H-Index vs G-Index: Which Academic Metric Should You Use?

The H-index has been the dominant bibliometric measure since Jorge Hirsch introduced it in 2005. But it has a significant blind spot: it treats a paper cited 20 times identically to one cited 20,000 times, as long as both are above the H-threshold. Leo Egghe's G-index, introduced in 2006, was specifically designed to fix this.

How the H-Index Works

The H-index is the largest number h such that you have at least h papers with at least h citations each. If your top 12 papers each have ≥12 citations, your H-index is 12. The 13th paper (if it has only 11 citations) doesn't count, regardless of how many other papers you have.

Example: Citation counts of 45, 30, 22, 18, 12, 9, 5, 3, 1 → H-index = 6 (6 papers with ≥6 citations).

How the G-Index Works

The G-index is the largest number g such that your top g papers together have received at least g² citations in total. It uses cumulative citation sums, so an exceptionally high-cited paper carries much more weight.

Example: Citation counts of 500, 45, 30, 22, 18, 12, 9, 5, 3, 1 — Cumulative sum after 6 papers = 626. Since 626 ≥ 6² = 36, and 626 ≥ 7² = 49, and 626 ≥ 8² = 64... the G-index keeps growing until the cumulative sum falls below g². Here G-index = ~25, versus H-index = 6. The 500-citation paper is finally being credited.

The Key Difference: One Breakout Paper

The defining difference is how each metric handles outlier papers. Consider two researchers:

Alice has a higher H-index despite having significantly less total impact than Bob. Bob's landmark 1,500-citation paper is almost invisible in his H-index. The G-index corrects this — Bob's G-index of 38 vs Alice's 10 correctly captures who has had more absolute scholarly impact.

When to Use H-Index

The H-index is better for measuring sustained, consistent research output. It rewards researchers who publish regularly at a quality level that attracts citations. It is particularly appropriate when:

• Benchmarking productivity across a career span
• Comparing researchers in the same field at similar career stages
• Evaluating promotion and tenure (where consistency is valued)
• Filtering large applicant pools for funding (as a crude screen)

When to Use G-Index

The G-index is better for measuring peak impact and breakthrough contributions. It is particularly appropriate when:

• Identifying researchers with transformative single contributions
• Evaluating fields where one landmark paper defines a career (physics, computer science algorithms)
• Assessing early-career researchers who have published one widely-cited paper
• Recognising that citation accumulation is skewed and the top paper deserves proper credit

What the G–H Gap Tells You

The difference between your G-index and H-index (G – H) is itself informative. A large gap indicates that your citation distribution is highly unequal — a small number of papers are pulling in a disproportionate share of citations.

• G – H ≤ 3: Citations are evenly distributed. You are a consistent producer.
• G – H = 4–10: Some papers have pulled ahead. You have one or two high-impact contributions.
• G – H ≥ 11: You have at least one breakthrough paper with exceptional citation counts. G-index gives a much fairer picture of your impact.

Which Metric Do Universities and Funding Bodies Actually Use?

In practice, H-index dominates institutional evaluation. It is used by most universities for tenure and promotion, by funding agencies as a screening criterion, and in national research assessment exercises. This is partly inertia — the H-index arrived first and became embedded in evaluation systems.

The G-index sees more use in bibliometric research, meta-analyses of scientific impact, and among researchers who want a more nuanced self-assessment. Some institutions in continental Europe have begun incorporating G-index into assessment frameworks, but it is not yet standard.

Practical advice: Know your H-index for institutional purposes. Know your G-index to understand your own impact more accurately, especially if you have one or two high-citation papers driving much of your recognition.

i10-Index: The Simpler Alternative

A third metric worth knowing is the i10-index, used by Google Scholar: it simply counts papers with at least 10 citations. It is the most transparent of the three — no threshold optimisation, no cumulative sums — but also the least nuanced. It tells you how many papers have crossed a basic citation threshold, which is useful for early-career researchers but becomes less discriminating as careers develop.

Summary Comparison