Go Going Fart rther r Together: r: The The I Impact mpact o - - PowerPoint PPT Presentation

▶
go going fart rther r together r the the i impact mpact o
SMART_READER_LITE
LIVE PREVIEW

Go Going Fart rther r Together: r: The The I Impact mpact o - - PowerPoint PPT Presentation

Sep 05, 2023 521 likes •774 views

Go Going Fart rther r Together: r: The The I Impact mpact o of So Social C cial Capit apital o al on n Sus Sustaine ned P d Partici cipa pation n in Open S in Open Sour urce ce Huilian Sophie Qiu , Alexander Nolte, Anita

slide-1
SLIDE 1

Go Going Fart rther r Together: r: The The I Impact mpact o

  • f So

Social C cial Capit apital o al on n Sus Sustaine ned P d Partici cipa pation n in Open S in Open Sour urce ce

Huilian Sophie Qiu, Alexander Nolte, Anita Brown, Alexander Serebrenik, and Bogdan Vasilescu @sophiehsqq @alexander_nolte @aserebrenik @b_vasilescu

slide-2
SLIDE 2

Ma Maintain, or

  • r su

suffer

1

(Greenstein and Nagel, 2016) https://w3techs.com/technologies/ history_overview/web_server

slide-3
SLIDE 3

Op Open-so source exp xperience boosts s resu sumes

2

Employers (and job seekers) use open-source experience to make inferences (or form impressions) about a candidate’s technical skills.

(Marlow et al., 2013)

https://codeformore.com/how-to-write-up-open-source-experience-when-you-dont-have-any/

slide-4
SLIDE 4

Kn Knowledge gap

3

developer

long-term developer

  • utsider

newcomer motivation

  • nboarding

retention

This work

?

slide-5
SLIDE 5

RESULT T HIGHLIGHTS: Social capital ex explains susta tained participation

4

Higher social capital Being part of teams with more diverse information, especially for women

slide-6
SLIDE 6

~40% 40% of

  • f G

GitHub c con

  • ntri

ributor

  • rs d

disengage

5

+ + + + + + + + + + + + + + + + + + + +

0.00 0.25 0.50 0.75 1.00 12 24 36 48 60

Time in months Survival probability

slide-7
SLIDE 7

Soci Social c capital t theor

  • ry f

for s

  • r sustainable p

part rtici cipation

  • n

Willingness to continue (Coleman, 1990)

6

Bonding social capital:

benefiting from strongly connected network

Bridging social capital:

benefiting from network with diverse info Opportunity to continue (Burt, 1998, 2001)

slide-8
SLIDE 8

H1: more soci cial capital ~ ~ mo more e prolonged ed en engagemen ement

Willingness to continue (Coleman, 1990)

7

Bonding social capital:

benefiting from strongly connected network

Bridging social capital:

benefiting from network with diverse info Opportunity to continue (Burt, 1998, 2001)

slide-9
SLIDE 9

Cohesive networks might foster discr crimination / excl clusion

8

slide-10
SLIDE 10

After one year ca. 70% of men are still active but only ca. 60% of women

9

On On GitHub, , women dise sengage earlier than men

slide-11
SLIDE 11

After one year ca. 70% of men are still active but only ca. 60% of women

10

Why do contributors disengage?

  • Developer survey in our paper:
  • Work-related (e.g., new job)
  • Personal* (e.g., different hobby)

*women cite more often than men

See also:

  • Miller et al, OSS 2019
  • Iaffaldano et al, SoHeal 2019

On On GitHub, , women dise sengage earlier than men

slide-12
SLIDE 12

H2: H2: Tea eams ms with mo more e diver erse e informa mation ~ mo more e pr prolong nged d eng ngagement, t, esp.

  • p. for

r women

11

Information diversity should reduce the risk of demographic- based echo chambers.

slide-13
SLIDE 13

Survey 88 replies Small sample 1,000 users

12

La Large-sc scale mixed-me method

  • ds s

study

Survey 88 replies Small sample 1,000 users

slide-14
SLIDE 14

In Infer erring ing gender ender from nam names es

13

gender Computer

https://github.com/tue- mdse/genderComputer https://www.namsor.com name features, e.g., the last two characters Naive Bayes classifier Binary gender prediction Public name lists & celebrity names, including 3,000 East Asian names

slide-15
SLIDE 15

Bon Bonding soc

  • cial capital – Te

Team Familiarity

14

End of data Project A Project B Project C Project X (de Vaan et al., 2011, Lutter 2015)

TIME

slide-16
SLIDE 16

15

End of data Project D Project C Project X (de Vaan et al., 2011, Lutter 2015)

TIME

Bon Bonding soc

  • cial capital – Re

Recurring Cohesion

slide-17
SLIDE 17

16

End of data Project X

Java C C++ C++ C

(de Vaan et al., 2011, Lutter 2015)

TIME

Bri Bridging soc

  • cial capital – La

Language D Diversity

slide-18
SLIDE 18

17

End of data Project X Project X Project X (de Vaan et al., 2011, Lutter 2015)

TIME

Bri Bridging soc

  • cial capital – Sh

Share of

  • f Ne

Newcome

  • mers
slide-19
SLIDE 19

18

active, major contributor, proj owner, social capital active, major contributor, not owner, social capital active, minor contributor, not owner, social capital left, minor contributor, not owner, social capital

CO COX regression model

Contributor Time Active Social capital Control variables 2008 Jan – Mar True

Team familiarity Recurring cohesion Language diversity Share of newcomers

Project size Project owner ……

2008 Jan – Mar True

Team familiarity Recurring cohesion Language diversity Share of newcomers

Project size Project owner ……

2009 Apr – Jun False

Team familiarity Recurring cohesion Language diversity Share of newcomers

Project Size Not project owner ……

slide-20
SLIDE 20

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

0.00 0.25 0.50 0.75 1.00 12 24 36 48 60

Time in months Survival probability Recurring cohesion

+ +

High Low 19

H1: more soci cial capital ~ ~ mo more e prolonged ed en engagemen ement

slide-21
SLIDE 21

0.00 0.02 0.04 0.06 0.08 12 24 36 48 60

Time in months Difference in survival probability Gender

Female Male

Survival difference between contributors with high and low language diversity

20

H2: Language diversity interact cts with gender

slide-22
SLIDE 22

Wha What ne next?

21

Recommend projects that can help build social capital Offer mentorship Use badges to show a community’s culture

slide-23
SLIDE 23

Soci Social c capital e explains p prol

  • lon
  • nged e

engageme ment

22

Acknowledgements:

Contact: Sophie Qiu @sophiehsqq hsqq@cmu.edu

Code and data:

@alexander_nolte alexander.nolte@udo.edu @aserebrenik a.serebrenik@tue.nl @b_vasilescu vasilescu@cmu.edu