Pressing the Button

I’ve been reading in today’s newspapers about a study at Heidelburg University in which people were shown a succession of images and had to press a button to show whether they considered each image Good or Bad. The purpose of the experiment was to measure the amount of time it took for people of various ages to make decisions of this type. 

According to the press release this involved the participants in three actions for each image:

1) An intuitive decision

2) Pondering it

3) Pressing the button

Apparently after the age of 60 the brain’s speed at carrying this process out declines considerably. As someone considerably older than that I can testify to the truth of the finding!

No further details were given in the articles, but two thoughts struck me immediately. 

The first was that it would be very easy to get a good score by cheating 

The second was that this is very applicable to the speed at which decisions are made from a to-do list about which task to do next.

So how can one cheat? There are multiple ways. For example:

- push the Good button each time. No decision is made so only step 3) is needed.

- push the Bad button each time. Ditto.

- push the Good and the Bad button alternately. Ditto.

The problem with all of these is that it’s obvious you’re cheating. But you can be a little bit more sophisticated.

- push either the Good or the Bad button regardless of what the image looks like. Step 2 - the longest - is left out.

- decide which button to push before the image is displayed. Assuming there is even a small gap between images, only Step 3 would be needed. 

The last one would be the most difficult to detect as cheating. 

Now it’s quite possible that there were measures in the actual experiment which would have prevented any of these happening, but they are not mentioned in the brief description.

So if the speed of the decision can be greatly increased by leaving out steps 1) and 2) how is this applicable to the speed of deciding which task should be done next off a list?

Well, there is one way which I left out. It’s not exactly cheating either. And that is to make an intuitive decision and press the button immediately without pondering it.

So what you are aiming to do is to decide immediately without any second guessing. That takes practice but is quite achievable.

As an exercise write down ten tasks which you could do now. Once you’ve written them down, and without reading through the list, run your finger or a pen rapidly down the list (go as fast as you can while still being able to read the tasks). Stop at any point and then do the selected task for real. Do the same several times again until you’ve got the feel for it. 

That’s how it’s done. And you will almost certainly find that making your choices quickly in this way will also speed up the work you do on the task.

An important point:

When you are running your pen down the list keep going forward. Do not go back if you think you’ve missed a task. if you find yourself thinking about whether you’ve made the correct choice, stop thinking and just do the chosen task.

Question: Was the choice of tasks any worse than if you had spent time pondering them?

Comments

[Seraphim]

Thanks Mark, this has been thought-provoking and has given me some ideas.

Much like the Random method, this experiment seems to challenge the assumption that the key to productivity is deciding what tasks to do, and what order to do them in, which is the focus of almost every time management system I've ever seen (except the Random method).

http://markforster.squarespace.com/blog/2014/1/22/random-time-management.html

The valuable activity is not the decision-making -- this is so counter-intuitive but I really believe it is true.

I've had a similar experience while using the Dendro app to manage my reading and researching. http://dendro.cloud/

So if the valuable activity is not the decision-making, then what part of time management actually does add value?

With Random, I think the value is simply writing down whatever occurs to us as needing to be done. There are a few other classes of tasks that never make it onto the list --
-- the things that we will do anyway, even if we never write them down
-- the things we never even consider doing
-- the things we consider but reject

So there really is quite a lot of mental processing, conscious and probably unconscious, just in the process of adding things to a list. And the result is a list of potentially valuable stuff that we could or should do. I wonder if perhaps we are not fully aware of how much mental processing goes into this -- it seems like such a simple and unassuming process, to write stuff on a list.

So perhaps, the decision of which tasks to do first is a much lower value activity than this first task of writing down, in any random order that occurs to us, what needs to be done.

But there is also something special in the Random algorithm -- not every task has an equal chance of being selected. There is more chance of older items getting selected than newer items, with the result that old items are not allowed to languish, you are always moving forward, there is less weight of old backlog items holding you down. There is also an element of clumping and attenuation -- another result of the combination of the list-making process combined with the sliding algorithm.

I am wondering how to take advantage of this, and if there might be some way to improve on it. I've been tinkering a little using Random on my AF4R lists. AF4R already has a built-in focus on recurring (giving an overall structure to the day) and unfinished (driving things to completion), which gives you a sense of your overall workload, which attenuates the desire to start items on the New list. Where would be the best place to use Random?

Recurring? I don't know, this doesn't feel like the right place. The "standing out" selection process seems to allow a natural order of the day to emerge spontaneously, and I am afraid Random would wreak havoc with that.

New? I don't feel this is the right place for Random, either. Handling urgent tasks requires an understanding of which things are urgent -- Random can't distinguish that. And starting a new project is always a matter of careful reflection and consideration.

Unfinished? In my experience so far, I do occasionally get bogged down in deciding what Unfinished task to focus on, especially when I am feeling some time pressure and nothing really seems to stand out more than anything else. I suspect it really doesn't matter that much. So maybe I should use the Random algorithm to process this list. I'll experiment with this and see what happens.

Hm, it also occurs to me that the Unfinished list is a lot like the DIT will-do list -- it functions as a "Closed List" much of the time (until you activate some New task from the New list). And one of the principles from DIT is that it doesn't matter what order you do the items on a Closed List -- the important thing is that ALL of them need to get done. Maybe the Random method will help me do that more effectively.

AF4R - http://markforster.squarespace.com/forum/post/1325521

[Alan Baljeu]

About making quick decisions (because as asserted here decision time is largely wasteful):

I have for a while been pondering a no-list system, that doesn't quite work for me but I think with a tweak could be very effective, which is:

If you think of something you can improve now, do so.

This simple directive takes advantage of the fact that you are always thinking of things so the decision time becomes 0. There's no scanning, etc., and most of the time you think of things in front of you so the activation time is small. The only problem is it's not quite balanced and there are things I ought to do that just don't come to mind often enough with this minimal system.

So if I could find a way to balance this, it could be very excellent. One idea is to time-bound this mode of operating, so you have everything covered elsewhere but for this period you just ACT. Another idea is to add reminder systems so the would-be-neglected things aren't neglected. Anyway, I haven't got this idea mastered yet.

Mike Vardy preaches about themed days wherein if it's Tuesday you do Tuesday things and thus don't have to think about that. Decision is premade.

[Seraphim]

Alan Baljeu -
<< If you think of something you can improve now, do so. >>

My problem with that is I can always think of 10-20 things to improve at any given moment. The problem is to decide where to focus.

I like the idea of "themed days". Reminds me of Laura Ingalls Wilder - << Each day had its own proper work. Ma used to say: "Wash on Monday, Iron on Tuesday, Mend on Wednesday, Churn on Thursday, Clean on Friday, Bake on Saturday, Rest on Sunday." >> ~ Little House in the Big Woods

[Alan Baljeu]

Yes we all can think of things to improve at any moment. Reread Mark’s blogpost. Skip the deciding and go with the first thing that occurs to you. That’s the scheme of this. So you walk into a room and you see something misplaced, you may then place it right. You see another thing that belongs at the other end of your house, but you are committed now to working on that job proposal now, so while you could improve where that other thing is, you don’t have the time now so you don’t now. It’s about paying attention to what’s immediate.

[Seraphim]

That works if each improvement can be completed quickly, but it has never been my experience that things are so uniform. Some of them become time sinks, and you can't always tell ahead of time which ones they will be.

If you happen to see something that needs improving, but it ends up taking a long time, everything else can languish. This can be very disruptive.

I think the principle of trusting your intuition to make fast choices is very valuable. But the choices still need to be structured and/or presented in such a way that the overall flow of work is not disrupted.

[Alan Baljeu]

Yes I agree with these points, which is why it seemed to me necessary to have the options. If the thing you observe is going to take too long, it’s best to plan to tackle it at a later date. To fit the scheme, I would say that creating a plan to improve the thing, is an improvement, provided of course there will be a future execution of that plan. I mean either a creating a plan, or setting a time, or establishing a reminder so you will get to it at a later time.

Or else to apply little and often, and just do a little now and a little the next time it shows up.

In these ways I think the risk of disruption is mitigated. Yet it still feels imbalanced overall, which is why I never followed through on writing about this before.

[Mark Forster]

A couple of relevant quotes:

‘If you're good at course correcting, being wrong may be less costly than you think, whereas being slow is going to be expensive for sure.’ Jeff Bezos

‘The startups that do things slowly don't do them any better. Just slower.’ Paul Graham (founder of Y-Combinator)

[Alan Baljeu]

My previous post was terribly written, so here's a rewrite:

Seraphim, I agree with your point that many things can't be improved in short order. This puts to test my dictum "If you can improve it now, do so". I would accommodate such cases by allowing deferred action. Instead of completing the work at this moment, make a plan to complete it. To fit my scheme, I would say that "create a plan to improve the thing", is a legitimate improvement to the status quo, provided of course the plan will actually be executed down the road. (Plans you never implement are generally waste and not improving anything.)

So when a thing comes up that needs addressing that you can't complete now, there are other improvements possible:

- creating a plan now
- put it in your calendar
- create a reminder (not necessarily time-based)
- apply little and often, and improve just a little now.

In these ways I think the risk of disruption by a long detour interrupting your flow of work is mitigated.

------------
New thoughts below.

My scheme is nothing more than the stand-out method of task picking but without a list.
I like the idea, yet I feel it remains too simplistic to work as a complete system. My current thought is to make this the secondary process.

The primary process [whatever it might be] would take care of the very few Big Important Tasks, while the undirected "improve what stands out now" would efficiently take care of all the myriad little tasks, whose exact plan of execution is immaterial. This would make managing the big stuff easier because you manage only them, and would make handling the little stuff easier because that's simplified to just "pressing the button" without the decision overhead or task management overhead.