letter, as a sequence of finger moves, etc. So in case one is lost or inactive, all other can be used instead.

Also, one there is a piece of information, you can relate (consciously or not) to other objects. E.g. you can additionally remember that there are two digits from your birthdate and there is 8, a number you like very much, ...

Actually, mnemonic techniques base on the three observations, to make it easier to learn or harder to forget.

As a side note, computational analogues (esp. number of bits) are usually very misleading when it comes to human mind. The later does not work as an easy in/out device storying faithfully digital information.

(BTW: very likely, still it is much easier to remember 4 digits (of base 10) than 14 zeroes/ones.)


The simple fact of the mater is you're talking Short Term Memory (STM) for the light switch versus Long Term Memory (LTM) for the PIN.

From Wikipedia's definition of Long Term Memory (which is fairly accurate, if simplistic)

Long-term memory (LTM) is memory in which associations among items are stored, as part of the theory of a dual-store memory model. According to the theory, long-term memory differs structurally and functionally from working memory or short-term memory, which ostensibly stores items for only around 20–30 seconds and can be recalled easily. This differs from the theory of the single-store retrieved context model that has no differentiation between short-term and long-term memory.

You never encode simple things like light switch positions into Long Term Memory because it would be useless; it's going to change and it's not that important (you can always check it's state readily). You pay very little attention to the light switch meaning you will likely not even keep it in Short Term Memory.

Your PIN on the other hand you have rehearsed dozens of times, tried very hard to remember, gotten some sleep (which helps the encoding process into Long Term Memory) and have built up many associations (this number gives me money, reminds me of my wallet, ect).

All forms of human memory are limited, so it's very important that we bias our memory toward important stuff; in STM you have to keep working with an item constantly to remember it; "call bob call bob call bob". But once you've elaborated on that item (I need to call bob at 3PM because he owes me money), rehersed it repeatedly, written it down (maintenance and elaborative rehearsal) it might slip into long term memory. But, since unlike your PIN, it will eventually be lost to forgetting (hopefully) since you don't continue to dwell on it, and the information stops being useful.

  • $\begingroup$ But you would encode a light switch with a LED indicator indicating whether or not the light is on into long term memory. :) ... proving your point I guess. $\endgroup$ – Steven Jeuris Jun 9 '12 at 18:02
  • $\begingroup$ @StevenJeuris not sure what you mean, that you would remember that "lit LED = light is on"? $\endgroup$ – Ben Brocka Jun 9 '12 at 19:44
  • $\begingroup$ I wasn't sure which one was which so I didn't mention it intentionally. :) We don't have those around here. But yes, that's what I was referring to. $\endgroup$ – Steven Jeuris Jun 9 '12 at 21:00
  • 1
    $\begingroup$ The question seems to have nothing to do with long vs short term memory. $\endgroup$ – Piotr Migdal Jun 9 '12 at 23:18
  • $\begingroup$ @PiotrMigdal not explicitly, but very few "1 bit" states would ever enter long term memory for the reasons I lay out $\endgroup$ – Ben Brocka Jun 10 '12 at 1:56

The short term v. long term memory answer is a good one. But what if they are both long term memory?

An alternative is that you just aren't counting bits correctly. Your 14 bit PIN is static. Once it is set one might actually argue it contains 0 bits within a given context. It has no probability of being anything else. The light switch is not static. You don't encode one instance of a light switch being on or off but multiple ones over time. The last episode of your kitchen light being on or off is one in a long string. If you were to consider the current state of your light at home being just the last of a very long list of times you had to remember that then the light switch actually contains vastly more information.

Another interpretation (sometimes used to further explain the one above) is not to consider information so much as interference. If you have one PIN# that you use within a given context then there's no interference from competing information. However, the extensive proactive interference from past traces of the state of the light switch limit your certainty about it's state. You might want to look up interference theory on wikipedia.

  • $\begingroup$ Thanks for your answer. I like the static/dynamic interpretation. $\endgroup$ – draks ... Jun 10 '12 at 11:26

A few thoughts:

  • Automaticity and consistency: It is much easier to automate a response that is consistent. Thus, if all the lights that you interact with follow the same rules then you will make the correct decision more often and you are more likely to automate the response.
  • cost of not recalling: The cost of not recalling your password can be high. This is particularly true for credit cards and so on. The cost of not knowing what is "on" on a light is minimal, just try the other option.
  • contingent responses: There are different ways that you can perform tasks. With the light switch, you could have a decision rule that says if I want to have light on and light is off, toggle switch. Such a decision rule does not require knowledge of which state is on and which is off. In particular, if your life is filled with inconsistent light switches then the toggle decision rule may be more adaptive.
  • automaticity and recall: If you have automated a response, you may even forget in a declarative sense what you are doing.
  • 1
    $\begingroup$ Thanks for your answer. Remark: the cost of not recalling might be higher, when it comes to change a broken light. You should be sure that the current is switched off :-) $\endgroup$ – draks ... Jun 10 '12 at 11:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.