Optical mark recognition (OMR) reads marked answer areas, such as checkboxes and bubbles, on paper forms. It identifies which answer areas contain marks; it does not read written comments by itself. Text and number fields need separate recognition methods.
Where OMR is used
OMR is used for surveys, tests, quizzes, evaluations, and other forms with structured answer choices. PaperSurvey combines mark recognition with handwriting recognition on forms generated by PaperSurvey.
Examples include:
- Course feedback forms
- Event satisfaction surveys
- Customer satisfaction surveys
- Resort evaluation surveys
Prepare a form for scanning
Create your questionnaire in the survey editor, inspect the paper preview, and generate the printable form. PaperSurvey adds the corner marks and page codes needed to connect scanned answers to the questionnaire. Importing questions from an existing document does not make previously completed copies of that original document scannable.
Test-print and complete a few copies before distributing a full batch. Scan the whole page clearly, with all corner marks and QR codes visible. Upload the files and compare the recognized answers with the scans. Use that pilot to check the layout, marking instructions, and review time.
Choose marking instructions that match the settings
PaperSurvey supports two marking conventions:
- Check or fill the box to respond. A cross or a filled box can indicate a selected answer. Filling a box is not a cancellation instruction in this mode.
- Check to respond, fill to unmark. A cross selects an answer; completely filling the box cancels that selection. Use this convention only when it matches the survey’s recognition settings and printed instructions.
Explain the selected convention before respondents begin and ask them to use a blue or black pen. Include corrected marks in your pilot, because a cancellation may be ambiguous even when the instructions are clear.
Read written answers as well as marks
OMR handles the selected options. Other recognition methods handle text:
- OCR, or optical character recognition, usually refers to reading machine-printed text.
- ICR, or intelligent character recognition, is commonly used for handwritten characters, including characters in separate boxes.
- HTR/HWR, or handwriting recognition, transcribes handwritten words and longer responses. Results depend on the handwriting, language, answer space, and scan quality.
PaperSurvey can transcribe supported written-answer fields and retain their source images for review. Choose recognition and verification settings for each field rather than assuming every kind of answer uses the same method.
Review before using the results
Recognition can misread a mark or a word. Validation and confidence settings flag some uncertain answers, but a high confidence score does not guarantee a correct result. Compare flagged answers and a sample of accepted answers with their source scans, correct errors, and expand review when the pilot or your study protocol requires it.
After review, explore the results in the app or export them to your analysis tool. Start a free trial to test a small batch with your own questionnaire and equipment.