What Is Research Evaluation and How Do You Assess Research Impact?

Research evaluation is the process of judging the quality, influence, and output of scholarly work — and in practice it usually means turning citation data into a defensible story about your impact. On Harzing.com, that work is supported by the Publish or Perish software for citation analysis, the Journal Quality List, and blog guidance such as "Finding the pearls in your citation record," which shows how to present a citation record to its best advantage. This explainer covers what gets evaluated, which metrics are commonly used and where they mislead, how to collect and analyse citation data, and how the results feed into promotion, funding, and career decisions.

What "research evaluation" covers

Evaluation happens at several levels, and the metrics differ at each:

Harzing.com frames research evaluation alongside academic publishing, creating research impact, and research diffusion through social media — a reminder that citation counts are only one channel through which impact is expressed.

Metrics you will meet, and their limits

• Citation count — how often a work is cited. Simple, but strongly age- and field-dependent; a recent paper in a slow-citing field will look weak next to an older paper in a fast-citing one.
• h-index — the point where a researcher has h papers each cited at least h times. It rewards sustained output but penalises early-career researchers and those who publish fewer, longer works.
• Journal-level indicators — used as a proxy for quality. The Journal Quality List on Harzing.com exists precisely because journal standing varies by field and by list, so a single ranking is rarely decisive.
• Altmetrics and diffusion — social media sharing, policy references, and practitioner uptake. Harzing.com treats research diffusion as a distinct activity from citation accumulation.

The practical rule: never present a single number as the verdict. Present the number, the context that explains it, and the specific works that carry the impact.

Collecting and analysing citation data with Publish or Perish

Publish or Perish is the citation-analysis tool hosted on Harzing.com. The general workflow:

1. Define the input. Decide whose or which works you are evaluating — an author name, a set of publications, or a journal.
2. Run the query against the citation data source the tool queries.
3. Inspect the results, not just the summary. Check for name disambiguation problems: common names, name changes, and institutional variants all split or inflate a record.
4. Clean the record. Remove duplicates and works that are not yours before trusting any aggregate metric.
5. Read the metrics in context. Export or note the citation counts, h-index, and the distribution of citations across works.
6. Verify the expected result. The cleaned record should match your own publication list; if it does not, the discrepancy is the finding.

The most common failure point is step 3. A high h-index built on a merged record of two different researchers is worse than a modest, accurate one — it will not survive scrutiny in a promotion or funding review.

Turning a citation record into a case

Harzing.com's "Finding the pearls in your citation record" describes the task directly: find the pearls, polish them, and string them into a necklace that tells your citation story effectively. In practice that means:

• Identify your highest-impact works and explain why they were cited — a methods paper, a widely reused dataset, a framing concept.
• Group citations thematically rather than listing them chronologically, so a reviewer sees a coherent contribution.
• Name the outliers. A single highly cited paper among otherwise modest output is a story about a specific idea landing, not a statistical accident to hide.
• Pair citation evidence with non-citation evidence — invited talks, adoption in practice, teaching use — because evaluation committees increasingly ask for impact beyond the count.

Where evaluation results are used

Citation and quality evidence typically feeds into:

• Promotion and tenure applications — where a structured, verifiable record matters more than a headline number.
• Funding applications — where track record is one criterion among feasibility and fit.
• Journal and venue selection — where the Journal Quality List helps you judge where a paper is likely to be visible and credible.
• Career strategy — Harzing.com's career guides cover journal publishing, research diffusion, research impact, and promotion applications as connected activities, not separate chores.

If you are early in your career, the site points to its Working in academia page as the best starting point, since evaluation expectations differ sharply by career stage and country.

A workable default

Treat research evaluation as a documentation habit rather than a once-a-decade panic. Keep a clean, disambiguated publication record; run it through Publish or Perish periodically so you notice trends early; check journal standing against the Journal Quality List before submitting; and rehearse the two-sentence explanation of your most-cited work. That preparation is what makes an evaluation conversation about your research rather than about your spreadsheet.