Motivation and Incentives: An Evidence-Based Approach to Community - - PowerPoint PPT Presentation

motivation and incentives an evidence based approach to
SMART_READER_LITE
LIVE PREVIEW

Motivation and Incentives: An Evidence-Based Approach to Community - - PowerPoint PPT Presentation

Motivation and Incentives: An Evidence-Based Approach to Community Management Prof. Jana Gallus UCLA @janagallus COMMUNITY MANAGEMENT TWO COMMON CHALLENGES @janagallus CHALLENGE 1: DIVERSITY & INCLUSION OSS WIKIPEDIA 2% 1%


slide-1
SLIDE 1

Motivation and Incentives: An Evidence-Based Approach to 
 Community Management

  • Prof. Jana Gallus

UCLA @janagallus

slide-2
SLIDE 2

COMMUNITY MANAGEMENT – TWO COMMON CHALLENGES

@janagallus

slide-3
SLIDE 3

OSS

2%

3%

95%

Men Women Non-binary

GitHub Open Source Survey 2017 Wikimedia Foundation Survey 2018

WIKIPEDIA

1%

9%

90%

Men Women Non-binary

CHALLENGE 1: DIVERSITY & INCLUSION

@janagallus

slide-4
SLIDE 4

Notes: “Active editors” are those with ≥5 edits/month. English language version only. Halfaker et al. (2013). The rise and decline of an open collaboration system. American Behavioral Scientist 57.

CHALLENGE 2: SUPPORTING PEOPLE’S MOTIVATIONS

Example: Wikipedia’s editor retention problem

@janagallus

slide-5
SLIDE 5

Standard economic incentives & contracts of limited use

  • Motivation crowding-out (intrinsic & image motivation)
  • Tasks cannot be contracted (creative, complex, prosocial)
  • Budget constraint

➡ Alternatives?

Community management challenges

@janagallus

slide-6
SLIDE 6

Recognition

Blood Donor's Medal UK US Nobel Peace Prize @janagallus

slide-7
SLIDE 7

WHAT WORKS? A DATA-DRIVEN APPROACH TO COMMUNITY MANAGEMENT USING FIELD EXPERIMENTS

Traditional firms Public sector Platforms

@janagallus

slide-8
SLIDE 8

1) SUPPORTING PEOPLE’S MOTIVATIONS

@janagallus

slide-9
SLIDE 9

Can symbolic awards help retain newcomers on Wikipedia?

Award on newcomer’s personal discussion page

Gallus (2017). Fostering public good contributions with symbolic awards: A large-scale natural field experiment at Wikipedia. Management Science 63: 3999-4015 (Open Access). @janagallus

slide-10
SLIDE 10

The award page

Can symbolic awards help retain newcomers on Wikipedia?

slide-11
SLIDE 11

The purely symbolic awards increase the share of newcomers remaining active in the month following the award bestowal 
 by 20 percent (p = 0.000).

Note: error bars indicate 95% confidence intervals. Data from 11 cohorts.

Main findings from field experiment

@janagallus Gallus (2017). Fostering public good contributions with symbolic awards: A large-scale natural field experiment at Wikipedia. Management Science 63: 3999-4015 (Open Access).

slide-12
SLIDE 12

Treatment effect persistence

Notes: Average values rounded to 2 decimal places. p-values from Chi-square tests in parentheses. The two higher-level awards, which are not randomly bestowed, fall into quarters 1 and 2; they could be received after months 2 and 5 after the initial award whose effects are being tested. The scheme includes no further awards thereafter. Adjustments for multiple comparisons do not change the significance of any of the tests.
 * p<0.05, ** p<0.01, *** p<0.001.

Gallus (2017, Management Science) @janagallus

slide-13
SLIDE 13

Symbolic awards can be used to sustainably motivate contributors

  • Effect on willingness to do tedious maintenance tasks
  • Awards foster confidence & identification with community

Implications

“Dear Edelweiß-Team, as an absolute newcomer I am very delighted about this award! I initially only wanted to make a few corrections every now and then, but this form of welcoming has highly motivated me! I am now working on my first article... Many heartfelt thanks in retrospect!” (emphasis in original)

@janagallus Gallus (2017, Management Science)

slide-14
SLIDE 14

2) DIVERSITY & INCLUSION

@janagallus

slide-15
SLIDE 15

Teams miss out on high-quality contributions

One important reason: Self-stereotyping, which is rooted in beliefs

  • Not about motivation (requiring incentives)
  • Not about lacking the knowledge (training, screening)
  • Not about discrimination – though that may come on top
  • Not (only) about under-confidence

@janagallus

slide-16
SLIDE 16

Can recognition correct beliefs? Which form works best?

Gallus & Heikensten. Shine a light (on the bright). AEA Papers & Proceedings, forthcoming.

  • Experimentally vary publicness:
  • Private feedback vs.
  • Virtual award vs.
  • Face-to-face ceremony
  • Focusing on collaborative, computer-mediated work on math

tasks in the lab

slide-17
SLIDE 17

0.19 0.26

.05 .1 .15 .2 .25 Difference in confidence to speak up between pre- and post-treatment periods Non-recipients Recipients

Note: Results hold when controlling for ability.

Recognition makes recipients more confident to speak up

@janagallus

slide-18
SLIDE 18

3.46 3.17 3.63 3.34 3.48 3.30 3.73 3.47 3.81 3.59 3.65 3.62

1 2 3 4

Confidence to speak up

Pre-treatment Post-treatment

Private Virutal Ceremony Private Virutal Ceremony

Recipients' confidence to speak up

Male Female

The gender gap across conditions

The form of recognition matters

@janagallus

slide-19
SLIDE 19
  • Self-stereotyping among high-ability women produces gender gap

in contributions when working on male-typed tasks in the lab

  • Recognition increases confidence to contribute 


(No effect on non-recipients)

  • The form of recognition matters: Face-to-face ceremony closes

gender gap

  • Public award seems to increase legitimacy & trust in award signal

Implications

Gallus & Heikensten. Shine a light (on the bright). AEA Papers & Proceedings, forthcoming.

slide-20
SLIDE 20

LIMITATIONS OF RECOGNITION & THE IMPORTANCE OF RIGOROUS EVALUATION

@janagallus

slide-21
SLIDE 21

Field experiment on two widely used award types

FRT p-value = .029 Robinson, Gallus, Lee & Rogers (2019). The demotivating effect (and unintended message) of awards. Organizational Behavior and Human Decision Processes, in press.

Note: Error bars show 95% confidence intervals. Sample: 15,329 students in grades 6-12

slide-22
SLIDE 22

FRT p-value = .029

+8% absences

Note: Error bars show 95% confidence intervals. Sample: 15,329 students in grades 6-12

Robinson, Gallus, Lee & Rogers (2019). The demotivating effect (and unintended message) of awards. Organizational Behavior and Human Decision Processes, in press.

Field experiment on two widely used award types

slide-23
SLIDE 23

… but it can backfire!

  • Crowding-out of motivation
  • Strategic gaming, multi-tasking
  • Hubris
  • Envy

➡ Needs to be designed carefully. Don’t forget:

  • Tasks that are less visible
  • People who have already made an effort

Recognition is intuitively appealing

Gallus (2018). The best ways to give employees performance awards. The Wall Street Journal

slide-24
SLIDE 24

Recognition can be used to sustainably motivate contributors. It can make a-typical users more confident to contribute. Not one-size fits all – the form matters (and there are many). Beware unintended effects!

Take-aways

@janagallus

slide-25
SLIDE 25

Institutionalize recognition, 
 design wisely, 
 and test

@janagallus

slide-26
SLIDE 26

2017 Oxford University Press

Please reach out if interested in evaluating community management practices: jana.gallus@anderson.ucla.edu @janagallus