What is the Difference Between MircoBatch and HyperBatch?

MicroBatch and HyperBatch are both ways to divide your study into multiple HITs that cut MTurk's fee by 50%. However, MicroBatch and HyperBatch run your study in different ways, either slowing down or speeding up data collection.

CloudResearch's Batching Tools

MicroBatch and HyperBatch are tools that break your study into multiple HITs, each with 9 assignments or less. Doing so, reduces Mechanical Turk fees from 40% to 20%. However, the primary difference between these batching tools is that MicroBatch runs HITs sequentially while HyperBatch runs HITs simultaneously.

MicroBatch vs. HyperBatch

When using MicroBatch, each HIT in your study is run in sequence (see top part of Figure 1). This means when your study is launched, the first HIT is available. The second HIT doesn't become available until 9 people have completed your study. Once 9 people complete the first HIT, HIT number 2 is launched. The study will proceed one HIT at-a-time until you reach your intended sample size.

The sequential nature of MicroBatch can be used to slow down data collection. While the sequential nature of launching HITs spreads out data collection some, you can also set a delay between when new HITs are launched. Setting a delay can help spread your data collection over the course of the day or several days in a week.

Unlike MicroBatch, HyberBatch runs all HITs in your study simultaneously (see lower part of Figure 1). This means some workers may complete HIT 2 while other workers are completing HIT 5 and all assignments have been completed for HIT 4. The simultaneous nature of HyperBatch ensures the fastest possible data collection by making your study available to all workers who may want to take it at the same time.

microbatch vs. hyperbatch