Free form notes on Life Balance.

Bias against Importance ratings and anything with a slider

I have a perhaps an irrational bias against presenting people with "scales, dials, meters and sliders" to "measure" things. Most people can't even estimate the length of a room (certainly not me). Trying to measure such things as "importance" and "effort" on a blank scale with no "concrete points of reference" is simply too hard and as a result irrelevant for a lot of people. Whenever someone asks me, what's the priority of this item? how urgent is this item? how important is this item? I start to hyperventilate and then I think to myself, with respect to what?

Measurements are important, but people do much better making "relative" assessments of importance then they do nudging a slider around trying to make "abstract, absolute" judgements.

In a list of 10 tasks, in order for someone to record the importance of say task# 4 on the slider, they have to remember what they put down for 1,2, and 3 and hope that 5,6,7,8,9, and 10 don't turn out to be so monumentally more or less important that you have to reconfigure everything according to a different scale. This is why some teachers grade on a curve, effectively rank ordering their students rather than trying to figure out some kind of absolute standard for As, Bs, Cs, Ds and Fs. It's the same problem Olympic judges face. How do you know what to give the first gymnast, when you haven't seen anybody else? A novice judge might be really impressed with a mediocre gymnast and give her a 9.8 and not realize she was mediocre until she sees the rest. At which point, it's too late because she has too little room above 9.8 to reflect how much better the subsequent gymnasts are.

Fortunately, novices aren't allowed to judge the Olympics and even the seasoned judges at the Olympics are given strict guidelines for how to give and deduct points. The question is, do people have or want such rigorous measurement standards for their task list?

Need for concrete reference points when asking users to rate items

The use of very generic terms (ie. none, below average, average, above average, maximum) to describe something like Importance or Effort is again asking users to come up with their own "reference points" for each "level of effort" which might change from project to project. What's maximum? 3 days seems like the maximum I would to spend on reorganizing my sock drawer, 30 days might be my maximum for teaching my kid how to ride a bike. More concrete levels such as: minutes, hours, days, weeks, months, years, lifetime would both make more sense and be more consistent across projects. The levels need to be questions users can answer with a definitive Yes or No. Will this task take an "average" amount of effort is too squishy, subjective and subject to context to be useful.

Lastly, all of these sliders feel like an annoying level of indirection that forces the user to manipulate their data via some 3rd party widget.

Why not allow users to just grab their items and explicitly order them with respect to importance?

More important, I'm not sure Life balance helps people answer the following questions:

1. How do you figure out what you should be focusing on right now?
2. How do you figure out if what you've decided you should be focusing on right now is right set of things?
3. How do you figure out if you can take on a project in 2 weeks?

Both the Outline view (which lists all of your tasks in an outline grouped by project) and the ToDo^? list (which lists all of the Tasks) are ordered with respect to "importance". Both lack the ability for the user to focus:

1. Just on things I'm should be doing now
2. Just on things I've deferred until later

There's no clear delineation between between Now and Later. Instead, it's simply a continuous flow of potentially hundreds of items from most to least important. If we know people are overloaded with trying to do too much all at once, how does this continuous help people draw a more digital line between what they can handle at this moment and what needs to wait until later?

The system also assumes that Importance is what people care about most. (It's clear from above that I don't even believe it's fair to even ask people to measure importance on some absolute scale.) Just because something is important doesn't necessarily mean I should work on it now. Planning my wedding might be important, but if I a status report due in 5 minutes and my wedding is scheduled for May 2006, I should probably do my status report, not call the florist.

There's some reprieve...or is there?

In the Outline view, you can use a checkbox in the detail view of any item and specify that you'd like to do "All subtasks in order". This feature is presumably meant to allow users to hide "irrelevant until later" subtasks so they can focus just on next actions. However, this still doesn't address a few issues:

1. Again, it's an indirect way for users to do something they should be allowed to do directly by dragging the item around in the summary view. It's so indirect that I'm not even sure users would necessarily think to use that feature as a way to focus on "Now" items. Instead, it looks more like yet another attribute the computer is asking you to specify that might or might not be helpful to you. Maybe the check box should say: Hide items you don't want to deal with right now.

2. It's also not clear that users would use it to hide "deferred until later" items because I will bet that most of time not all substasks need to be completed in order. Or a different way to put it would be that "the need to perform subtasks in order" does NOT necessarily equal "tasks are deferred until later". Users may want to defer tasks for myriad other reasons:

1. Not that important
2. Dependent on another item that the user wouldn't think to enter into the task list (ie. when my landlord finally fixes my refrigerator, I'll pay the rent)
3. Can't deal with right now and not relevant for another 2 weeks

The question is: What do I do if for some project, subtasks 2,4 and 6 are dependent on each other but 1,3 and 5 aren't? And what if I need and can do tasks 2,3 and 4 right now? How do I keep those sub-tasks in view and get rid of the others?

There's also no way to say that an entire Project is deferred until Later.

Comments

MimiYin 20050425 I think the big philosophical difference I see behind Chandler and LifeBalance^? is not the substance of WHAT you can do, but HOW you do it. Yes you can use both designs to display data wrt NOW v. LATER, but LifeBalance^? follows an "application-centric" model for tweaking the display of data whereas Chandler is aiming for a more explicit, user-controlled experience for manipulating data views.

In LifeBalance^?, users indirectly affect how their data is displayed by: 1. twiddling dials and knobs on individual items and then the 2. application aggregates all of the dial and knob settings and 3. pumps it through application-defined rules to 4. come up with a data display BASED ON the user-defined per item dial and knob settings

In Chandler, users directly control how their data is displayed by: 1. dragging and dropping their data in the view

Chandler philosophy can be summed up as: 1. Primary interaction should be: direct, explicit manipulation of how users see things in the UI (ordering, hiding and showing of items).

2. Less dependence on the system (app) to do the right thing automatically based on system rules that were not set by the user.

3. Less dependence on data inputs from the user (ie. dial and knob settings on individual items).

All of this comes from a fundamental belief that PIMS and information managers in general, should be more of a complementary aid to human memory, processing and organization and less of a "replacement repository" for everything in your head. The belief stems from:

1. Disbelief that people will be disciplined about writing everything down 2. Belief that more often than not, people can't articulate in words, what's in their head 3. Belief that users don't need "replacement repositories". Replacement repositories take too much work. Instead, users need systems that a) allow them to quickly record "memory ticklers" (just enough information to help you find information buried in your brain) and b) directly manipulate how and when they view these "memory ticklers".

=

A different way to look at this issue is that LifeBalance^? is that when an user wants to "Defer a task" (aka Get it out of my face for now...), LifeBalance^? asks, why do you want to do that? And the user must respond, "...because this task is not due for 2 weeks and I only need 1 day of lead time before it's due."

Chandler is less interested in the WHY an user might want to get something "Out of their face". There are too many reasons to model and even if we could model all of them, users either don't often know why and/or aren't interested in recording why.

I want this item out of my face...even though it's due today...because I'm mad at the person who assigned me the task...because other things that are more important came up and I can't deal and frankly, I don't know when I'll get to it again. OR The user could have a 2-minute task that is due in a week, but they just want to keep it around, just because.

[As a result, users might not make the connection between the systems notion of WHY they might want to "Get something out of their" face and their desire to "Get something out of their face". The user is still left wondering, how do I get this item out of my face?]

The reasons are too many, too tangled and too irrational for the system to try and model. Instead, it is safer to simply provide users with the explicit affordance to "Get something out of my face" and not worry too much about the WHY.

Furthermore, reasons change second to second as people receive more and more data inputs via email. I can see how LifeBalance^?'s system works within a task manager, but once you factor in email, I think it's simply too onerous for users to be constantly trying to control their view of data by tweaking settings on individual items.

(This also relates to the belief that people are not uniformly rational enough to be trusted with absolute measurement devices such as importance and priority ratings.)

Again, I agree that you can simulate a lot of the behavior I'm describing by checking the right boxes and setting the right dates. But that's sort of the point. Chandler is trying to shift away from "you fill out the attribute fields and the system will figure out how to display your information BASED ON the data you put in..." TO "here's you're data, tell the system directly how and when you want to see it".