Impact factor of a journal depends on number of citation per number of papers published in a year or during 5 years. Thomson Reuters provide journal citation report each year and it is considered as official impact factor source all over world. Though impact factor has been considered as a potential parameter to evaluate or grade any research activity now a day H-index is being more popular. Because it doesn't make any sense that one has cumulative impact factor of 100 but his/her research doesn't have any profound influence on other research activities i.e. nobodies citing those research reports worth 100 IF.
H-index evaluates personal influence of a researcher on science community by calculating individual citation per paper and number of paper he/she published.
Therefore, H-index is going to overtake IF in this sense. This is what I think.
Even H-index is a biased one. Because, if a journal have free full text option, they tend to have more citations and more H-index when compared to a journal which charge for the full text.
I feel that eigenfactor is good among all the rating scales.
see various papers by Ludo Waltmann et al. here on RG, e.g. Also recommended: papers from Woeginger et al about the metric properties ('characterisation') of h-index and close associates of it
Article Generalizing the H- and G-Indices
Article Woeginger's axiomatisation of the h-index and its relation t...
Article A systematic empirical comparison of different approaches fo...
H Index is best parameter to evaluate a journal. This is the most conveniently known by h5 score in Google Scholar Metrics (GSM). The other parameter is Median h score, which is also a good parameter to evaluate a journal. The Journal Impact Factor is a complteley meaningless average cites score in old "legacy" databases for all the articles published in the last 5 or 2 years. It is far worse than the h5 median score. According to Stephen Curry, "If you publish a journal that trumphet its impact factor in emails, you are statistically illiterate".
As per the recent information, Source normalized impact per paper (SNIP) – The measure is calculated as SNIP=RIP/(R/M), where RIP=raw impact per paper, R = citation potential and M = median database citation potential. Are considered by far the best measures and introduced by Elsevier in 2012.