incentive models for longitudinal studies

30 Oct 2008 - 4:28pm
5 years ago
4 replies
675 reads
Meredith Noble
2010

Has anyone done a longitudinal study before? What was your incentive
model like?

We're asking participants to commit to 8 sessions, happening roughly two
weeks apart. We are fully prepared for drop-outs but want to minimize
this, obviously.

We'd considered giving people some money for each session (say $50), and
then reserving the same amount of money for the very end, which would be
handed over if they complete all 8 sessions. So they'd get $50 every
session for 8 sessions, and get an extra $400 at the end if they
attended all 8. If they miss one of the 8, the bonus money wouldn't be
paid out.

The trouble is, if someone misses session 5, we are worried that such a
model might make them give up on sessions 6, 7 and 8. If they've already
blown the big pay-out at the end, why bother continuing?

Any suggestions on how we can use incentives to encourage long-term
participation?

Thanks,

Meredith

Meredith Noble
Information Architect, Usability Matters Inc.
416.598.7770 x16
meredith at usabilitymatters.com
http://www.usabilitymatters.com <http://www.usabilitymatters.com>

Comments

30 Oct 2008 - 5:04pm
djetsa
2008

Hi Meredith,

Why don't you progressively increase the amount of money you are giving?

1 - 30
2 - 35
3 - 40
4 - 45
5 - 50
6 - 55
7 - 60
8 - 65
=====
380
+ 420 for those who attended to all sessions

If they miss session 6, the next one they will get $55 instead of $60.

Cheers.

Ricardo Grzeca

30 Oct 2008 - 7:15pm
Jeff Howard
2004

Hi Meredith,

I've conducted week-long diary studies, with half up-front and half on delivery (no dropouts), but nothing like a three month multi-session study. I'd recommend also asking this question over on the Anthrodesign Yahoo Group; it's much more focused on anthropology and design ethnography.

http://tech.groups.yahoo.com/group/anthrodesign/

Offhand, I'd imagine that progressive compensation would be too complex. Instead, build in some flexibility. Maybe give everyone who attends the first session one free pass so they can qualify for a (half?) bonus if they miss a future session, or hold more than eight sessions of which they can attend any eight to still get the payout.

Inter-session postcard reminders (with progress updates/goal) might also go a long way toward promoting repeat attendance. Something tangible that they can keep around. A set of stickers for their calendar, maybe a punched card like at a pizza place...

// jeff

31 Oct 2008 - 9:14am
Meredith Noble
2010

Thanks Ricardo. I think that sort of breakdown would be good for a
small number of repeating sessions but when you get to eight, I think
I lean towards Jeff's feeling: it might be too complicated for
people. Still a good idea though; I will keep it in my back pocket
for a shorter study.

Jeff, all great ideas, and I hadn't thought of asking anthrodesign.
Figures that I just unsubscribed a month ago :)

We're asking ourselves just how important it is to have repeat
attendance (perhaps we don't need to go to all this trouble). The
unfortunate thing is we don't know yet.

It's certainly important from a recruitment standpoint -- we don't
want to have to re-recruit every two weeks. But in terms of properly
testing the application, because it's being developed using agile,
we don't really have a sense of the dependencies between features
yet, and therefore don't know how important it is for the same
people to test the application week after week.

At this point we're thinking that to keep things simple we might
just make sure the immediate incentive is enough to make people want
to come as often as possible, and leave it at that, no bonus at the
end. We'll probably have a mixture of repeat people and new people
in each test, and that may be a good thing.

Thanks for all the suggestions!

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=35086

31 Oct 2008 - 9:14am
Meredith Noble
2010

Thanks Ricardo. I think that sort of breakdown would be good for a
small number of repeating sessions but when you get to eight, I think
I lean towards Jeff's feeling: it might be too complicated for
people. Still a good idea though; I will keep it in my back pocket
for a shorter study.

Jeff, all great ideas, and I hadn't thought of asking anthrodesign.
Figures that I just unsubscribed a month ago :)

We're asking ourselves just how important it is to have repeat
attendance (perhaps we don't need to go to all this trouble). The
unfortunate thing is we don't know yet.

It's certainly important from a recruitment standpoint -- we don't
want to have to re-recruit every two weeks. But in terms of properly
testing the application, because it's being developed using agile,
we don't really have a sense of the dependencies between features
yet, and therefore don't know how important it is for the same
people to test the application week after week.

At this point we're thinking that to keep things simple we might
just make sure the immediate incentive is enough to make people want
to come as often as possible, and leave it at that, no bonus at the
end. We'll probably have a mixture of repeat people and new people
in each test, and that may be a good thing.

Thanks for all the suggestions!

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=35086

Syndicate content Get the feed