A PhD student in computer science may publish 8 conference papers and zero journal articles before graduating — and be considered highly productive. A PhD student in chemistry who publishes 8 conference papers and zero journal articles may struggle to find a postdoc position. These are not contradictions. They reflect how fundamentally different disciplines think about publication formats.
What Are Scopus and SCI?
Scopus (Elsevier) is the largest abstract and citation database, covering over 27,000 peer-reviewed journals, books, and — importantly — conference proceedings. A Scopus-indexed conference paper is fully searchable, citable, and appears in author citation metrics.
SCI (Science Citation Index, part of Web of Science / Clarivate) covers roughly 9,200 high-impact journals across science, technology, and medicine. SCIE (the expanded version) covers ~9,500 journals. Conference proceedings are generally NOT indexed in SCI unless explicitly noted.
A key difference: SCI indexes journals. Scopus indexes both journals and conference proceedings. So when someone says "SCI paper," they mean a journal article. When they say "Scopus-indexed conference paper," they mean proceedings — which SCI typically would not cover.
Where Conference Papers Reign Supreme: Computer Science
In computer science and related fields, the flagship conference is the primary publication venue. A paper at NeurIPS, CVPR, SIGCOMM, or SOSP is a career milestone that most hiring committees weight higher than a paper in many journals.
Reasons:
- CS conferences have peer review as rigorous as top journals, with acceptance rates of 15–25% at flagship venues
- Turnaround time is faster — 3–4 months from submission to notification vs. 6–18 months for journals
- The community reads conference proceedings, not journals, for current work
- Citation patterns reflect conference-centricity: the most-cited CS work is almost always conference papers
Where Journals Reign Supreme: Natural and Life Sciences
In chemistry, biology, physics, medicine, and environmental science, the journal article is the definitive unit of scientific output. Conference proceedings in these fields are often not peer-reviewed, not archived, and carry little professional weight.
An SCI journal publication in Nature Chemistry, JACS, or Angewandte Chemie signals that your work has survived rigorous external review and is considered a permanent contribution to the literature. A poster at a chemistry conference signals attendance, not publication.
The Middle Ground: Engineering and Social Sciences
Electrical engineering, biomedical engineering, and some social sciences occupy a middle ground:
- IEEE flagship conference papers (IEEE INFOCOM, IEEE ICASSP) are considered primary research contributions — but IEEE journal publications in the same society's transactions are often weighted higher for promotion decisions
- Economics and social sciences have begun to treat top conference papers (NBER, AEA) as significant research outputs — but journal publication in QJE, JPE, or Econometrica is still the definitive measure of quality
What Hiring Committees Actually Look At
Ask yourself: what do papers look like in the CVs of people who recently got the kind of job you want? This is the most reliable signal.
For a CS faculty position at a research university: NeurIPS/ICML/CVPR papers count as much as journal publications. A strong conference record is sufficient.
For a biology or chemistry faculty position: three or more first-author SCI journal papers (ideally in Impact Factor 5+ journals) is the minimum threshold at most research universities. Conference papers do not appear in the "publications" section.
For industry research positions: publication record matters, but recent, relevant work at top venues (in whatever format) tends to outweigh format considerations.
Impact Factor vs. Conference Ranking
- SCI journals are ranked by Impact Factor (IF) — average citations per article over two years. IF 1 is low; IF 10+ is excellent; Nature and Science are above 40.
- CS conferences are ranked by CORE rankings (A*, A, B, C) or by the h-index of the venue as measured on Google Scholar Metrics.
An A* CORE conference and an IF 10 journal are broadly comparable in prestige within their respective fields — but you cannot compare them directly across disciplines.
Practical Advice
Publish where your community publishes. That sounds obvious, but early-career researchers sometimes make the mistake of chasing format (journal for prestige) rather than venue (the conference where their actual peers will read and cite their work). Read the CVs in your field. Talk to your supervisor. Then publish there.