Study participation opportunity: usability evaluation for OpenDP's DP Wizard!

[Joe Near]

OpenDP and the University of Vermont are conducting a remote usability
study to evaluate OpenDP’s DP Wizard tool for differentially private
data releases. We are looking for participants who:

- Are at least 18 years
- Reside in the US
- Can speak and read English
- Have prior data analysis experience with Python

In this study, you will be asked to:
- Watch/read a tutorial about a data science tool interface
- Complete a series of data analysis tasks using the interface on a shared screen
- Speak out aloud your thought process while you complete the tasks
- Complete post-task survey and interview

All study procedures will take approximately 1 hour via a Microsoft
Teams meeting, the study session will be screen and audio recorded.
After completion, you will receive a $40 electronic gift card of your
choice (e.g., Amazon.com, Walmart) as a thank-you for your time. If
interested, please fill out this online eligibility survey:

https://qualtrics.uvm.edu/jfe/form/SV_dnaZyLvHAwnJaIK

If you are selected to participate, we will reach out to you via email
typically within 2 weeks. If you have additional questions about
participating in this study, please contact the principal investigator,
Onyinye Dibia by email (Onyiny...@uvm.edu).

Thanks!

[Howard Smith]

To Whom It May Concern:

The following question is not fully specified. Releasing two differentially private statistics, one with ε₁ = 0.1 and the other with ε₂ = 0.5, results in a total privacy loss of:

The total privacy loss (ε_total) from releasing multiple statistics depends on whether they are computed on:

1. The same dataset (sequential composition)

2. Disjoint datasets (parallel composition)

   Scenario	Total Privacy Loss (ε_total)
   Same dataset	ε₁ + ε₂ = 0.6
   Disjoint datasets	max(ε₁, ε₂) = 0.5

Both of these could be correct and they are both available in the answer choices.

--
You received this message because you are subscribed to the Google Groups "opendp-community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to opendp-communi...@g.harvard.edu.
To view this discussion visit https://groups.google.com/a/g.harvard.edu/d/msgid/opendp-community/87h5zx81xr.fsf%40uvm.edu.