Use AI Chat to Select Indicators
AI Chat helps users turn research questions into candidate data sources and indicators. It is most useful before export.
When to Use AI Chat
| Situation | How to ask |
|---|---|
| Unsure which source to use | Describe the research question, unit of analysis, geography, and period. |
| Need dependent and explanatory variables | State the intended model and ask for indicator suggestions. |
| Coverage is incomplete | Provide the indicator name or short code and ask for proxies. |
| Need logic checks | Ask AI to identify frequency, unit, endogeneity, or missing-data risks. |
| Need reporting help | Ask for a data summary, chart idea, or export workflow. |
Prompt Examples
I want to study the effect of FDI on provincial economic growth in Vietnam from 2010 to 2023.
Suggest the dependent variable, explanatory variables, controls, Ecodata sources, and coverage checks before export.
I selected these indicators: GDP growth, FDI inflow, labor force, CPI.
Check whether they are suitable for a province-level panel model and suggest additional controls.
AI Chat -> Dashboard -> Export
- Ask AI Chat for candidate sources and indicators.
- Go to Dashboard or the relevant module to search for indicators.
- Open indicator details and review metadata.
- Copy selected short codes.
- Return to AI Chat to validate model logic and coverage.
- Preview data.
- Export data with metadata.
Combining With Econometrics
After selecting indicators, open Econometrics to run a preliminary model. AI Chat can help interpret results, suggest missing variables, or flag assumptions to verify.
Notes
AI Chat supports selection and interpretation, but it does not replace metadata review or source-data validation. Research conclusions should rely on previewed/exported data and a properly tested method.