Peter John (MS PowerPoint , 681kb)

If I can do to it so can you: How leader and
personal endorsement affects participation in
Peter John, Beth Carley, Oliver James,
Alice Moseley, Liz Richardson, Matt
Ryan and Gerry Stoker
Recruitment: leader endorsement
• Invitation to volunteer from a ‘leader’ who is already doing
the activity – a politician, celebrity, another citizen
• May motivate on basis of:
– modelling (providing positive example) (Bekkers & Wiepking, 2011;
Glӧckner et al., 2010)
– legitimisation/ confirmation of quality of org/ activity (Bekkers &
Wiepking, 2011; Vesterlund, 2003; Andreoni, 2006)
– leader self-sacrifice (De Cremer et al., 2009 ; Glӧckner et al., 2010).
• Politicians have authority but effect of ‘anti-politics’? (Hay,
2007); many volunteers a-political/ motivated by lack of
institutional trust (Eliasoph, 2009; John et al., 2011)
Research Design
• An experiment that assesses the impact of
endorsements from people in different types of
social positions on the decision to volunteer.
• The idea is that potential volunteers will be
influenced by endorsement in the form a request
in which the endorser is cited as having
undertaken volunteering activity themselves (‘do
as I do’) rather than merely because the endorser
recommends volunteering as a desirable activity
(‘do as I say’).
More on design
• Five participating universities with a total of
100,974 students: UCL, Southampton, Salford,
Exeter, Plymouth
• Block randomisation within each university (done
by York Trials unit- thanks esp to Hannah)
• The instrument was a standard e-mail invitation
to volunteers that went to all students, but with
additional endorsement information included in
the treatment messages from three different
types of endorser.
• Those in the control group receive a simple invitation
from the volunteering unit at their respective
university encouraging them to volunteer and sending
them a link to the volunteering unit’s website.
• Those in the treatment groups receive a similar
invitation, but delivered in the framework of an
• Treatment 1 informs the recipient about a politician
who has volunteered
• Treatment 2 concerns a celebrity volunteer;
• and Treatment 3 conveys endorsing information about
the activity of another student volunteer;.
Detail on endorsements
• The individual politician and celebrity endorsers included in
these messages are people who have been identified
through internet research as having undertaken
volunteering and were invited by the research team.
• Student endorsers were contacted and their agreement to
participate sought through the volunteering units.
• In order to avoid confounding endorsement effects with
organisation name or brand recognition effects, each
endorsement will convey the type of volunteering activity
undertaken by the endorser, e.g. volunteering for an
organisation that helps the elderly, rather than referring to
a specific named organisation.
More on endorsers
• Endorsers not prominent or controversial
• Politicians were members of Parliament, but not
members of the government and selected from
across the main political parties, although party
affiliations were not be mentioned
• The student endorsers selected from outside of
the institutions in the study to reduce personal
acquaintance effects.
We are writing to you directly from Community Action, part of the University of Exeter Students’ Guild, to
invite you to sign up to our service.
Anyone can volunteer and all sorts of people do, including national politicians like Gloria de Piero, Dominic
Rennie Raab, Andrew Percy and Roberta Blackman-Woods - MPs who support, develop and promote charitable
projects, organisations, and causes.
Everyone needs a break from the routines of student life, so why not try something enjoyable that can make a
difference to other people too?
Visit the Community Action website to find out more:
And registering with our group just takes a couple of clicks:
Volunteering is a great way to give something to others while gaining new skills and confidence, making friends
and exploring career options. From one-off activities, to more extended roles, there is something for everyone.
We look forward to hearing from you!
Community Action, University of Exeter
In partnership with the Giving Time research project
Our priors
• We did a research design document presented
in May 2013 with analysis plans and study
• Stated expected hierarchy of treatment effects
from literature (hard to compare with our
• Celebrities > peers > control > politicians
• Heterogeneity by university (but no expected
Monitoring outcomes
• All invitations contained a link to the student
volunteering service where students interested in
volunteering register.
• There are two binary outcome measures: firstly
whether the recipient clicks through the link contained
in the e-mail and registers with the volunteering unit,
as a measure of intention to volunteer; secondly,
whether the student takes up a volunteering role, as a
measure of volunteering activity, based on information
provided by the volunteering unit at the institution in
• A limited set of covariates for analysis were
provided from student records, including gender,
ethnicity, year of study and subject discipline.
• Multiple outcome measurement ranging through
no response, clickthroughs, registered intention
to volunteer, taken up volunteering role and
• Measured by Mailchimp, student records from
volunteering units and survey
Clickthroughs to UCL’s VSU website
Summing up the findings
• Endorsements have no positive effects on
click-throughs to volunteering units
• Student endorsements are negative, but
would other kinds of endorsement work?
• Impact of politicians on uptake of training
• Patchy impact of celebrities
• Some little heterogeneity by university and
ethnic background

similar documents