Scribd
Upload
89 views

Single VS. Multiple Textbox Event Creation Experiment

Presentation slides summing up paper documenting an experiment of creating events through an event management web application. Is Natural Language Processing (NLP) a better choice to specific requested text fields? Two designs of information input were tested: a multiple textbox fields and a single textbox. Users completed the event creation process faster and with fewer errors using the single textbox design.

Uploaded by

Filip Mares
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
89 views

Single VS. Multiple Textbox Event Creation Experiment

Presentation slides summing up paper documenting an experiment of creating events through an event management web application. Is Natural Language Processing (NLP) a better choice to specific requested text fields? Two designs of information input were tested: a multiple textbox fields and a single textbox. Users completed the event creation process faster and with fewer errors using the single textbox design.

Uploaded by

Filip Mares
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 26

Single VS.

Multiple
Textbox Event
Creation Experiment
Filip Mares – TIM Carleton University
Outline

 Background
 Experiment
 Method
 Analysis
 Future Experiments
 Conclusion
Twegather

Create an  Started in Summer 2009


event with
3 mouse  Joint project for TeamCamp
clicks
members
 Team of 5 active members
 Close to a Minimum Viable
Product (MVP)
Difference from competition
 Simple event creation process
 Registration not required for creating
 RSVP via Social Networks
 Twitter
 Facebook (soon)

 Free
Experiment
Experiment
 Natural Process
 Increase event creation
 Reduce creation process difficulty
 Re-design front page
Hypotheses
 H1: Participants will create events quicker on
average using the proposed page design
opposed to the current page.
 H2: Participants will make fewer mistakes when
creating events using the current design
opposed to the proposed design.
 H3: Participants will prefer the proposed
Twegather event creation page to the existing
event creation page design.
Method
 A/B test
 Design A – Current
 Design B – Proposed

 8 participants
 Recorded screen with software
 Pre-thought event
Metrics
 Completion Time
 Errors Encountered
 Questionnaire
 LIKERT

 Designs Preference
 Designs Difficulty

 Comments
Hypothesis 1
Participants will create events quicker on
average using the proposed page design
opposed to the current page.
Completion Time

Design A Design B

 Mean: 41.5  Mean: 22.75


seconds seconds
 SD: 13.6 seconds  SD: 7.28 seconds
Completion Time
 P Value: 0.0091
 Group comparison:
 Design A pval: 0.779
 Design B pval: 0.7648

 Hypothesis Verified
 Design A is faster to
complete
Hypothesis 2
Participants will make fewer mistakes
when creating events using the current
design opposed to the proposed design.
Errors Encountered

Design A Design B

 Mean: 1.38 errors  Mean: 0.625


 SD: 0.52 errors errors
 SD: 0.625 errors
Errors Encountered
 P Value: 0.0199
 Group Comparison:
 Group A pval:
0.6376
 Group B pval: 0.39

 Hypothesis
rejected
 Design B had less
errors
Hypothesis 3
Participants will prefer the proposed Twegather event
creation page to the existing event creation page
design.
Design Preference

Design A Design B

 Mean: 3.5/5  Mean: 3.63/5


rating Rating
 SD: 0.93  SD: 1.19

Note: (really disliked) 1 – 5 (really liked)


Design Preference
 P value: 0.8565
 Group Comparison:
 Group A pval: 1
 Group B pval: 0.788

 Hypothesis
rejected
 Therewere no
obvious preferred
design
Threats to Validity
 Typing skill level
 Internet connection
 Test Environment
Participant Feedback
 Single textbox hard at 1st, easier
afterwards
 Multiple textbox requires less thinking
 Multiple textbox followed order of other
services
 Good parsing of regular phrases in single
textbox
Future Experiments
 Event syntax auto-completion vs.
preview
 Event creation workflow
Conclusion
 Event creation completed faster with
single textbox
 Less errors encountered in single
textbox design
 Participants did not have a strong
favorite design
Thank You
Questions?

You might also like