A method for data input in which marks made on preformatted documents are sensed by photoelectric means. The marks are interpreted as either characters or values according to their position on the form. This provides an efficient type of data input in applications in which there are relatively few answers to a limited number of questions. No further data preparation activity or machines are required since the reader can be connected directly to the system. Special readers are available to handle large documents. OMR documents are printed on paper, and information to the user as to where the marks should be put and what significance they will have is usually printed in a colour that will not be detected by the photoelectric sensors. Machine-readable timing marks are preprinted along the edge of the document.