Typically, "flawed" data are identified through comparisons between multiple data sources for the same or closely related properties. An example would be data for a mixture for which the endpoints show very large deviations from numerous literature sources. If the endpoints are grossly in error, the mixture data cannot be correct. In such cases, the compound may be mis-identified in the article, but unfortunately, such errors can rarely be repaired.
"FLAWED DATA" are identified in the Reject column of data tables in TDE.