The aim of this article is to present a trend in research on measurement error in survey data and to suggest some problematic aspects of this approach. The article describes the Multitrait Multimethod experimental design and its modifi cation into a 2 Split-ballot Multitrait Multimethod (2 SB MTMM), which is used for experimental data collection in the European Social Survey. The text shows how to analyze 2 SB MTMM data to obtain estimates of construct validity, reliability and common method variance for a single questionnaire item, and how to make use of these estimates. It also points to some problems encountered in 2 SB MTMM data analysis., Johana Chylíková., and Obsahuje bibliografii