A point has recently come up of describing the tendency of some authors to include many references in their bibliographies that they have not actually consulted themselves. Often, these references are taken from other, cited work, as can be seen in the faithful preservation of spelling errors, number transpositions, or even just non-existent works. These references give the impression that the author is quite widely read, and some of them may be used as a scientific basis for "facts" that are not, in fact, true.
I came across one of these when a student was looking for a reference for that statement that people retain some percentage of what they read, more of what they hear, and even more of what they do. She found some references, but was unable to find the source given. At that time we looked around, but were just not able to find an original study that would support this statement commonly accepted as gospel truth in didaktics.
I just found a good blog entry discussing this problem that Will Thalheimer wrote in 2006: People remember 10%, 20%...Oh Really?. The blog details his quest for the source, and shows lots of fascinating graphs of this "fact", often with "references" given.
At VroniPlag Wiki there is also a case documented that has about 2200 references listed (on 170 pages). The link is to a page that started documenting the erroneous entries, although after about E or F the group lost interest, as there were so many errors. This doctorate has since been rescinded.
I need a name for these kind of bogus or fictitious references. It has been suggested that they be called Petersilienreferenzen in German, as one often finds a sprig of parsley on one's plate in Germany in an attempt to spruce up the looks of a dish in a restaurant. I think that generalizing this to garnish references is a clearer name in English. Or is there a generally accepted term for these? Do any of my readers have good links to examples of this kind of reference?