Blog #1: Imaginary interfaces

Title: Imaginary Interfaces

Authors: Sean Gustafson, Daniel Bierwirth and Patrick Baudisch from Hasso Plattner Institute in Potsdam, Germany.

Summary:

For User Study #1:  BASIC SHAPES AND DRAWINGS

Hypothesis: 

The purpose of this study was to test if Imaginary Interfaces are subject to the difficulties in connecting stroked reported by Ni and Baudisch. The author’s Hypothesis was that participants would perform fewer Graffiti recognition errors than reported by Ni and Baudisc. (Ni and Baudisc explained the idea of Disappearing mobile devices).

Methods:

Participants’ task was to reproduce a series of predefined sketches. Each sketch  required participants to align or connect a stroke to a stroke drawn earlier.  

The trials were split into three tasks:

I) Graffiti – The participant would draw seven letter characters, URDGPO because these leters are more difficult to draw eyes-free.

 2) Repeated drawing – Participants were to draw a simple triangle and square repeatedly over each shape none stop till they drew each shape five times.

3) Multi-stroke drawing- The participants were to  draw a set of five simple sketches  each of which involved multiple drawing strokes. For example: Drawing a curve that slides up to a peak then down into a trough, then being asked to circle either the peak or trough to exact precision in the imaginary space.

Results:  they found for the Graffiti trials only 5.5% of the text entrees were not recognizable. Ni and Baudisc predictions  were 15% of text entries were not recognizable. Ni and Baudisch does not apply to Imaginary Interfaces, so there truly isn’t enough information to suggest that they were wrong. But the data supports the authors’ hypothesis that visuospatial memory helps fill in the blanks.

For the repeated drawing of shapes, Overall the average error from previous for the diamond was 2.20cm (s=0.90cm) and for the triangle was 3.25cm (s=2.31cm). The decreasing error distance from the previous suggests that participants built up visuospatial memory of the shape with repetition and the fact that users kept drifting away from the first suggests that later loops overwrite earlier one.

For the multi-stroke drawing task the alignment between strokes seems to decrease with the complexity of the sketch. The individual letters of the ABC string are well drawn; participants, however, condensed whitespace, causing letters to overlap. Relative scale appears reasonably correct. Misalignment appears to be caused mostly by translation errors derived from choosing the wrong starting point.

For User Study #2: RETURNING TO A DRAWN FEATURE

Hypothesis: 

The Authors wanted to understand what causes visuospatial memory to fade when using Imaginary Interfaces. There hypothesis was that “that intermitted actions affect visual memory and thus reduce accuracy”.

Methods:

Participants started with their hands at their side, and then are shown a simple glyph on the computer. Each glyph consisted of 4 strokes in order to be able to draw it. Then Participants clicked a lever with their foot and replicated the glyph with the dominant hand, while using their none-dominant as the plane reference. Then the participant would see another glyph and draw without their reference hand.

Part 2 of the experiment would have the participant rotate 90 degrees after seeing the glyph on the screen, then be able to draw the glyph in there imaginary space. The authors hypothesized that this part would increase the errors drawing the glyphs. The same tasks of using a reference hand and none, applied to the turning.

Results:

They made an ANOVA data sheet, and found The  hand condition had an average error of 5.4cm (s=0.4cm) and the none condition had an average of 6.3cm (s=0.3cm). Showing that having a reference had made a significant difference. Also, the  stay  condition (mean=5.1cm, s=0.3cm) had significantly less (F1,11=16.007,  p=0.002) average error than the rotate condition (mean=6.6cm, s=0.4cm).

So the Authors hypotheses for both methods were supported. And found that forcing the participant to turn 90degrees, created significantly higher error rates. However, the effect was significantly lessened by using the reference hand.

User Study # 3 POINTING BASED ON COORDINATES

Hypotheses: 

The Authors wanted understand what caused visuospatial memory to fade when using Imaginary Interfaces. There hypotheses was that intermitted actions affect visual memory and thus reduce accuracy.

Methods:

For each trial, participants started in a neutral position with their hands held loosely at their sides. Participants received the target location as two digits via audio and displayed on a monitor. They pressed a footswitch which started the timer and started the trial. Participants then raised their hands, formed an ‘L’ gesture with their left hand and acquired the respective target by pinching with their right hand. Participants committed the acquisition by pressing a footswitch again. This recorded the 3D location and rotation of the body and both hands, played a confirmation sound and completed the trial. In preparation for the next trial, participants dropped their arms and rotated approximately 90° to simulate mobile use.

Results: 

The Authors expected that error would increase with distance from the two main landmarks, i.e., the fingertips of the ‘L’ hand. As expected the fingertips were the most accurate locations, with button sizes of 0.35 thumbs wide and 0.21 index fingers high. As the distance increased from the nearest finger tip error increased linearly.

Discussion: 

Imaginary Interfaces seem to be a interesting technology in the future.  If researchers are able to prefect the technology, I can see people being able to do simple tasks without the aid of a visible screen. The idea of having screen-less mobile devises certainly would miniaturize that technology to its lowest level. I watched a program a few weeks back on history channel, exploring the possibility of having the internet literally wired into our brain. Imagine the possibilities if we could communicate via internet without having a phone or computer. Using imaginary interfaces may very well be that next step in the communication boom that this generation is going through.