[Contents Page] [Previous section] [Next section: Appendix] Back to my homepage
Using recent theses, we have tested and confirmed the Barnett equation as a description of the age distribution of citations. The formula is valid regardless of whether we count all citations, or merely apply it to the bibliography (provided that the bibliography is complete, but not over-stuffed).
The estimated curve for Social Sciences was as expected. However the sharpness and proportion of the curve for Humanities subjects was not as the original model postulated, because the rate at which both areas adopt innovations is not significantly different.
The rates at which sources are disadopted are quite different. From our sample, Humanities resources obsolesce at half the rate of Social Science resources. This has a direct impact on the age at which printed items may be put into off-site storage.