Professional Documents
Culture Documents
m.valletorre@student.tudelft.nl
ABSTRACT
Technology and innovation are creating new solutions for
all kinds of tasks, sometimes helping humans in their performance and others completely substituting the need for
their participation, but at the same time new doors are
opening, transforming the job market and economy as
we know it. Crowdsourcing is growing quickly and it is
starting to get attention as a job substitute. In this work
we analyze the literature regarding motivation and rewards of crowdsourcing in an attempt to determine if the
current situation allows this substitution to happen.
1.
INTRODUCTION
2.
Amazons Mechanical Turk or MTurk is a platform dedicated to the performance of crowdsourcing; where requesters create Human Intelligence Tasks (HITs) that the
crowdworkers execute. As mentoned before, most of the
times these HITs consist of simple annotations, labelling
of an image, audio transcriptions and many others; which
usually take less than 10 minutes and are rewarded by
payments of a few cents. MTurk allows a wide variety of
options for the execution of tasks, like restricting workers
to perform only once, or splitting a HIT in many equal
tasks for different workers so the results can be aggregated for an "average" [2].
2.1
Research in MTurk
3.
PROBLEM
The process of consolidating the concept of crowdsourcing is still in infancy and there are plenty of different
views regarding its many features, such as motivation,
design or execution; in this case we will focus more towards the reward perceived by the typically self-selected
volunteers or part-time workers contributing to the CS
community. For example, some definitions state that
this type of contribution should be done for free or significantly less than the cost of a regular employee [7],
while some state that it needs to fulfill at least one [28]
of Maslows individual needs: either the aforementioned
financial aspect, development individual skills, to simply
have fun, share knowledge, or form part of a community
[21]. On the other hand, some researchers consider that
CS has to be rewarded in an economic way [18], as well
as demographic studies point out that around 20% of the
crowdworkers on MTurk have a household income of less
than $15,000 U. S. dollars [13] (compared to the average $55,000 in the United States [31]) and rely on the
financial reward to make ends meet.
There has been work trying to pinpoint what the motivation for crowdsourcing should be and why it is important; whether it is the money, skill development [3],
task autonomy or work variety [15]. On the contrary, research towards the reward itself is scarce and while some
of them just say that money is the prime reward, for
example stating specific amounts such as remunerating
tasks with $0.01 to $0.10 [7]; others say that the feeling
of belonging, recognition or participation are enough.
Most approaches focus on how to motivate workers to get
the best quality from CS, but fall short when defining the
compensation. As mentioned by Estells and Gonzlez
[5], the aspect of reward is surprisingly not mentioned
enough in relevant literature. One big trend is to note
aspects such as that the integration of consumers in the
process, which is intrinsically the reward [29]. The problem is that there is not that much research directed towards crowdsourcing rewards as a regular job, like a new
alternative to outsourcing tasks, such as Human Resource
screening [10]. This leads to the question for this work:
Can crowdsourcing substitute a traditional job income?
In this review the goal is to find the literature available
and set base to determine if the current state of crowdsourcing, technologically and socially, would be a viable
option for a person to make a living. The fact that crowdsourcing has not been defined concretely makes it difficult
to take in account or disregard both good and bad examples of how it can substitute traditional jobs. Along
with theory and experiment anaysis, we revise different
types of CS and present examples of successful ways that
4.
RELEVANT LITERATURE
4.1
Motivation in Crowdsourcing
4.2
Studies on Rewards
5.
DISRUPTIVE CROWDSOURCING
6.
CONCLUSION
www.uber.com
www.reflik.com
8
www.airbnb.com
9
www.helloalfred.com
10
https://www.uber.com/driver-jobs
7
7.
[10]
[11]
[12]
[13]
[14]
[15]
FUTURE WORK
8.
[9]
[16]
[17]
[18]
REFERENCES
[19]
[20]
[21]
[22]
[23]
[24]
[25]