Optical scanning of forms
The Unit for Applied Clinical Research can provide assistance in designing forms for optical scanning for use in research projects.
Optical scanning of forms
The Unit for Applied Clinical Research provides assistance in designing forms for optical scanning to be used in research projects, and offers help with the scanning of those forms after data collection. At present (des 2009) more than 150 studies have used this tool.
Optical scanning means the direct transfer of data from a written form to an electronic file. We use the program Teleform to design and scan the forms after data collection. The program compares the written data in the form with what is expected from the specifications for each field in the form. The data are then transferred to an SPSS file. Every field in the original form has a format which determines whether it should be a number, a letter, or a mark in a tick box.
When and why should you use this method?
Large studies require effective and accurate methods for collecting data. This method is particularly applicable when participants are asked to fill in forms themselves, but considerable time can also be saved when health personnel are responsible for filling in the forms. For multi-centre studies, you should consider using our remote data entry solution NetCRF. Optical scanning is most effective in large studies involving many patients, and less practical in small studies.
What can the Unit for Applied Clinical Research do for you?
We can make forms designed for scanning on the basis of your original forms. The original forms should be in Word format or another format that is easy to copy from. We can either scan completed forms ourselves if the volume of data is limited, or provide training and advice so that researchers can perform this task themselves in larger studies. It is normally possible to produce Teleform forms within two to three weeks, but if our workload is heavy or if the forms are very complicated, the process may take longer.
What do we expect you to do?
You should deliver forms in Word format or another format that makes it easy to extract text. The final layout is designed in the Teleform program. The format of tick boxes and fields for numbers or letters is determined in cooperation with the research team to ensure that scanning is as accurate as possible. The form should be proof-read carefully before transfer to Teleform, since it is much more complicated and time-consuming to make corrections in Teleform. If in doubt, we will be pleased to give general advice on how to edit forms that will do well in optical reading before you start editing your forms.
After data collection, you return the complete forms to us. Scanning and verification is carried out on our computer system. For better results, someone from the research group should scan the forms. We provide training and advice.
Verification involves accepting or changing data that the program does not recognise. It is an advantage if the person making the corrections is familiar with the study and what information is expected in each field. Carrying out the corrections also gives the research team insight into how the forms are filled in and which fields are most likely to give rise to errors and misunderstandings.
How accurate is the method?
We performed a test on 100 forms of 10 pages. The form had 335 variables and 581 fields, including both tick boxes and fields for numbers and letters. The forms were filled in by parents of young children, who should be as a representative group of adults. The results shows that the system is reliable, but not that there are no sources of error.
- 0.01 % of empty tick boxes were registered as marked
- 0.04 % of ticked boxes were not registered
- 28 % of deleted ticked boxes were registered as marked
- 0.2 % of the numbers were not registered
- 1.5 % of the numbers were incorrectly registered
Not unexpectedly, deleted ticks are often registered as marked. This can be prevented by providing more than one alternative. If another alternative is chosen, the first one is deleted and the right alternative will be registered. Otherwise, the results indicate that the method is very accurate. Letters and numbers should only be used when necessary. The best results are obtained by using tick boxes and providing all relevant alternatives on the form.
What do we charge?
Research projects that are fully financed externally will be charged for our services as will investigator initiated research projects from outside our health region. Investigator initiated research at the Faculty of Medicine and at hospitals in the Central Norway Health Region will not be charged.
Contact Berit M. Bjelkåsen for more information.