DID-ACTO / FIELD GUIDE
A plain-language glossary
The terms you are most likely to encounter in a research conversation.
Research terms
| Term | Meaning |
|---|---|
| Unit | An entity observed in the study, such as a store, region or person. |
| Panel | The same units followed across multiple periods. |
| Treatment / intervention | The change or exposure whose effect you want to investigate. |
| Cohort | Units grouped by a common treatment start period. |
| Comparison group | Observations used to represent the change that could have occurred without treatment. |
| Counterfactual | The unobserved outcome under an alternative treatment state. |
| Estimand / target | The precise effect, population and window the analysis is meant to describe. |
| ATT | Average treatment effect on the treated, for the specific supported target. |
| Event time | Time measured relative to the start of treatment. |
| Parallel trends | An assumption about comparable changes without treatment, possibly after specified adjustment. |
| Anticipation | A response to treatment before its recorded start. |
| Interference / spillover | Treatment of one unit affects outcomes of another. |
| Support | The observations and comparisons available to identify the declared target. |
| Confidence interval / band | A measure of uncertainty under the specified inference procedure. |
| Sensitivity analysis | Analysis of how conclusions behave under specified changes or departures from assumptions. |
Workflow terms
| Term | Meaning |
|---|---|
| MCP | The protocol connecting your assistant to DiD-Acto’s tools and resources. |
| Upload ID | An opaque reference to a file in your private upload store. |
| Dataset ID | A reference created when the workflow registers a dataset. |
| Plan hash | An identifier binding the exact analysis choices you reviewed. |
| Run ID | The durable identifier for one submitted analysis. |
| Exploratory / quick | A first pass for learning about the data, effect and design before a fuller study. |
| Confirmatory | A declared purpose with stricter requirements; eligibility still depends on the completed evidence. |
| Provenance | The record of the source and choices behind a result. |
| Zero-refit rendering | Creating a report from saved results without running the estimators again. |
| Quarantine | A saved record awaiting an administrator’s integrity or compatibility review before it can be opened. |