Options
Design and application of Self-Generated Identification Codes (SGICs) for matching longitudinal data
Publikationstyp
Conference Paper
Date Issued
2016-09
Sprache
English
TORE-URI
Citation
Annual Conference of the European Society for Engineering Education - Engineering Education on Top of the World: Industry-University Cooperation, SEFI: (2016-09)
In this article, we presented a list of criteria for high quality SGIC-items and discussed the advantages and disadvantages of our chosen items for the purpose of pre-/post-testing and possible application in long-term study designs. We compared a handwritten and a bubble-sheet version of code forms and presented the algorithm used for matching. With accepting off-1 matches, we could improve our matching rate to 76 % compared to 72 % matching via official matriculation numbers. A handwriting analysis indicates no false-positive and only 8 % false-negative matches. Based on our results, we conclude that the advantages of using SGICs can outweigh the disadvantages, providing that the items and the matching algorithm are chosen carefully. We thus encourage other researchers to consider using SGICs for their data linkage.
Subjects
Matching algorithm
Subject-generated coding