Statistical Data Converter

Open SAS, SPSS, and Stata datasets and export complete UTF-8 CSV or JSONL with an optional metadata JSON sidecar.

Dataset file

Files stay in the conversion workspace for up to six hours. The upload limit is 1 GB.

Source format

Dataset summary

Bounded preview

Only the preview is capped. The completed export includes every row.

Statistical metadata

Export complete dataset

Output format

Values

Dates and times

CSV delimiter

Missing values

Include the source variable names as the CSV header Export the metadata JSON sidecar

CSV and JSONL do not inherently preserve variable labels, value-label dictionaries, tagged or user-defined missing values, display formats, or exact numeric and date metadata. Keep the sidecar when those semantics matter. External SAS catalog files are not resolved.

Supported statistical formats

SAS7BDAT, SAS XPT/XPORT, SPSS SAV, compressed ZSAV, SPSS POR, and Stata DTA versions 104 through 119 are read locally with pyreadstat and ReadStat.

Complete exports, bounded memory

Rows are read and written in chunks. If the complete text export cannot fit in the workspace, conversion stops with a clear error instead of returning a truncated file.

Statistical Data Converter

Inspect supported SAS, SPSS, or Stata datasets and export complete tabular values with metadata.

Input: One supported .sas7bdat, .sav, or .dta dataset.
Output: Dataset metadata plus CSV or JSONL rows.
Real limitations: Format versions, compressed encodings, value labels, missing-value semantics, and very wide data can limit fidelity or support.
Privacy and data handling: one supported .sas7bdat, .sav, or .dta dataset is uploaded to a compatible Convertr worker only after you start the operation. Processing is automated, and temporary source and result files are deleted automatically after processing rather than retained as permanent storage.

How to use this tool

Choose one supported .sas7bdat, .sav, or .dta dataset and wait for the page to validate it.
Set the available output or inspection options, then start processing.
Review the reported result and download dataset metadata plus CSV or JSONL rows before the temporary link expires.

Common failure modes

An unsupported version, damaged header, unknown encoding, or inconsistent row data can prevent parsing.

Useful example

Upload survey.sav to inspect variable labels and export all response rows to CSV.