In the paper with Rosalie Gorter and Jos Twisk (BMC Medical Research Methodology, 15, 2015) we show the negative impact of using sum scores as outcomes in a longitudinal data analysis in comparison to the use of item response theory scores. We have used plausible values for the scores for the longitudinal data analysis to avoid effects of scale differences. The presented methodology makes it possible to get all estimated scores on the same scale (measurement analysis) and ready for secondary analysis (structural analysis).
The paper is already highly accessed (over 2000 views).
Gorter, R., Fox, J.-P., and Twisk, J.W.R. (2015). Why item response theory should be used for longitudinal questionnaire data analysis in medical research. BMC Medical Research Methodology, 15(1):55. DOI:10.1186/s12874-015-0050-x. (Highly Accessed)
Open Access: http://www.biomedcentral.com/1471-2288/15/55