Milestone 3 Sky TrustIdea 1: Collective Statistics Panel (CSP)

From crowdresearch
Jump to: navigation, search

Abstract

Having search based on requesters reputation would increase the trust of worker ---> requester. So we need to give workers this opportunity to search HIT according to their trust level. According to our need finding in milestone 2 workers wants to trust two things. First Requester's intention and second HIT description.

We believe statistics that come from collective data posses valuable information that could be used by both workers and requesters in order to establish trust.

Section 1: Realtime HIT Data

As we mentioned we need something to fill the gap of trust between workers and requesters using a more precise description for HIT using collective statistical data. We can have a panel that have many statistics. This panel must be real time based on previous HIT attempts by collecting the following data:

  • Average completion time : time that took previous workers to complete
  • Average Handy score of HIT
  • Safety score of HIT
  • Clarity score
  • Fun score:
  • etc.

This panel contains data based on previous HITs so they are trustable. we will explain it in details in the next step.


Section 2: Realtime Requester Data

One other collective data gathering is about having realtime statistics for:

  • Requesters approval rate ( in all of his/her previous HITs)
  • Number of approve/rejection for each HIT (average is not enough)
  • Response rate
  • Worker's rating average
  • Number of total HITs
  • Payment and average hourly payment (payment /average completion time)
Section 3: Realtime Worker Data

In this section we provide realtime collective statistical data on different workers based on:

  • Completion rate
  • Requester review
  • Payment rate
  • Approval rate
  • Response rate

Objective

  • Building collective trust overtime
  • Helping workers to trust the descriptions provided
  • Helping workers to trust requester payment intention
  • Helping to get fair payment
  • Workers can make sure that HIT is clear before starting the task
  • Predict requester response rate in future

implementation

We propose three coherent methods in order to collect these information:

  • A survey upon HIT completion for workers
  • Gathering requesters footprint overtime
  • Gathering workers footprint overtime
  • Gathering HIT data until final deletion
  • Review and rating

Implementation notes

NOTE : The most important gathering tool is the survey that will be filled immediately after HIT completion by workers. We ask workers about HIT related experience (e.g. Section 1: Realtime HIT Data ). There could be more parameters extracted from these data based on specific needs.


= UseCase

Statistics Panel & pop-up window with Mouse-over