WinterMilestone 1 himangi774

From crowdresearch
Jump to: navigation, search

Experience the life of a Worker on Mechanical Turk

Reflect on your experience as a worker on Mechanical Turk. What did you like? What did you dislike?

I am based in Hyderabad, India and have used the worker sandbox.

It was fun finding tasks and completing them. Its an easy to understand interface. What I liked the most was that some tasks had a very less acceptance turnaround time and I was able to make $1.20 within a couple days.

There were some drawbacks thought. I found some hits that required qualification, however there is no clear explanation of what a qualification is and how to obtain it. There is a qualification tab but I am not allowed to request any of the seven qualifications, nor do the qualification names make much sense. Another issue is there are some tasks that I have completed and required considerable time have not received any acceptance/rejection response after 4 days. This is a serious concern for freelancers and workers who might be expecting the return.

Worker.png


Experience the life of a Requester on Mechanical Turk

Reflect on your experience as a requester on Mechanical Turk. What did you like? What did you dislike? Also attach the CSV file generated when you download the HIT results.

I have used the requester sandbox here.

The creation of a task was an interesting process. A lot of information was needed to be filled in for the registration and payment modes, which is essential to authenticating requesters. Also, they provide templates for the projects and allow creation of HITs very smoothly. The requester only needs to worry about the content leaving the formatting and other concerns to Mechanical Turk.

I had applied a timer on the tasks, but I noticed people completing it in time frames not practical. I guess the HIT is visible before and the worker has the answer before the timer starts(when he clicks - accept HIT).

File:Results himangi.csv

Explore alternative crowd-labor markets

Compare and contrast the crowd-labor market you just explored (TaskRabbit/oDesk/GalaxyZoo) to Mechanical Turk.

TaskRabbit : The nature of tasks is quite different. While we have tasks collective micro-tasks like classification and annotation at Mechanical Turk, TaskRabbit has labour tasks like moving, cleaning, delivering etc. Thus, while MTurk has digital tasks we have onsite mechanical tasks requiring more labour at TaskRabbit. Thus, task completion is very unambiguous and requesters cant deny it.

Upwork (Odesk) : The nature of task here is that of big programming projects. Its a connect based system with a membership program. Often the requesters interview the potential workers. They are long-term projects, often with milestone type increments, thus requiring higher expertise and committment.

Readings

MobileWorks

This is a mobile-based crowd sourcing platform.

  • What do you like about the system / what are its strengths?

1. Accessibility : It is available to the low-income sections of the society.

2. Automated quality checks: They distribute tasks to multiple workers and check for match.

3. Accuracy: Something like a reputation score for the users helps decide payments and trust results for the requesters.

  • What do you think can be improved about the system?

1. We can have a qualification set for each task, and match worker qualifications before offering the task to a specific worker.

2. Tasks can be prioritized and the system can intelligently attract workers to high-priority tasks.(like displaying them at the top of the list for most workers or increasing pay as the deadline approaches)

3. Have a highly responsive design as the workers might have different smartphones probably with extremely small screens causing strain.

Daemo

  • What do you like about the system / what are its strengths?

1. Prototype task workflow: A worker is expected to complete the prototype task, develop better understanding of the larger task and contribute to improvement of the problem statement.

2. Rating the workers based on their performance and feedbacks, thus helping build trust in the community.

3. A democratic system for taking decisions and policy-making where workers, requesters and experts have an equal say.

  • What do you think can be improved about the system?

1. A seamless system for communication between the workers and requesters - something like an IRC integrated.

2. Have a qualification set for the workers and match tasks based on experience/qualifications.

3. Allow teams to work on tasks.

4. The rating system can take in feedback from peers in case of multiple workers collabrating on a task.

Flash Teams

  • What do you like about the system / what are its strengths?

1. Team composition system : It allows co-ordination of on-demand expertise by creating dynamic teams with people working remotely from around the globe.

2. The tasks have well-defined modules that enable collaboration and completing tasks on time with teamwork.

3. The concept of Blocks with an expert and DRI(Directly Responsible Individual) per team helps in the management of the tasks better.

  • What do you think can be improved about the system?

1. There should be a good mechanism for credit assignment as there can be conflicts when tasks are reassigned. Also, a conflict resolution policy should be framed.

2. Have a good time-management scheme and deadlines for the tasks. There can be deadline extensions with penalties and the worker can be asked to request on that. This is essential as some modules might be dependent on others and hence blocked.

3. We can have a rating system that especially counts team-work and coordination apart from work efficiency. This will help create more compatible teams that will have less differences.

Milestone Contributors

@himangi774