Substance or Representation? An Observational Investigation of Criteria for Recompense Winning Sites.


73 views
Uploaded on:
Category: Education / Career
Description
Substance or Illustrations? An Experimental Examination of Criteria for Honor Winning Sites Rashmi Sinha, Marti Hearst and Song Ivory College of California, Berkeley Maya Draisin Universal Foundation of Advanced Expressions and Sciences For more data: http://sims.berkeley.edu/~sinha/WebbyAwards
Transcripts
Slide 1

Substance or Graphics? An Empirical Analysis of Criteria for Award-Winning Websites Rashmi Sinha, Marti Hearst & Melody Ivory University of California, Berkeley Maya Draisin International Academy of Digital Arts and Sciences For more data: http://sims.berkeley.edu/~sinha/WebbyAwards

Slide 2

Question : How would we be able to recognize attributes of good sites on a vast scale? Question : How would we be able to transform these attributes into experimentally accepted rules?

Slide 3

Find a corpus of sites effectively recognized as great! Utilize the WebbyAwards database Conduct Usability Studies: Hard to do on a huge scale

Slide 4

Talk Plan Details about WebbyAwards 2000 Qualities of exceptionally appraised sites Relative significance of Content & Graphics Empirically approved configuration rules Future Plans: WebbyAwards 2001, client concentrates on and so on

Slide 5

Criteria for accommodation to the WebbyAwards Anyone who has a present, live site Should be open to the overall population Should be overwhelmingly in English No breaking point to the quantity of sections that every individual can make 7 th Conference on Human Factors and the Web

Slide 6

Site Category Sites must fit into no less than one of 27 classifications. For instance: Arts Activism Fashion Health News Radio Sports Music News Personal Websites Travel Weird Sites can be recorded in various classes 7 th Conference on Human Factors and the Web

Slide 7

3 Stage Judging Process Review Stage: From 3000 to 400 destinations 3 judges rate every webpage on 6 criteria, and make a choice on the off chance that it will go to the following stage Nominating Stage: From 400 to 135 locales 3 judges rate every website on 6 criteria, and make a choice in the event that it will go to the following stage 7 th Conference on Human Factors and the Web

Slide 8

Final Stage: From 135 to 27 locales Judges makes choice for most loved webpage People’s Voice: 135 to 27 destinations Anyone can vote in favor of their most loved website among the ones in the last stage 7 th Conference on Human Factors and the Web

Slide 9

Criteria for judging 6 criteria Overall Site Experience Five particular Criteria Content Structure and Navigation Visual Design Interactivity Functionality Scale : 1-10 (most noteworthy) 7 th Conference on Human Factors and the Web

Slide 10

Content : is the data gave on the website. Great substance is drawing in, important, proper for the gathering of people you can advise it\'s been produced for the Web in light of the fact that it\'s reasonable and succinct and it lives up to expectations in the medium … 7 th Conference on Human Factors and the Web

Slide 11

Structure & Navigation : is the association of data and route. Locales with great structure and route are steady and successful. They permit you to frame a mental model of the data gave … 7 th Conference on Human Factors and the Web

Slide 12

Visual Design : is the webpage\'s presence. Great visual configuration is high caliber, proper, and applicable for the group of onlookers and the message it is supporting … 7 th Conference on Human Factors and the Web

Slide 13

Interactivity : is the way a website permits a client to do something. Great intuitiveness is more than sound impacts, and a Flash activity. It permits the client to give and get. Its information/yield in ventures, visit rooms, ecommerce etc.… 7 th Conference on Human Factors and the Web

Slide 14

Functionality : is the utilization of innovation on the website. Great usefulness implies the webpage stacks rapidly, has live connections, and any new innovation utilized is utilitarian and pertinent for the target group … 7 th Conference on Human Factors and the Web

Slide 15

Overall Experience : The general experience incorporates substance, structure and route, visual configuration, usefulness, and intuitiveness, yet it additionally includes the intangibles that make one stay or leave… 7 th Conference on Human Factors and the Web

Slide 16

Webby Judges Internet experts who work with and on the web: new media writers, editors, web engineers, and other Internet experts have unmistakably self evident nature with the classification which they audit 7 th Conference on Human Factors and the Web

Slide 17

Review Stage : 3000 to 400 destinations, every website judged by 3 judges Nominating Stage : 400 to 135 locales, every locales judged by 3 judges Criteria for judging : Content Navigation Visual Design Functionality Interactivity Overall Experience Quick Review of Judging Process 7 th Conference on Human Factors and the Web

Slide 18

Are we testing from the entire scope of sites (great to awful) in the Review Stage? 7 th Conference on Human Factors and the Web

Slide 19

Review Stage: The entire scope of locales, great to terrible Mean = 6.01 SD = 1.59 Overall Rating 7 th Conference on Human Factors and the Web

Slide 20

Can votes be anticipated by particular criteria? Measurable Technique : Discriminant examination Question : Can we foresee the votes from the 5 particular criteria? Could general rating be anticipated by particular criteria? Measurable Technique : Regression investigation Question : What % fluctuation is clarified by 5 criteria Percentage change clarified = 89% Classification Accuracy for Sites = 91% 7 th Conference on Human Factors and the Web

Slide 21

What criteria contribute most to general rating? 7 th Conference on Human Factors and the Web

Slide 22

Nominating Stage Analysis 6 criteria Content, Structure & Navigation, Visual Design, Functionality & Interactivity Overall experience 400 locales 3 judges evaluated every webpage 7 th Conference on Human Factors and the Web

Slide 23

Nominating Stage: Top destinations for every classification Mean = 7.6 SD = 1.66 Overall Rating 7 th Conference on Human Factors and the Web

Slide 24

What criteria add to general rating at Nominating Stage? 77% change clarified in general rating 7 th Conference on Human Factors and the Web

Slide 25

Unique Contribution of Content and Visual Design People’s Voice Ratings likewise show that individuals vote in favor of destinations with preferred substance rather over better visual configuration 7 th Conference on Human Factors and the Web

Slide 26

Summary of Findings The particular appraisals do clarify general experience. The best indicator of general score is content . The second best indicator is intelligence . The most noticeably bad indicator is visual outline 7 th Conference on Human Factors and the Web

Slide 27

Are there contrasts between classes? Expressions Activism Fashion Health News Sports Music News Personal Websites Travel Focus on a couple of Categories Art, Commerce & Radio 7 th Conference on Human Factors and the Web

Slide 28

Art 7 th Conference on Human Factors and the Web

Slide 29

Commerce Sites 7 th Conference on Human Factors and the Web

Slide 30

Radio Sites 7 th Conference on Human Factors and the Web

Slide 31

Conclusions The significance of criteria shifts by class. Substance is by a long shot the best indicator of general site experience. Intelligence comes next. Visual Design does not have as much prescient force with the exception of in particular classes Can we create plan rules by distinguishing attributes of good site pages? 7 th Conference on Human Factors and the Web

Slide 32

Study 2 An exact base up way to deal with creating configuration rules Challenge: How to go use Webby criteria to illuminate website page outline? Answer: Identify quantitative measures that describe pages 7 th Conference on Human Factors and the Web

Slide 33

Quantitative Measures Page Composition words, connections, pictures, … Page Formatting textual styles, records, hues, … Overall Characteristics data & design quality 7 th Conference on Human Factors and the Web

Slide 34

Quantitative page measures Word Count Body Text % Emphasized Body Text % Text Cluster Count Link Count Page Size Graphic % Color Count Font Count 7 th Conference on Human Factors and the Web

Slide 35

Quantitative Measures: Word Count 7 th Conference on Human Factors and the Web

Slide 36

Quantitative Page Metrics Webby Ratings Model Accuracy Within Categories Across Categories Highly Rated Sites Top 33% 76% 67% Word Count, Body Text %, Text Cluster Count, Link Count and so forth. Low Rated Sites Bottom 33% 63% 83% Study Design 7 th Conference on Human Factors and the Web

Slide 37

Classification Accuracy Comparing Top versus base Accuracy higher for inside of classifications 7 th Conference on Human Factors and the Web

Slide 38

What page measurements anticipate website quality All measurements assumed a part However their part contrasted for different classifications of pages (little, medium & vast) Summary Across all pages in the specimen Good pages had essentially littler design rate Good pages had less accentuated body message Good pages had more hues (on content) 7 th Conference on Human Factors and the Web

Slide 39

Role of Metrics for Medium Pages (230 words by and large) Good medium pages Emphasize less of the body content Appear to arrange content into groups (e.g., records and shaded table areas) Use hues to recognize headings from body content Suggests that these pages Are simpler to output 7 th Conference on Human Factors and the Web

Slide 40

No Text Clustering No Selective Highlighting Low Rated Page 7 th Conference on Human Factors and the Web

Slide 41

High Rated Page Selective Highlighting Text Clustering 7 th Conference on Human Factors and the Web

Slide 42

Why does this methodology work? Shallow page measurements reflect more profound parts of data construction modeling, intelligence and so forth 7 th Conference on Human Factors and the Web

Slide 43

Future work Do comparable examination for Webby2001 Add more measurements Aspects of information, route, and visual depiction Category-based profiles Use bunching to recognize various types of good and poor locales These can be utilized to propose option plans Do client studies to confirm rules 7 th Conference on

Recommended
View more...