Virtual Keyboard Enhancement

Role

UX Researcher & UI Designer

Responsibility

• Prioritized the main finds

• Prepared question scripts

• Conducted the observation & interviews

• Delivered the document, wireframes & high-level mock-ups

Deliverable

Insights and Recommendation Deck




The keyboard is an essential part of any smartphone experience. However, the ordinary and third-party keyboards on smartphones are generally flawed. When I was working at Foxconn, we discovered that the third-party keyboard was not as accurate as we had hoped, and clients were not satisfied. So we decided to investigate the problem and make suggestions for improvement. We advised the vendor based on the experiment’s findings and helped them integrate the improvement into the virtual keypad, which not only improved accuracy by 10% but also led to the acquisition of a patent.

Experimental Programs
Participants

There were typing interfaces, which were each designed with different layouts and touch areas. Twenty-four participants typed an alphabet letter randomly shown on their screen. Each subject typed a total of seventy-eight letters in each experiment.

testing programs
Testing programs
Participants
Participants
Term definition for the experiment
Participants
Term and definition for the experiment
Research

We analyzed several virtual keypads from Android and iOS platforms and discovered that the individual touchable area and key dimension might be the most important factors in improving accuracy. We began to construct by building the hypothesis from the touchable area.

Research Process
Research Process
First experiment
Hypothesis

A larger touchable area would give the user more space while typing the keypad. More room for users, thus allowing users to touch accuracy.

Prediction
If the text printing is moved upward and the key’s height is decreased, then users will shift downward and concentrate on the proper place. Test

{···} - Variable – touchable area


We created two prototypes and put them into the programs, one with a touchable area that was the same size as a single key and the other significantly bigger. The testing prompts an English letter of the alphabet each time, and the user has to enter in accordance with the English letter prompted, with the next English letter appearing when the input is complete. Each test program has 78 English letters (26 English letters x3).

Data Analysis
Data Analysis
Result

We didn’t detect any improvement from group B in the initial round of testing, although we did discover some trends.


Hardly touch the top area*

Right-handed users tend to press to the right and lower down when striking the target key.

Tend to touch the right and lower right

Users sometimes incorrectly strike the intended key because they attempt to avoid striking the space between adjacent keys

Think of outside margins on each row as the hit area

Users sometimes intuitively believe that the hit area includes the two outside margins of the first and last key of each row

*Left-handed users tend to press to the left and lower left when striking the target key.
Second experiment
Hypothesis

Because of the user’s perspective, when they are using a virtual keypad, they tend to press the low position of the key to read the letter.

perspective
User Perspective
Prediction
If the text printing is moved upward and the key’s height is decreased, then users will shift downward and concentrate on the proper place. Test

{···} - Variable


letter placement


touchable area


extend the touch area of two end


image
Enlarge and lower to the right and low-right
image
Moved letter placement higher up on the physical key
image
Enlarge the hit rate area of the first and last keys on each row
Result

The second experiment validated our assumptions, raising the accuracy by 8%. However, even though decreasing the height improves accuracy, it might cause the user to feel uncomfortable when using it.



Position deviation

The user’s point of view affects the touch deviation of position.


Printing location

Because of the printing location, users tend to press the lower and lower-right area


Expended area

Utilizing the two-end area as a touchable area to increase the accuracy rate


image User also said….

I’m afraid to press the key because the vertical dimension is too small.

image
Data analysis
Final experiment
Hypothesis

In the second experiment, the height of the key is too short, and the user feels uneasy to use the keyboard, so the increasing height of the key will help to fix this issue.

Prediction
If each key’s height increases, then users feel more confident in using it.
Test

{···} Variable – the height of the key


Reult

The third trial confirmed our hypothesis, which was that lowering the height of the key was beneficial. It boosts users’ confidence when using the keyboard and improves accuracy by 9.79%.

Data Analysis
Data Analysis
Share Insights

Created the presentation to share the outcome and recommendations with the teams and vendors.

final deck
Final Deck
Takeaway

Even though the hit-rate accuracy improved by 10%, as a result, it was just for the right-hand user. We believe that we can use an accelerometer to detect input and change the display of each key on the virtual keyboard to offer an appropriate contact area on the virtual keyboard.

Next Step

Work with the vendor to prioritize the work to introduce the virtual keypad

Analyze the visuals and layout of the keyboard

Apply for a patent

Next Case Study.

Image

Power Management System

How I developed proprietary Solar’s Performance Management System while working with Formosa

UX, UI Design

Copyright ©. All Rights Reserved.

Background Research 1st Experiment 2ne Experiment Final Experiment Takeaway