Social Interfaces
3 likes1,080 views
The document discusses three main challenges of crowdsourcing data: 1) crowdsourcing data, 2) human motivation, and 3) behavioral design. It explores theories of human motivation and how interface design can affect behavior. Specific examples are discussed, such as default choice bias and how interface elements in applications like Snackbot can influence user choices. The document advocates for social interfaces to draw from social and cognitive science regarding crowd data analysis, motivation, and designs that compel human participation and empowerment.
![[Haiti Earthquake Usahidi] Can unbounded crowdsourcing (non-
representative sampling) via SMS predict actual damage?
http://irevolution.net/2010/10/13/crowdsourced-prediction/
Tuesday, 7 May 13](https://image.slidesharecdn.com/mvk-2012-october-sociam-social-interfaces-130513114359-phpapp01/85/Social-Interfaces-12-320.jpg)
![8,163 copies of War
and Peace
(June 2011)
¡°Is It Really About Me? Message Content in Social
Awareness Streams¡± [Naaman et al. CSCW ¡¯10]
information sharing (IS) 0.2
self promotion (SP) 0.05
opinions/complaints (OC) 0.25
statements and random thoughts 0.25
me now (ME) 0.4
questions to followers (QF) 0.05
presence maintenance (PM) 0.05
anecdote by me (AM) 0.05
anecdote by others (AO) 0.01
Tuesday, 7 May 13](https://image.slidesharecdn.com/mvk-2012-october-sociam-social-interfaces-130513114359-phpapp01/85/Social-Interfaces-13-320.jpg)
![Formal Methods: Whence and Whither? [Martin Fr?nzle Festkolloquium, 2025]](https://cdn.slidesharecdn.com/ss_thumbnails/mf2025-250305164811-a0930761-thumbnail.jpg?width=560&fit=bounds)
More Related Content
Recently uploaded (20)
Featured (20)
Social Interfaces
- 1. SOCIALinterface smax van kleek sociam retreat pony land, new forest 1 october 2012 Tuesday, 7 May 13
- 2. 3 challenges 1 - crowdsourcing data 2 - human motivation 3 - behavioural design Tuesday, 7 May 13
- 3. 1 - crowdsourcing data 2 - human motivation 3 - behavioural design 3 challenges Tuesday, 7 May 13
- 4. Tuesday, 7 May 13
- 5. motivation is clear; Tuesday, 7 May 13
- 6. mumbai terror crisis motivation is clear; Tuesday, 7 May 13
- 7. mumbai terror crisis haiti damage report motivation is clear; Tuesday, 7 May 13
- 8. mumbai terror crisis haiti damage report japan nuclear crisis motivation is clear; Tuesday, 7 May 13
- 9. mumbai terror crisis haiti damage report japan nuclear crisis motivation is clear; data challenges remain. Tuesday, 7 May 13
- 10. C1. violations of statistical assumptions C2. variability in expertise: unknown experience, biases motivation and intention factors C3. gaming the system Tuesday, 7 May 13
- 11. C1. violations of statistical assumptions C2. variability in expertise: unknown experience, biases motivation and intention factors C3. gaming the system non-probability sampling methods convenience sampling, snowball sampling statistical data analysis, multi-factor latent process analysis robust statistical ranking, scoring game-proo?ng design Tuesday, 7 May 13
- 12. [Haiti Earthquake Usahidi] Can unbounded crowdsourcing (non- representative sampling) via SMS predict actual damage? http://irevolution.net/2010/10/13/crowdsourced-prediction/ Tuesday, 7 May 13
- 13. 8,163 copies of War and Peace (June 2011) ¡°Is It Really About Me? Message Content in Social Awareness Streams¡± [Naaman et al. CSCW ¡¯10] information sharing (IS) 0.2 self promotion (SP) 0.05 opinions/complaints (OC) 0.25 statements and random thoughts 0.25 me now (ME) 0.4 questions to followers (QF) 0.05 presence maintenance (PM) 0.05 anecdote by me (AM) 0.05 anecdote by others (AO) 0.01 Tuesday, 7 May 13
- 14. likert ordering pickbest User visits http://twiage.me and registers 1. 2.A question/tournament is chosen One of three types of rounds are selected randomly 3. The user¡¯s score is tallied and their rankings are displayed. Next round starts.4. Twiage - Van Kleek et al. CHI 2012 Tuesday, 7 May 13
- 15. Q1: Q2: Q3: twiage user selections/ratings EWR ELO LR rank set (66%) Rank agreement Classification agreement (Top 5%) test set (33%) EWR ELO LR Agreementmeasures randomly divide Tuesday, 7 May 13
- 16. Empirical Win Elo Raw Likert All 69.7 (0.9) 72.8 (0.9) n/a Pick-Best 71.2 (3.0) 75.5 (0.8) n/a Likert 67.7 (1.6) 71.4 (1.6) 64.8 (2.0) Ordering 63.3 (1.9) 62.4 (2.2) n/a ranking scheme roundtype rank agreement Tuesday, 7 May 13
- 17. 1 - crowdsourcing data 2 - human motivation 3 - behavioural design 3 challenges Tuesday, 7 May 13
- 18. Tuesday, 7 May 13
- 20. Tuesday, 7 May 13
- 21. Reddit, Digg, Slashdot, HackerNews, .Crowd?ltering Tuesday, 7 May 13
- 22. Yahoo! Answers, Knowledge NI, QuoraSocial Q&A Tuesday, 7 May 13
- 23. Software development + debugging assistancSocial Coding Tuesday, 7 May 13
- 24. Software development + debugging assistancSocial Coding Tuesday, 7 May 13
- 25. Tumblr - Pinterest/Pinspire - Path 76.6 Million Blogs, 32.6 Billion Posts 50GB of posts added?each day; follower list updates are roughly another 2.7 terabytes daily Youtube, Vimeo Content creation and curation networks: Blogging, Tumblogging, Vlogging, Geotumbling, Facebooking. Wordpress, Livejournal, Tuesday, 7 May 13
- 26. what attracts participation in these communities? what drives continued involvement? what causes communities and platforms to fail ? Tuesday, 7 May 13
- 27. theories of human motivation instinct incentive drive arousal identity/self-expression role ful?llment brand/image/community association social contact Tuesday, 7 May 13
- 28. theories of human motivation instinct incentive drive arousal identity/self-expression role ful?llment brand/image/community association social contact what does this have to do with interface design? { fb / G+ } - real-name policy the saga of orkut, myspace and dead communities the un-instagrams Tuesday, 7 May 13
- 29. 1 - people as data sources 2 - human motivation 3 - behavioural design 3 challenges Tuesday, 7 May 13
- 30. HOW DO interface ARTEFACTS affect human behaviour? Tuesday, 7 May 13
- 31. Luce¡¯s Model of Choice Violations of principle of proportionality Similarity Hypothesis (Tversky et al. 1972) Regularity Hypothesis Asymmetrically Dominated Alternatives (Huber et al. 1982) Tuesday, 7 May 13
- 32. Tuesday, 7 May 13
- 33. SNACKBOT Min Kyung Lee, CMU CHI 2011 Tuesday, 7 May 13
- 34. default choice bias present-biased preferences asymmetric choice SNACKBOT Min Kyung Lee, CMU CHI 2011 Tuesday, 7 May 13
- 35. tomorrow you would like : salad - sandwich for a snack: apple - cookie default choice bias present-biased preferences asymmetric choice SNACKBOT Min Kyung Lee, CMU CHI 2011 Tuesday, 7 May 13
- 36. tomorrow you would like : salad - sandwich for a snack: apple - cookie You selected a salad for lunch today would you like to keep this choice or switch to a Roast Beef Sandwich what would you like tomorrow? default choice bias present-biased preferences asymmetric choice SNACKBOT Min Kyung Lee, CMU CHI 2011 Tuesday, 7 May 13
- 37. tomorrow you would like : salad - sandwich for a snack: apple - cookie You selected a salad for lunch today would you like to keep this choice or switch to a Roast Beef Sandwich what would you like tomorrow? default choice bias present-biased preferences asymmetric choice apple - celery - cookie SNACKBOT Min Kyung Lee, CMU CHI 2011 Tuesday, 7 May 13
- 38. summary ¡°Rapid Assembly of Social Machines¡± FLOSS Software Platforms (Tech. Infrastructure) To make truly effective Social Machines social interfaces can borrow/beg/steal from social and cognitive science Crowd-data distillation Group coordination and mobilisation Core motivation Lightweight structures of distributed management and coordination Interface design choices that compel humans to be empowered, participate, motivate. Tuesday, 7 May 13