Rate 63 Short English Passages for a Research Study (90 min)

We are conducting a research study on machine-generated text and require independent human readers to evaluate short English passages. This is a one-time task with a fixed price.

the task:
- you will receive one html file containing 63 short passages, each approximately 100-120 words long. Each passage will be preceded by its preceding text for context.
- A comprehensive one-page guide will be provided, detailing the scoring criteria.
- For each passage, you will assign two scores on a 0-10 scale: surprise (how unexpected the passage is given the preceding text) and coherence (how well the passage holds together as text).
- The guide includes clear anchor points for both scoring scales to ensure consistency.
- Upon completion, the page will generate a small text file (JSON) which you will send back to us. This JSON file is the sole deliverable.
- The estimated time for completing the entire task is 60-90 minutes.

what we need from you:
- native or fully fluent english reader with a strong habit of reading literary prose. Ideal candidates include editors, proofreaders, or teachers of English/literature.
- Your own genuine first-impression reading. It is crucial that you do not use any ai tools (such as chatgpt, claude, gemini, etc.) Or conduct web searches for any part of this task. The study aims to compare human and machine readings, and any AI-assisted rating would render the data unusable and cannot be compensated.
- No prior knowledge of AI or research methodologies is required. We are happy to provide an explanation of the study's objectives after you have delivered the results.

Please include a brief sentence about your usual reading habits and confirm your agreement with the no-AI rule in your proposal. We will provide the necessary files immediately upon acceptance and process payment upon successful delivery of the JSON file.