Contexts Question- What are your Contexts? Posted in Gear by mistake as well

TesTeq

Registered
7 or 5 to 9.

Desultory said:
It's impractical for me. If I see more than a few things (like *five*) on the list for a given context, I feel overwhelmed.
As far as I know people feel comfortable with lists of 7 items plus/minus 2 (5 to 9 items).
 

Brent

Registered
TesTeq said:
As far as I know people feel comfortable with lists of 7 items plus/minus 2 (5 to 9 items).

You know, I've heard this for years, but I don't think I've ever seen evidence of it.

I've seen evidence that people can keep no more than 7 (+/- 2) unrelated pieces of data in their mind at once, but that's it.
 

andersons

Registered
Brent said:
You know, I've heard this for years, but I don't think I've ever seen evidence of it.

I've seen evidence that people can keep no more than 7 (+/- 2) unrelated pieces of data in their mind at once, but that's it.
There are gazillions of studies confirming Miller's short-term memory limitation of 7 +/- 2 items, but yes, they are distinct items (not necessarily unrelated). "Chunking" is one way we increase our capacity to use information, by grouping related items into a single higher-order one. This is why hierarchical organization is so powerful. If you give someone a list already grouped, they will remember many more items than if the list were ungrouped and unsorted.

In real life, the limit is probably more like 5. We can get people to expend extra effort and energy to do their best in lab tests, but in real life it's too expensive to perform at our lab-best all the time. In our lab, we literally pay subjects bonuses for better performance. For short bursts of time they do their best, then head off to the pub when they're done. :)

Long-term memory is another way of increasing capacity. If you review a list of 100 things often enough, you build up strong long-term memories of it. So when you look at your list to choose from it, each item is a trigger for a strong existing memory. Contrast this with the case in which someone else gives you a list of 100 actions you've never seen or thought about before.

With my list of 100 actions sorted by priority, I have reviewed and consulted it enough times that I can reproduce the top 20 or 30 items from memory, no problem. If my PDA crashed, I would still know what I want to do today. That's because of long-term memories created by review. I could also probably remember many or most of the lower-priority items, but I would not be able to reproduce all of them without some trigger.

There's also a big difference in what we subjectively feel we remember versus actual memory performance. People can perform way above chance on a recall test, yet still believe they can't remember a thing. This is called "implicit" memory: your brain knows stuff that you don't know you know. :)
 

Brent

Registered
andersons said:
There are gazillions of studies confirming Miller's short-term memory limitation of 7 +/- 2 items

Where are they? I'm not trying to challenge you; I'd genuinely like to read them.
 
R

ritz

Guest
My contexts

@Anywhere
@Computer: Offiline
@Computer: Online
@Errands: [town I live in]
@Errands: [town I work in]
@Errands: [town I visit]
@Home
@People (my designation for @Agendas)
@Phone (my designation for @Calls)
@Site: [branch office name here]
@Site: [branch office name here]
@Site: [volunteer worksite here]
@Site: [main office here]

Rather than starting with the standard GTD contexts in my list, I started with no contexts, and added each context as required to file my NA's.

I'm careful only to create true contexts - i.e. groups that imply that the basic resources are available to perform the next actions contained within. I avoid creating other types of lists in order to subdivide next actions because in my experience, that results in overlap and having to hunt for NA's. For me, there is no @Waiting context, because when I have an NA that is waiting for an outside requirement to be met, I file it in the context in which it will eventually be performed, and annotate it with the requirement it is waiting for.
 

andersons

Registered
mostly not in my database, but here are a few

Brent said:
Where are they? I'm not trying to challenge you; I'd genuinely like to read them.

If you would like to read some of the current debate about capacity limitations, try
"The magical number 4 in short-term memory: A reconsideration of mental storage capacity" by Nelson Cowan, Behavioral and Brain Sciences, 2000, 24(1):87-185.​
The 99 pages show that without doubt there are memory and attention limitations, but considerable debate about the specifics. Cowan argues for a limit of 4 chunks. Behavioral and Brain Sciences is an excellent journal, because it publishes many leading researchers' responses to the main author's paper (hence the 99 pages of debate). The main author then answers the points debated by the other researchers. There are some 10 pages of references if you want more reading.

Be warned that this research perspective is to understand the brain's memory functions by isolating short-term memory, not to understand how people actually overcome it in their real-world tasks. (I have some papers about that too, but not handy.) Rehearsal, elaboration, encoding strategies, chunking, and long-term memory all increase memory performance. Researchers in short-term memory limitations specifically try to eliminate those strategies to better understand the underlying capacity limits.

You may be more interested in skilled memory performance, which has much more real-world application. K. Anders Ericsson claims that the critical aspect of exceptional memory performance is not the amount of information stored, but rather how information is stored in and retrieved from long-term memory -- precisely what is controlled for in the capacity-limit research. But exceptional memory performance does not come from increasing the capacity of short-term memory. Nice overview paper: "Expert performance: Its structure and acquisition" published in American Psychologist, August 1994, pp.725-804, by K. Anders Ericsson and Neil Charness. His website lists more.

The oft-cited paper by Miller was published in Psychological Review in 1956, volume 63, pp. 81-97: "The magical number seven, plus or minus two: some limits of our capacity for processing information." It relies heavily on information theory.
 

TesTeq

Registered
Is @WaitingFor a context? And is it really useful?

ritz said:
Rather than starting with the standard GTD contexts in my list, I started with no contexts, and added each context as required to file my NA's.
I agree totally with you but some people do not feel comfortable enough with the idea of contexts to make a disciplined approach to the incremental context definition process.

ritz said:
I'm careful only to create true contexts - i.e. groups that imply that the basic resources are available to perform the next actions contained within. I avoid creating other types of lists in order to subdivide next actions because in my experience, that results in overlap and having to hunt for NA's. For me, there is no @Waiting context, because when I have an NA that is waiting for an outside requirement to be met, I file it in the context in which it will eventually be performed, and annotate it with the requirement it is waiting for.
I don't think @WaitingFor is a real context. What resources are available in this context? Besides it interferes with other contexts (some follow-ups can be done in the @Call context only, some in the @Computer context, and some in one of the @Agenda contexts). So treating @WaitingFor as a context is misleading somehow (it is rather another class of the project's "bookmark").

To summarize - in GTD we can "close an open loop" for a given project by defining all NextActions (active - you should do something) and all WaitingFors (passive - somebody should do something).

But isn't it simpler to define the NextAction "Check if John prepared report" in your @Ofiice context instead of putting "John prepared report" in the artificial @WaitingFor context?
 

Desultory

Registered
TesTeq said:
I don't think @WaitingFor is a real context. What resources are available in this context?
As I understand it, "Waiting For" is a list on the same level as "Projects" and "Someday/Maybe." Depending on your implementation, I think a lot of us are treating the list as a category. (After all, I think I've seen reference to @Projects and @Someday, and I don't think anyone's confusing those for contexts!) The edges may be getting blurred because things we're "Waiting for" tend to revert to actionable items.
TesTeq said:
...isn't it simpler to define the NextAction "Check if John prepared report" in your @Ofiice context instead of putting "John prepared report" in the artificial @WaitingFor context?
Trouble is, "Check if John prepared report" isn't an NA until you've Waited a reasonable amount of time for him to do it. You need some other kind of trigger in the future, perhaps in the Tickler, perhaps in a Daily Review, to find out if you need to take action make sure John does turn in the report on time. At which point, the NA could be anything from "Ask John if he needs help" to "Read/Review John's Report".
 
Top