Category Archives: spatial

The Spatial Demands of Assembly Instructions

I purchased a “build it yourself” cabinet over the internet, naively thinking that it couldn’t be too complicated if they really expected me to assemble it. Boy, was I wrong.

Take these instructions for example:

“Tap in post with a hammer to start and screw down with a screwdriver or with a small allen wrench until the shoulder of the post hits the wood. Then unscrew until the hole is parallel with the groove on the bottom of the Drawer Front.”

  1. I don’t know what a shoulder is.
  2. When is a hole parallel to a groove? Is it when the hole is facing the groove making them both “face” the same way,  or when it is going along in the same direction as the groove? In retrospect this seems more obvious than it did at the time.

Next, I’ve posted before how “bugs”  are different than human factors issues. I would like to add that minor errors in the instructions for a very complex spatial task can make it nearly impossible. For example, the instructions end with “To understand how the bastion fastening system works see next page.”  I finally did find that explanation two pages previous to the current page, at which point I’d forgotten what I was looking for.

Last, hardware is referred to by size and a number. For example, a diagram would have a picture of a screw going into a board and have the callout “1 1/4 #8.” Because there are multiple screws included that are 1 1/4″ long, I needed to reference the second page of instructions that listed all the hardware. On the hardware page are pictures and diagrams of the tops and side views of the hardware with the label below. Unfortunately, the label is equidistant from the diagram above it and below it, so I had to keep reminding myself by looking at the top screw picture to see if the label was for the graphic above or below. On this list the name is followed by a number in ( ), which seemed to be the number included in the package. So, back to my search for the 1 1/4 #8, it turns out there are two different 1 1/4″ screws, two different Phillips Flat Head #8s, and one of those happened to have (8) in the package. At this point the conjunction search became too much for me and I had to take a break.

On returning, I decided to ignore the front of the instructions that told me “Assembly is very easy if you read and follow the instructions step by step.” Once I just started putting it together using the overall picture and clamps and screws from my own workshop it went much faster. I had a few instances where I needed to backtrack, but overall it a successful assembly. Was this the triumph of a good mental model over step by step instruction or do I just lack the mental resources to follow steps that depend on referencing other steps and pages?

I suppose I should also mention what was well done in the instructions. The pictures of each screw with the difficult labels included: 1) a top and side view 2) a box drawn to scale (so you didn’t need a ruler to measure if the screw you were holding was 1 1/4″). The instructions also divided the steps into small enough independent pieces on each page that they were not overwhelming. Good job on that!

3D is better than 2D, right?

It seems that every few years, 3D technology is in the zeitgeist (with 3d movies).  User interfaces are not immune to the frenzy of 3D.  However, there is quite a bit of past research in 3D interfaces (I won’t even scratch the surface but see this simple Google Scholar search to start). Much, but not all, relate to navigation in virtual environments, while other research relates to the inclusion of use of depth/perspective.  There are still many outstanding issues in the use of 3d in user interfaces; some of which are: use/interaction (input, output), effects on workload, and effects on learning.

In general, 3-dimensional displays (like a perspective view) are perceived to be more natural and possibly require less mental integration than 2-dimensional displays (see this very well-researched U.S./FAA report on multifunction displays; warning PDF).  Some of the logic goes like this:  when I view a 2d map, I usually turn it into a 3d representation in my head.  Showing a 3d representation removes this step (in addition to showing more information).  Compare the two types of information displays:

2d and 3d representations of a hiking trail
2d and 3d representations of a hiking trail

These images come from a user study examining user preferences in map presentation (2d or 3d). The research showed that it depends.  The preference data was very complex (see paper) but the preferences were evenly split but those aged 26-40 preferred 3d maps. Males preferred 2d maps while females preferred 3d maps (which seems surprising).

Personally, I switch between 2d and 3d view when I can because each offers information the other does not. I like to examine hikes after the fact (collecting and mapping GPS data). See below; each view gives you different information:

Left: 2d, top-down view. Right: 3d, perspective view

The 2d view gives a good general overview and the intricacies of the trail but shows no elevation information while the 3d view shows terrain but obscures the path (part of it is hidden behind the terrain).

More subtle uses of 3D on websites is the use of the parallax for the illusion of depth. This website showcases some creative uses of this effect. Most websites use it for aesthetic effect, however, I noticed the new Google Nexus One phone uses it in a subtle but useful way to indicate that you are on a different screen (a type of low-level feedback). See the video below. When the user slides to another screen horizontally, the animation of the galaxy changes perspective:

Notice the slight perspective shift of the galaxy background
Notice the slight perspective shift of the galaxy background after the user swiped the screen

Embedded video (skip to 61 seconds in):

In some cases, when used appropriately, learning can be enhanced by the use of 3d.  Researchers Avi Parush and Dafna Berman (in a 2004 paper in the International Journal of Human-Computer Studies) were interested in the use of 3D interfaces for navigation and orientation in a virtual environment.  The virtual environment contained the objects that one would normally have on their computer desktop (e.g., files, applications).  Will the use of a 3d environment enhance learning and performance?  They manipulated whether subjects had two kinds of aids to help: landmarks or a route list.  They found that both types of aids help in the learning process but a crucial point was that landmark placement (in either 2d or 3d) was critical.

Example of the 3d condition (left, no landmarks; right, landmarks). From Parush & Berman (2004)

One commercially available tool that gives users this kind of view of their computer is the BumpTop desktop:

BumpTop Desktop

The BumpTop desktop introduces the further complication (on top of 3d) of the nature of interaction.  You are using a 2d surface (the touch pad or mouse) to navigate the 3d environment and in some cases using multi-touch gestures (using more than 1 finger).  Very cool…but useful?  See for yourself:

(post image: