To convert paper survey responses to SPSS with PaperSurvey.io, prepare the questionnaire in PaperSurvey before collection, print its generated forms, then scan and review the completed responses. You can export the resulting data as an SPSS .sav file or as CSV.
PaperSurvey only processes forms generated by PaperSurvey. It cannot automatically read a batch of completed questionnaires created in Word, another survey platform, or a Remark workflow. Importing a blank Word, PDF, or text questionnaire brings its questions into the builder; you must check them and generate new forms before respondents fill them in. See the form compatibility guide.
If you already have completed external forms, use a separate capture or data-entry workflow for those responses. Plan a PaperSurvey pilot for your next study or collection round. Teams considering that move can also review the Remark Office OMR comparison.
Plan data quality before choosing an entry method
Typing each response into SPSS takes time and can introduce errors. Double data entry, in which two entries are compared and differences reconciled, is one way to check the transcription. Scanning can reduce typing, but it still needs a review process appropriate to your study.
Plan for the same data-quality questions whichever method you use:
- Traceability: keep a response identifier that lets a reviewer find the original form.
- Review effort: include correction and checking time when estimating the work for a batch.
- Consistent coding: define answer codes, blanks, and invalid responses before collection.
- Early feedback: pilot the questionnaire so ambiguous items can be corrected before the full print run.
PaperSurvey keeps scanned answers available alongside recognized values, applies your configured coding, and flags answers for review according to the recognition settings. Automatically accepted answers can still contain errors, so include spot-checks or full review where your protocol requires them.
From printed questionnaire to scanned image
The starting point is a new PaperSurvey-generated form printed on ordinary paper. Create or import your questions, check response options and codes, and generate a few copies for a pilot. Complete them with practice answers, including blanks, multiple selections, and handwritten fields, then scan and review them. Check a pilot export in SPSS before printing the full study batch. See how to create a paper-based survey for the design steps.
Getting completed forms back in is flexible:
- Office scanner: batch-scan a stack to PDF or TIFF and upload the file.
- Email-in and Dropbox: forward scans to a dedicated address or sync a folder.
- Shared upload page: let field staff or teaching assistants drop images without an account.
- Phone capture: scan a form with a phone's document scanner and email it in when no scanner is nearby.
Generated pages carry identifiers that support matching pages to responses. Check page and response counts after upload, especially when a page is damaged or an identifier is unreadable.
How checkbox, handwriting, and numeric items become variables
Once pages are in, recognition maps each question to a variable. Different item types are read by different engines, and their accuracy characteristics matter for analysis:
- Checkboxes and bubbles: optical mark recognition reads designated answer areas on generated forms. Faint marks, corrections, and multiple selections can require a reviewer to interpret the original. See how optical mark recognition software turns marks into data.
- Handwritten text: AI transcribes cursive and printed handwriting in supported languages. Test the languages and writing styles you expect, and review the transcriptions against the scans.
- Numeric fields: numeric recognition reads written ID numbers, ages, and scores into numeric variables. Check them during review, since one misread digit changes the value.
Recognition settings control which text answers are accepted or flagged. Flagging does not guarantee detection of every error. More detail on the end-to-end pipeline is covered in processing paper surveys with scanning.
Variable labels, value labels, and .sav vs CSV import
The difference between a usable SPSS file and a pile of numbers is metadata. Because you define each question and its answer options when you build the survey, that structure carries through to export instead of being reconstructed by hand.
- Choose variable names using each question's Column Header setting, then check them in the exported file.
- Check answer codes and labels against your study codebook, including any recoded values.
- Review variable types, measurement levels, and missing values in SPSS before running analyses.
Use the SPSS .sav export when you want an SPSS data file. CSV is useful when you prefer to configure the import yourself; check types and apply variable and value labels from your codebook after importing it. PaperSurvey also offers Excel files, R import scripts, and PowerPoint chart exports. The export guide lists the formats and settings.
Inspect a sample SPSS file and codebook
Download the SPSS example dataset and its codebook to inspect the variable structure before your next collection. The 18 fictional training-feedback responses cover three sessions, two trainers, ratings, multiple selections, comments, and missing answers.
These are synthetic records passed through the export workflow, not responses recognized from scans. They do not establish recognition accuracy or processing speed. The sample-data walkthrough includes Excel and CSV versions, intended answers, and an illustrative review log for checking how corrections and missing values affect the final data.
Verifying flagged marks before you analyze
For flagged answers, compare the scanned answer with the proposed value and confirm or correct it. Leave genuinely missing answers distinct from invalid or unreadable marks according to your coding rules.
The number of corrections depends on scan quality, handwriting, form design, and recognition settings. For written fields, each question's recognition setting decides what is flagged: Good accuracy flags readings below 90% confidence, and Verify always flags every response, which suits identifiers and primary outcome fields. Validation rules separately flag issues such as two marks on a single-choice item. Because unflagged answers can still be wrong, check a documented sample of accepted values against the scans and report that procedure in your methods section.
Handling open-ended and numeric items
Mixed instruments are common, and both non-checkbox item types export as their own variables:
- Open-ended responses become text variables. Bring the verbatim answers into SPSS as string variables for coding, or export to Excel or Google Sheets for qualitative and thematic analysis.
- Numeric write-in items such as age, count, or exam score can be collected with number fields. Check variable types, ranges, and missing-value codes in the export before computing results.
- Hybrid collection is supported when part of your sample answers on paper and part on the web. Both methods can feed the same dataset; check collection-mode information and any repeated submissions before analysis.
A REST API, webhooks, Zapier, and Google Sheets integration can send responses to other systems. Check plan features for the access and capacity your workflow needs.
Built for research data governance
Before collecting research data, review the security statement and AI disclosure against your institution's requirements. Request a DPA where needed and agree on access, retention, identifiers, and review procedures. Include these checks in the pilot before distributing participant forms.
Try It Free
Start with one blank questionnaire for your next study round. Start your free trial, import or create the questions, generate practice copies, and check the recognition and review workflow. Plan the SPSS export with your analyst so the final variables, labels, and missing-value codes match the study codebook before full collection begins.