Well, interesting question indeed. But not all journal are listed in SCI and the sources for IF are very different. It's still unclear to me but certainely the uses of the parameters to compute the IF are also different!
As far as my knowledge about the impact factor is concerned, the impact factor of a journal depends on how much frequently its articles are cited in the original research articles. More the articles are cited from a particular journal, higher will be its impact factor.
For example, NATURE journal has articles based on fundamental physics and its articles are cited in most of research articles published in different journals. As a result it has higher impact factor.
May be it is not the only parameter for calculating impact factor for a journal.
there are journals indexed in SCI with IF of about 0.2 in 2012 (e.g.- Multidisciplinary respiratory medicine-0.15, INTERNATIONAL JOURNAL OF MATERIALS & PRODUCT TECHNOLOGY- 0.258)
Some jurnals with higher IFs are not listed (e.g.- Global Research Analysis- 0.2714, Indian Journal of Research- 0.3208)
OK so these 2 are not listed by WoS or the JCR. It is likely that they do not satisfy Thomson Reuter's criteria to be indexed. This could be that they do not publish regularly enough, there are not enough existing citation links to titles in WoS, the peer review for the journal may not satisfy TR's requirements or they are not deemed to be at a level of quality (what ever that means).
The fact that Global Research Analysis publishes an impact factor makes me worry. It actually doesn't have an Impact Factor. You only get one granted to you by TR which they don't.
That they put this fictitious impact factor on the homepage makes me concerned about the ethics and quality of the publication ...and perhaps therein is your answer to why it isn't indexed.
Impact Factor uses Thomson Reuters (ISI Web of Knowledge) citation data. Obviously, if a journal is not listed in ISI Web of Knowledge CAN NOT OBTAIN an IF (at least in the sense that we know it).
Journals like Global Journal for Research Analysis are really tricky. Its so called IF has not anything common with IF added by TR. It only express frequency of citations of this journal in country where this journal was established. In conformity with previous comment try to find these journals on WoS... in JCR... Result? .... zero!
Maybe the list is very old, they should update it! Maybe there are some other things effect the list. Whatever, we all know which one is good and which one is garbage, I think that's enough.