What the work involves

This is application-grounded data collection rather than abstract prompt writing. You work inside EViews on real econometric tasks — building and diagnosing ARIMA and VAR specifications, running unit-root and cointegration tests, estimating GARCH models, constructing forecast series, managing workfiles and page links, writing EViews programs — and capture what you do as high-resolution screenshots and written step accounts. Some tasks ask you to demonstrate the correct path through the interface; others ask you to judge an AI-generated answer about EViews output or econometric procedure and explain precisely where it goes right or wrong.

The recurring value you provide is menu-level and output-level specificity. Reviewers care whether you can name the exact dialog, the option checkbox, the equation object method, and what the estimation output table actually reports — not a generic description of "running a regression." Critiques of model AI answers are scored on whether your correction is econometrically defensible and tied to what EViews genuinely does in that version.

What the platform screens for

  • Verified access today. EViews installed and openable on a Windows machine you control. Mercor does not issue licences, virtual machines, or remote desktops.
  • Display capability. Screenshots above 2.5 MP, via a single high-resolution monitor at the stated thresholds or a dual-monitor pair captured as one spanning image. Two 1080p screens qualify as a pair; one alone does not.
  • Applied econometrics depth. Expect follow-ups on stationarity testing, lag selection, impulse response interpretation, and forecast evaluation — answered the way a practitioner would, with reference to what EViews reports.
  • Evaluation judgment. The ability to separate a wrong answer from an unconventional but valid one, and to write a correction someone else could act on.

Logistics

Remote and asynchronous, task-based, with work claimed from a queue rather than scheduled shifts. Contributors commonly report 10–20 hours per week, with the observed band for this listing at $45–55/hr — rates vary by task type and review outcome and are not guaranteed. Expect a screening conversation, a paid or unpaid calibration task, and ongoing quality review that can adjust your task access either direction.