I observed some very good math journals are not indexed by Thomson Reuters. The basis of my evaluation is the composition of the editorial board, the university or press that house the journals and the authors that publish with them. My question is what is likely the reason for this non-indexing?
The only mathematical journals that I think highly of and that I have noticed at any point not being indexed by Thompson Reuters where simply too recent. For example, Forum of Mathematics (both Pi and Sigma parts) is currently not indexed, but I bet it will soon be. Of course, I may have missed others.
Added in response to user's comment: you did find one I missed. More precisely, Real Analysis Exchange is known to me and not indexed. Here I guess that the issue is that the journal is rather low-profile, in the sense that its articles receive few citations on average (not to mean that its article are not good, I know much too few of them to judge even roughly). JCR only indexes the journals that receive a certain amount of citation (see the "Citation Analysis" section in the link provided in Sean Elvidge's answer). If one looks at MathSciNet, which is more comprehensive for maths, one sees that the MCQ (analogue of the 5 year Impact factor but using the database of MathSciNet) of Real Analysis Exchange is lower than half the average MCQ of MathSciNet's journals.
The selection process used by Thomson Reuters for choosing journals is described, in detail, on this page: http://wokinfo.com/essays/journal-selection-process/
The basic standards (see the link for more information) are:
Timeliness of publication is a basic criterion in the evaluation process. It is of primary and fundamental importance. A journal must be publishing according to its stated frequency to be considered for inclusion in Web of Science. The ability to publish on time implies a healthy backlog of manuscripts essential for ongoing viability. It is not acceptable for a journal to appear chronically late, weeks or months after its cover date. To measure timeliness we need to see three consecutive current issues, one after another, as soon as they are published.
INTERNATIONAL EDITORIAL CONVENTIONS
Thomson Reuters also determines if the journal follows international editorial conventions, which are intended to optimize retrievability of source articles. These conventions include informative journal titles, fully descriptive article titles and author abstracts, complete bibliographic information for all cited references, and full address information for every author.
FULL TEXT ENGLISH
English is the universal language of science. For this reason Thomson Reuters focuses on journals that publish full text in English, or at very least, bibliographic information in English. There are many journals covered in Web of Science that publish articles with bibliographic information in English and full text in another language. However, going forward, it is clear that the journals most important to the international research community will publish full text in English. This is especially true in the natural sciences. There are notable exceptions to this rule in the Arts & Humanities and in Social Sciences topics. This is discussed further below. Nonetheless, full text English is highly desirable, especially if the journal intends to serve an international community of researchers. In addition, all journals must have cited references in the Roman alphabet.
Application of the peer-review process is another indication of journal standards and signifies overall quality of the research presented and the completeness of cited references.6 Inclusion of Funding Acknowledgements is also strongly recommended. Not only do they help create a greater context for the journal, these acknowledgements also function as a confirmation of the importance of the research presented.
Once again, for a more detailed description visit http://wokinfo.com/essays/journal-selection-process/
Edit: In response to Benoit Klockner's comment, I thought I would also quote the 'Citation Analysis' that is used by Thomson Reuters.
The Thomson Reuters Journal Selection Process is unique in that Thomson Reuters editors have a wealth of citation data available to them. Because Web of Science is a true citation index, all cited references from every item in every journal covered in Web of Science are indexed whether or not the cited work is also covered as a source publication. Through these data it is possible to measure the citation impact of journals under evaluation.
The importance of interpreting and understanding these data correctly cannot be emphasized too strongly. Using quantitative citation data to measure impact is meaningful only in the context of journals in the same general discipline. For example, smaller fields like Agricultural Engineering do not generate as many articles or citations as larger fields like Biotechnology or Genetics. In some areas, particularly in the Arts & Humanities, it may take a relatively long time for an article to attract a meaningful number of citations. But in other areas, such as the life sciences, it is not unusual for citations to accrue rapidly and peak after two or three years.7 These facts must be taken into consideration if citation data are to be used correctly.
Citation analysis takes place on at least two levels. Thomson Reuters looks for citations to the journal itself, as expressed by impact factor and/ or total citations received. For new journals that do not yet have a citation history at the publication level, analysts examine the citation record of the contributing authors and editorial board members. This allows them to see if the journal is able to attract contributions from scholars whose prior work has been useful to the research community.
Likewise, established journals that are not covered are often re-evaluated. These journals can experience new growth in citation impact resulting from changes such as translation into English, change in editorial focus, change in publisher, publishing medium, etc.
As noted above, because Thomson Reuters captures all cited references from each of the over 12,000 journals covered, citation information is available on journals not covered as well as those that are covered.
Self-citation rates are also taken into consideration. The self-cited rate relates a journal's self-citations to the number of times it is cited by all journals, including itself. For example, journal X was cited 15,000 times by all journals, including the 2,000 times it cited itself. Its self-cited rate is 2/15 or 13%.
It is entirely normal for authors to reference the prior work that is most relevant to their current results, regardless of the source journal in which the work was published. However, there are journals in which the observed rate of self-citation is a dominant influence in the total level of citation. For these journals, self-citation has the potential to distort the true role of the title in the context of the literature of its subject.8
Among all journals listed in the 2010 JCR Science Edition, for example, 85% have self-citation rates of less than 15%. This shows that self-citation is quite normal for most journals. Significant deviation from this normal rate, however, prompts an examination by Thomson Reuters to determine if excessive self-citations are being used to artificially inflate the impact factor. If we determine that self-citations are being used improperly, the journal's impact factor will be suppressed for at least two years and the journal may be considered for deselection from Web of Science.