Seminars & Conferences

2018: Lecture TU Delft

Crowdfunding science-only projects: A good alliance, or bad marriage?

There is an increasing focus on third party funding to support scientific research, since it can be a resourceful way for scholars and academics to fund. An example of such a less traditional funding method is crowdfunding. Crowdfunding is an online funding method that builds on a large group of individuals each donating a small amount. Can crowdfunding close the gap for potential and promising but unfunded projects? Compared to the more traditional funding methods, the amounts assembled through crowdfunding might appear small and insignificant. However, crowdfunding should not be overlooked as it has several opportunities. Nevertheless, it is the question whether science projects can fit the crowdfunding model. In addition, if science projects can be supported by crowdfunding, we need to determine whether crowdfunding is a savior, a substitute or a complimentary source for science projects. Is it a good alliance, crowdfunding and science?

Date:                Thursday October 4, 11:30 – 13:00 (incl. lunch)
Venue:             Theaterzaal, X TU Delft (former Sport & Culture), Mekelweg 8-10, 2628 CD Delft

2018: DVDS – conference

Follow the crowd:

Social information and crowdfunding donations in a large field experiment

Claire van Teunenbroek & Rene Bekkers.

Center for Philanthropic studies, VU University Amsterdam

Slides here: DVDS 2018 _ van Teunenbroek

 

Abstract

Guiding human decision making by a discrete suggestion, that is a nudge, is increasingly popular. One particular promising nudge is to provide individuals with the amount donated by others, i.e. social information. Social information is frequently used by practitioners to increase charitable giving, while the literate suggests that the effects are context specific and inconsistent. Also, the use of social information could decrease donation amounts and, therefore, prove costly for practitioners. Our article, based on a large natural field experiment (n =  24,070), adds to the literature by researching if social information is an effective stimulant for crowdfunding, a relatively new fundraising instrument in an online context. The literature has spent limited to no time to researching social information effects in crowdfunding, while the growing use of crowdfunding emphasizes the importance of understanding stimulants for crowdfunding donations. Our key hypothesis is that social information increases the amount donated. In addition, we tested three additional hypotheses stating that the strength of the effect depends on: (1) social norm, (2) reactance and (3) project funding period. The data was collected at Voordekunst, the largest Dutch culture and arts crowdfunding platform. We used a straightforward A/B test with two conditions: (1) control condition: without adding social information, and (2) treatment condition: adding social information using the following sentence: “Did you know that the average donation amount at Voordekunst is €82?”. We show that social information increases the amount donated by 16%, similar to previous research.

Keywords

Crowdfunding, decision making, donation behavior, natural field experiment, online donations, social information

2018: SABE/IAREP- conference

Follow the crowd:

Social information and crowdfunding donations in a large field experiment

Claire van Teunenbroek & Rene Bekkers.

Center for Philanthropic studies, VU University Amsterdam

Abstract (941 words )

A common technique in fundraising is the use of social information, i.e. information on donations made by other donors. Previous research in offline contexts suggests social information on average increases donations by 15%. What is the effect of social information in online contexts? We present a large natural field experiment, testing the effect of social information in crowdfunding, an online fundraising instrument. Also, we test three additional hypotheses: on (1) social norms; (2) reactance; and (3) project funding time.

The social norm hypothesis predicts that when the amount displayed in the social information condition is larger than the intended donation amount, the amount donated increases, and when the amount displayed is lower the amount donated decreases. By making social information available, we hint at the ‘expected’ behavior and we try to reduce uncertainty about the appropriate amount.

The reactance hypothesis predicts that the larger the difference between the amount displayed in the social information condition and the amount donors intended to give, the smaller the effect of the amount displayed.

The project period hypothesis predicts that the effect of social information is strongest in the middle of the campaign, where most of the donors consist out of individuals searching for cues. Crowdfunding projects are advertised for a limited period of time. In the beginning of a campaign the donors mainly consist out of family and friends, while donors in the middle of the campaign are less likely to have social connections with the solicitor. Friends and family of the solicitor are likely to have some sense of the social norm through previous contact with the solicitor, in contrast to the ‘crowd’ that consists of latent and weak ties of the solicitor.

Methods: In this preregistered study (https://aspredicted.org/u5w9u.pdf) we report results from a large natural field experiment among visitors (n = 24,072) of voordekunst.nl, the largest crowdfunding platform advertising culture and arts campaigns in the Netherlands. We implemented a straightforward  A/B test with two conditions (1) control: the regular website, without adding social information; (2) treatment: adding social information using the following sentence: “Did you know that the average donation amount at Voordekunst is €82?”. The information provided is the actual average. The experiment was conducted among 24,072 unique viewers. 11,974 visitors were assigned to the treatment condition and 12,098 to the control condition.

Preliminary findings: A total of 2,657 visitors (11.0%) made a donation. The idea that social information increases donation amounts is not supported by the data in a straightforward comparison of means (F = .071, n = 2,657, p = .790), but did receive support in a non-parametric test (F = 5.240, n = 2,657, p = .022). There were a few exceptionally large donations (1.39%, N=19), i.e. more than 3 standard deviations above the average. In an analysis without these outliers, social information did increase donation amounts (F = 4.723, n = 2,638, p = .030). Donors donated about €11 higher amounts when they had received social information (M = 80.73, SD = 146.96, n = 1,274) than in the control condition (M = 69.54, SD = 116.74, n = 1,364). This is a 16% difference and in line with previous research. A  regression analysis of the amounts donated with project fixed effects and excluding the outliers confirmed that social information increased amounts donated by 16% (b = 11.119, n = 2,638, p = .029).

The social norm hypothesis: Figure 1 shows the distribution of the amounts donated per condition. We expected that a larger proportion of donations in the treatment condition would be close to but above €82 than in the control condition. This was not the case. Instead, the treatment lowered the proportion of donations below €25 (low amounts: 45.1% in control vs 39.3% in treatment) and increased amounts between €501 and €1,000 (high amounts: 51.0% vs 56.1%), and intermediate amounts (€26-€500, 3.9% vs 4.6%). A Chi-square test indicates that social information affected the distribution of amounts over these categories (X2=9.146, p=.010).

Figure 1. Histogram of the amounts donated per condition

 

Reactance hypothesis: We tested whether the mean absolute difference from the suggested amount differed between the two conditions. We found no difference in the mean absolute difference between the control (M = 92.899, SD = 334.756, p<.118) and treatment condition (M = 91.802, SD = 249.216). Excluding the outliers, the mean absolute difference was somewhat larger (75.541, SD = 126.040) in the treatment condition than in the control condition (M = 68.705, SD = 95.186), but not significantly so (F = 2.491, p = .115).

 

Project funding time hypothesis: We computed a project funding stage variable based on the number of days since the project launched and defined three stages: beginning (until day 26), middle (from day 26-35) and end (after day 35). The average amounts donated were higher in the middle (€160) than in the beginning (€69) and final stage (€111). Excluding outliers, the amount increased from €60 in the beginning stage to €90 in the middle and €99 in the final stage. As figure 2 shows, the pattern we predicted did not emerge: the social information effect was absent in the middle stage, but increased giving in the beginning stage and the final stage.

Figure 1. Line graph of the social information effect at different funding stages by condition.

Discussion: The main hypothesis, social information increases amounts donated, was not supported by the data including all donations. However, excluding outliers, we found a modestly positive effect of about 16%, which is in line with previous research. The major contribution of our study is that we show that the effect of social information varies over the project funding stage. A limitation of our data is that we have no information on individual donors. More research is needed to further specify for which donors social information is more effective.

2018: ISTR- conference

Follow the crowd:

Social information and crowdfunding donations in a large field experiment

Claire van Teunenbroek & Rene Bekkers.

Center for Philanthropic studies, VU University Amsterdam

ISTR- Amsterdam, The Netherlands, July 10-13, 2018

Abstract

In practice, donors respond to how much other people have given, i.e. social information. People like to follow and the mention of the donation amount of other’s makes this easier. Although researchers have suggested that social information could increase amounts donated, a recent study suggest that social information effects are context specific. Our article, based on a large natural field experiment, adds to the literature by researching if social information is an effective stimulant for crowdfunding, an online fundraising instrument. The literature has spent limited to no time to researching social information effects in crowdfunding, while crowdfunding could use a stimulant as it is relatively ineffective in assembling donations. We show that SI increases the amount donated by 16%.

 

2018: ISTR- conference

A helping hand:

Challenges for the Dutch philanthropic sector and the possibilities of crowdfunding

Claire van Teunenbroek & Rene Bekkers.

Center for Philanthropic studies, VU University Amsterdam

ISTR- Amsterdam, The Netherlands, July 10-13, 2018

Abstract

Instead of taking the readers by the hand and running them down the hill, I want to lead them along the different topics and importance of my dissertation. I want to take some time to really argue why this research is important: the societal value. In doing so I discuss the challenges of the cultural sector and their increasing focus on philanthropy. Next I present why a focus on philanthropy might not be ideal, as they are experiencing a decrease in donations and charities are have increasing difficulties to reach out to potential donors. After this I want to present crowdfunding as a possible ‘helping hand’, and discuss if this type of solicitation fits the demands on donors. While crowdfunding could work, as it is transparent and democratic like the public is demanding, it makes up for only a small amount of the total philanthropic amount. Thus, in order to use crowdfunding as an effective solicitation method for the third sector it could benefit from a stimulant to make it more effective in terms of assembling donations. The main contribution of this dissertation is the focus on social information (yes again SI) as a possible stimulant for crowdfunding specifically and charitable giving in general. I finish with explaining the meaning and potential of social information.

2017: WINK – conference

Are we Joining the Crowd? A natural field experiment on the impact of social information in crowdfunding campaigns. 

Claire van Teunenbroek & Rene Bekkers.

Center for Philanthropic studies, VU University Amsterdam

WINK – the Nudge conference, Utrecht, The Netherlands, June 23-24, 2017

Abstract

Philanthropic crowdfunding is an online funding method with a growing popularity. In this study we quantify the effects of information about the donation behavior of previous donors, also known as social information. We report results from a large natural field experiment among visitors (n = 23,676) of a crowdfunding platform advertising campaigns for artists and nonprofit organizations in the field of arts and culture. Visitors who were exposed to a reminder about the average donation amount of previous donors were not more likely to donate than visitors who were not exposed to this information. Neither did visitors who decided to donate give higher amounts.

Van Teunenbroek, P.S.C. & Bekkers, R. (2016). Joining the Crowd: Social Information Effects on A Crowdfunding Platform. Paper presented  at the 2017 WINK – nudge Conference, Utrecht, June 24, 2017. Session 04C, Saterday, June 24,  11:00-11:30 AM. Preregistration here. Slides here: Are we joining the crowd (van Teunenbroek &amp; Bekkers, 2017) WINK conference June 2017.

wink-banner4

2016: 16th ICPPE Conference, Rome

Text based on my conference paper : van Teunenbroek, P. S. C. (2016). Social Aspects and Successfully Funding a Crowd-Funding Project: The Impact of Social Information. World Academy of Science, Engineering and Technology, International Journal of Social, Behavioral, Educational, Economic, Business and Industrial Engineering, 10(6), 1765-1776.  http://www.waset.org/publications/10004657

During the presentation I presented the preminarly results of my second pilot, consisting out a class-room experiment with  181 students from the VU University. The results suggest that social information, in the form of the average donation amount of previous donors, positively influences the donation behavior. It increased not only the average an individual (i.e. intensive influence) donates but also the number of donors (i.e. extensive influence) and therefore increasing the total amount assembled.

TABLE IX

PROPORTION DONATING AND AMOUNTS DONATED

Control Treatment Test statistics Significance
Median  amount donated €10 €20 X2=7.47 p=.006
Minimum donation €0.00 €0.00
Maximum donation €500.00 €250.00
Mode donation €10.00 €20.00
Number of donors 81% 94% X2=.19 p=.012
Total contribution €1402 €2097 X2=1.00 p<.001

As can be seen in figure 1, the reference amount (15 euros)  was donated twice as often donated in the treatment condition. However, contrary to the expectations the suggestion amount (15 euros) was not the most popular amount. The most popular donation amount was €20 (donated by 18.1%).

bar graph pilot 2 experiment 1 - psc van teunenbroekFig. 1 Bar graph of the frequency of donation rates per condition

It could be that individuals prefer to donate slightly above the reference amount because they want to signal their wealth and generosity. In other words, could this finding be explained by stating that  individuals are ‘outigiving’? We base this statement on our additional following finding: individuals who have a concern for their reputation (i.e. find a positive reputation important) donate an amount that is higher than the suggestion amount.

These are just some of the findings. Please keep in mind that this research is in an early stage, we will keep you updated with possible new findings.

Powerpoint link: Presentation ICPPE 2016 -VU Claire van Teunenbroek

Forthcoming- Accepted for 16th ICPPE 2016 Conference, Rome

Paper: Social aspects and successfully funding a crowdfunding project: The influence of social information on a real life crowdfunding platform.

.

     Keywords: online donation behavior, philanthropic crowdfunding, social information, social influence, social motivation

Picture1.jpg

2015 – 44th ARNOVA conference, Chicago

One of the advantages of being a PhD candidate is the chance to visit and present your paper at conferences. I visited my first conference in December last year, I presented my first conference paper at the Association for Research on Nonprofit Organizations and Voluntary Action (ARNOVA). The paper consists out of an overview of the published articles reporting social information effects on donation behaviour (i.e. systematic literature review).

Powerpoint link: Powerpoint ARNOVA conference 2015 – P.S.C. van Teunenbroek

2015 – Helsinki, PhD progress seminar

June 7, 2015 – June 9, 2015

Place: Calliola Conference Center, Snappertuna (Finland)
Universities: Hanken University (Finland), University of Liverpool (England) and VU University Amsterdam (the Netherlands)

PhD Finland

Forthcoming- Accepted for ARNOVA’s 2015 Conference

I’m excited about the opportunity to be one of the speakers at the ARNOVA conference 2015. Together with my colleague Irma Borst (also VU University, Faculty of Social Sciences) we will present the following paper at ARNOVA’s 2015 Conference in Chicago, IL November 19-21.

‘Social information as a charitable stimulant: A literature review of the effects of social information on donation behaviour’

Recently, crowdfunding -raising external funding online- emerged as a new funding instrument for non-profit organizations. Practitioners assume that social information influences donation behaviour in crowdfunding. We provide a literature review of social information, in particular in the context of charity giving. Since the number of publications on crowdfunding is limited, our literature review is mainly based on traditional charity giving. We will theorize whether effects of social information found in traditional charity giving -mainly offline- apply to the online crowdfunding context. As a result, we develop a theoretical framework explaining and undermining effects of social information on online donation behaviour.

For more information regarding the conference : http://www.arnova.org/?page=conference

ARNOVa l

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s