Introduction
Gesture Research
Goal 
Have 40 participants in each of 9 countries to create gestures for 28 actions
Challenge #1
Ensuring repeatability
Challenge #2
Recording gestures
Gestures are used in an ever increasing number of products
Some gestures vary by culture
Challenge #3
Classifying consistently
Series double tap; curve around object
1. Analyzed the top-generated gestures, and the very infrequent gestures
Sorted gestures created for each action by frequency and statistically determined the top-generated gesture(s) and the very infrequent gestures

2. Analyzed for cultural differences - conducted 3 primary analyses
Did one country use a particular gesture more than another?
Did one country use a particular gesture type (direct manipulation vs symbolic) more than another?
On visual inspection, did a pattern of responses emerge?

3. Analyzed for Novice vs Experience differences - 2 analyses
Computed agreement scores for Novices and Experienced participants
On visual inspection, did a pattern of responses emerge?

Dan Mauney
Method
Data Analysis
Results
Results
1. Analyzed the top-generated gestures, and the very infrequent gestures

2. Analyzed for cultural differences - conducted 3 primary analyses
On visual inspection, did a pattern of responses emerge?

3. Analyzed for Novice vs Exerience differences - 2 analyses
On visual inspection, did a pattern of responses emerge?
There was a higher agreement on gestures for actions that could be performed through direct manipulation and lower agreement on gestures for actions that were more abstract in nature
Discussion
1. Most of the variability in participant responses was not based on culture
China's use of symbolic gestures was the primary exception to this finding
Good news for designers of global products

2. Back, Forward, Scroll Up, and Scroll Down exhibited the largest pattern of differences
that is most likely due to experience with finger-driven touchscreen interfaces
Few other differences found between Novices and Experienced participants

3. When a gesture to elicit an action was not immediately identifyable, participants
generally wanted to tap on the screen to bring up a menu

4. For actions with no clear top gesture other than tap, a detailed examination of this dataset would be a strong first step toward identifying an appropriate gesture. 
One pattern we noticed when examining these ourselves is that participants often used the first letter of the action or drew the familiar windows icon when trying to create a shortcut
Potential Follow-on Research
Take the top gestures generated by participants and provide 
them to a new group of participants in a survey as multiple 
choice options and request a large number of participants 
in each country to select the gesture that makes the most 
sense to them

Once participants have made all selections, allow them to go back and change previous answers after considering the gesture set as a whole. Allow iteration until participant feels the set is ideal.
Checkmark
Swipe Object
Multi-Tap (anywhere)
Swipe down
Swipe Right
Letter/word other than X
Swipe/drag to neutral space
Swipe across object
X
Letter/word other than X
Question mark
Swipe
Curve or Circle CCW
Swipe right
Spread
Swipe up
Swipe diagonal
Pinch
Swipe down
Swipe object
Tap object sequentially
Tap multiple objects simult.
Carat or arrow betw. words
Drag/swipe to new location
Drag object to neutral space
Letter/word other than X
Curve or circle CW
Swipe right
Curve or circle CW
Curve around object
Sim press & hold & swipe
Check Mark
Letter/word other than X
X
Letter/word other than X
Swipe object
X on Object
Belgium * Chile * China * Finland
France * Germany * India * Italy 
Spain * Sweden * UK * USA
3. Analyzed for Novice vs Experience differences - 2 analyses
Computed agreement scores for Novices and Experienced participants
Sylvia Le Hong
Delete object: % of participants 
performing an x on the to-be-deleted items
Research Questions
• Common gestures for familiar software actions? 

• Cultural differences in the use of gestures? 
 
• Differences between experts and novices?

A cross-cultural gesture set
Discussion
Gesture set
This study aims at helping designers understand how users expect to interact with a gesture-enabled touchscreen.

Therefore, we took an initial approach to suggest a gesture set for the 28 investigated actions ... with a few caveats:

Gestures were collected independently
Gestures are interdependent and need to be evaluated as a complete set
A lo-fi prototype was used to collect data >> no system feedback >> no learning and adaptation from users >> but intuitive behaviour unaffected by the system's response 


Accept / Verify
Menu:
Press & hold object
Shortcut: 
Letter C anywhere
Back
Swipe right anywhere
Close folder
Tap object
Continuous scroll down
Swipe up repeatedly anywhere
Copy
Delete
Cut
Check mark on object
Menu:
Press & hold object
Shortcut: 
Flick object
X on object
Swipe left anywhere
Forward 
Minimize
Home
Swipe diagonal
Question mark anywhere
Menu:
Press & hold anywhere
Shortcut: 
Tap anywhere with 3 fingers
Magnify
Spread from object
Help
Multi select
Tap object
Press & hold object then swipe 
across multiple objects
Drag / Swipe object
Press & hold object
Move object
Open menu
Open folder
Redo
Press & hold anywhere and 
curve / circle clockwise
Press & hold object 
and swipe any direction
Menu:
Press & hold object
Shortcut: 
Letter P anywhere
Menu:
Press & hold anywhere
Shortcut: 
Carat
Pitch
Print
Paste
Scroll down
Menu:
Press & hold anywhere
Shortcut: 
Letter S anywhere
Swipe down anywhere
Swipe up anywhere
Scroll up
Save
Rotate image
Curve / circle clockwise
on object
Zoom in
Press & hold anywhere
and curve / circle counter
clockwise
Tap anywhere
Spread from object
Pinch with any # of fingers
on object
Undo
Zoom out
Stop scroll
Sorting Gestures
2. Analyzed for cultural differences - conducted 3 primary analyses
Did one country use a particular gesture more than another?
-->For the top 38 gestures (used by more than 40 people), there was no significant difference across country in their use of those gestures

Did one country use a particular gesture type (direct manipulation vs symbolic) more than another?
-->Country was significant at the p<0.001 level

Thanks for your Attention!
For further details, contact:
Sylvia Le Hong - sylvia.lehong@uid.com
Dan Mauney - dmauney@humancentric.com

Follow our blog at: www.touchthinking.com

View our Prezi at: http://prezi.com/ycecrqkvlj7i/