Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 185
APPENDIX E High-Speed Scanning Technology This appendix describes the use of high-speed scanning for data entry from paper forms. This technology represents, in some ways, the leading edge of survey forms handling, and the techniques and tools described here are still being developed and improved. For the latest information in this area, the reader is directed to the Internet for further research. Data Entry by Scanning The process of scanning forms as a data entry method has been around since the days of punch-cards and early development of technology for school standard testing. Students would complete a form consisting of columns of multiple choice answers and signify their response by darkening the appropriate box or oval. Things have come a long way since then. The old structured format of a fixed pattern of boxes or ovals that can be filled in has been replaced with free-form capabilities, and the scanning speeds and hardware have improved to allow thousands of forms to be scanned--both sides--in a matter of hours. The following sections describe each of the components of this high-speed scanning system. There may be differences in available products and services, but the main principles will still apply. Hardware The first component is the scanner itself. These are not ordinary office scanners, but special purpose scanners that are designed for high volumes of work. The hardware includes a hopper for thousands of completed forms. Each form is fed into the scanner, where a scanning pass takes one second or less. The page is then reversed and the other side is scanned. The scanned images are then processed. Software The software that comes with these systems provides the user with the ability to program the areas of the forms to be scanned and information on what to with the results. There is a certain degree of character recognition built into the software. The requirements for legibility and well-formed letters can be quite high. E-1
OCR for page 185
E-2 Guidebook for Conducting Airport User Surveys For questions with tick-box answers, the process is relatively straightforward and also rela- tively lenient as to whether a box is ticked or not. For written responses, the optical character recognition (OCR) software can be quite crude. Depending on the particular service or system being used, those completing the forms may need to be trained on how to enter each letter and number. Sloppiness at the time of the interview translates into reduced productivity in converting the data into an electronic file. Interview surveys with many free-form entries and time constraints or respondent self- completed surveys are not recommended for such a process due to the inherent sloppiness of the responses. Scan Design Process Perhaps the longest part of the process is the scan design. This step only takes place after the survey form has been finalized. The software is then used to designate each question on the form, where the responses are, and what each tick box means. In this manner, each question and each response is linked to an entry in the resulting database. This process can take several hours and must be thoroughly tested. Purchase or Rent It is possible to buy the hardware and software or purchase such services from a local agency. A recent search on the Internet revealed a number of businesses that offer high-speed scanning and data entry services. Careful research would be necessary to match the capabilities of different vendors with the requirements of the survey. Purchasing of the hardware and software should be considered if large volumes of surveys are done on a frequent basis. This equipment is relatively expensive and the cost would most likely not be justifiable over a single survey project.