The dielectric losses of a material are directly related to its resistance. High losses means low resistance. The losses are often expressed with the imaginary part of the dielectric permittivity e''. The quotient of the imaginary and the real part of the dielectric permittivity (e''/e') represents the loss tangent. The loss and loss tangent depend on your material and the material processing parameters. It is not possible to predict these parameters reliably for a given sample. I suggest you to measure e'' and e' of your samples using dielectric spectroscopy. Personally, I find the values of real dielectric permittivity e' and the resistivity of a sample the most meaningful parameters. The loss tangent is difficult to be interpreted, because it is a mix of permittivity and losses. Personally I don't like it. You are loosing information by mixing e' and e'' together.

Agreeing with the previous note. Inorganic materials exhibit dielectric loss mainly because of crystal defects, which act as dipoles. If the material is semiconductor (e.g. mixed valences combined with hopping or band conduction) you will have both localized loss and conduction loss. Both can be strongly frequency dependent. Ideally you should study single crystals for identifying the local defects, but most oxide samples are polycrystalline or powders. In this case you will have phenomena related to interfaces (e.g. interfacial polarization, surface conduction etc:) which may complicate the picture. Even in the case of single crystals you will have additional capacitive and resistive components from the electrode contact (charge injection, or blocking electrode etc.). Anyway, it is advisable to use possibly wide frequency range, and to use Cole-Cole plots (or Z' vs. Z" plots) to distinguish relaxation processes. Sometimes the electrical modulus picture works better than the permittvitiy represeantation. Anyway, it is worthwhile to study first the literature on the specific oxide you want to investigate.

The problem with ferric oxide that most articles found in google using the keywords "iron oxide dielectric loss" are related to ferrites rather than pure ferric or ferrous oxide. Nevertheless teh methodolgy may be useful for you

Laboratory measurement of dielectric constant and loss tangent of ...

www.annalsofgeophysics.eu/index.php/annals/.../4788

http://prr.hec.gov.pk/Chapters/1861-7.pdf

http://panda.unm.edu/Kenkre/papers/Art116.pdf

https://pangea.stanford.edu/departments/geophysics/dropbox/SRB/public/docs/theses/SRB_024_NOV84_Knight.pdf

You can easily find further referemces on the internet.

Fully endorsing Rainer Schmidt saying about tan loss, scientifically its a bit trcky to handle loss tan as explained by the presence of defects and competition between localized and delocalized charge carriers by Gyorgy Banhegyi. However, from engineering and device formation point of view tan loss is very very important.