RESEARCH METRICS

INTRODUCTION
NECESSITY OF RESEARCH METRICS
TYPES OF RESEARCH METRICS
JOURNAL METRIC
ARTICLE METRIC
DR. PARDEEP SINGH AUTHOR LEVEL METRIC
PROFESSOR , SCHOOL OF COMPUTER SCIENCE & ENGINEERING
GALGOTIAS UNIVERSITY, GREATER NOIDA
+91-9896305860
[email protected]
INTRODUCTION
Research metrics are quantitative tools used to assess the quality and impact
of research outputs. Metrics are a way to see the impact of a scholar's work.
Tools to track and measure impact are designed to help researchers
understand how their work is being used and how it fits into the scholarship
of their field overall. For example:
• How widely is my work being cited in other scholarly publications?
• How widely is my work being read?
• Who is reading and using my work? What subsequent scholarship have I
influenced?
• How do articles that I published in a particular journal compare to articles
that I published elsewhere?
• What influence does my work have outside of academia?
NECESSITY OF RESEARCH METRICS
Evaluating Research Impact: Metrics help assess the influence and
reach of research findings.

Decision Making: Metrics inform decisions about funding,

promotions, and tenure.

Benchmarking: Metrics allow researchers to compare their work to

others in the field.

Identifying Trends: Metrics can help identify emerging trends in

research.

Understanding the impact of one's research can help scholars build

tenure and promotion cases, select publication outlets for future work,
and identify potential collaborators.
TYPE OF RESEARCH METRICS

They can be categorized into two main types:

1. Bibliometrics:
•Citations: Measures the number of times a research article has been cited by
other researchers.
•Impact Factor: A measure of the average number of citations received by articles
published in a particular journal.
•h-index: A metric that reflects both the productivity and impact of a researcher's
publications.
•g-index: A refinement of the h-index, which places more weight on highly cited
publications.
•Eigenfactor score: Measures the influence of a journal within a specific field,
taking into account the quality of citing journals.
TYPE OF RESEARCH METRICS

They can be categorized into two main types:

2. Altmetrics (Alternative Metrics):
•Social Media Mentions: Tracks the number of times a research article is
mentioned on social media platforms.
•News Mentions: Measures the number of times a research article is mentioned in
news outlets.
•Policy Documents: Tracks the number of times a research article is cited in policy
documents.
•Public Engagement: Measures the number of times a research article is
downloaded, viewed, or commented on.
•Plum Analytics: A comprehensive platform that tracks a variety of altmetrics,
including social media mentions, news coverage, policy documents, and public
engagement.
TYPE OF RESEARCH METRICS
• Additional Metrics:
1. Journal Level Metrics
• Impact Factor (IF)
• CiteScore
• Source Normalized Impact per Paper (SNIP)
• SCImago Journal Rank (SJR)
• Eignfactor
2. Author Level Metrics
• h-index
• g-index
• i10-index
• m-index
3. Article Level Metrics
• Citations
• Altmetrics
• Downloads
• Views
JOURNAL LEVEL METRICS
Impact Factor: Journal Impact Factor, or JIF, is the most
commonly-used metric used to indicate journal quality. As the
formula below shows, JIF is a 2-year calculation based on the total
number of citations to articles published in the journal. A 5-year
citation window is an alternative 5-year JIF.
Impact Factor (IF) = Citations in Year X / Citable Items in Years X-1
and X-2

Example:
To calculate the 2023 Impact Factor for a journal, you would:
1.Count the number of times articles published in 2021 and 2022
were cited by other articles in 2023.
2.Count the total number of citable items published in 2021 and
2022.
3.Divide the number of citations by the number of citable items.
JOURNAL LEVEL
CiteScore: CiteScore METRICS
is a metric used to measure the average
number of citations received by articles published in a specific
journal over a four-year period. It is calculated by dividing the
total number of citations received by articles published in the
journal during the past four years by the total number of
documents published in the same period.
Here's the formula:
CiteScore = (Total Citations in Years X-3 to X) / (Total
Documents Published in Years X-3 to X)
For example, to calculate the CiteScore for 2023, you would:
1. Count the total number of citations received in 2020, 2021,
2022, and 2023 for articles published in those years.
2. Count the total number of documents (articles, reviews,
conference papers, etc.) published in 2020, 2021, 2022, and
2023.
3. Divide the total number of citations by the total number of
JOURNAL LEVEL METRICS
Difference between Impact Factor and CiteScore:

Feature Impact Factor CiteScore

Data Source Web of Science Scopus
Time Period 2 years 4 years

All document types

Document Types (articles, reviews,
Included Articles and Reviews conference papers, etc.)
Calculation Citations / Citable Items Citations / Documents
Transparency Less transparent More transparent
 JOURNAL
Source LEVEL
Normalized Impact METRICS
per Paper(SNIP): A metric used to measure the contextual
citation impact of a journal. Unlike the traditional Impact Factor, which doesn't
account for differences in citation practices between fields, SNIP adjusts for these
differences, making it more suitable for comparing journals across different subject
areas.
How SNIP works:
1.Citation Impact per Publication (CIP): This is calculated by dividing the total
number of citations received by a journal's articles in a specific year by the total
number of articles published in that year.
2.Database Citation Potential (DCP): This measures the average citation
potential of all articles in the same subject field as the journal. It takes into account
the overall citation behavior in that field.
3.SNIP Calculation: SNIP is calculated by dividing the CIP by the DCP.
Key advantages of SNIP:
•Contextualized Impact: SNIP considers the specific citation practices of a
journal's subject field, providing a more accurate measure of its impact.
•Fairer Comparisons: It allows for more equitable comparisons between journals
in different fields, as it accounts for differences in citation rates.
 JOURNAL LEVEL METRICS
SCImago Journal Rank (SJR) is a metric used to measure the scientific impact of scholarly journals. It
considers both the number of citations received by a journal's articles and the prestige of the citing
journals.
How SJR Works:
1.Citation Impact: The SJR algorithm calculates the citation impact of each article in a journal. This is
done by considering the number of times the article has been cited and the prestige of the citing
journals.
2.Journal Prestige: The prestige of a journal is determined by the average citation impact of its
articles. Journals with a higher average citation impact are considered to be more prestigious.
3.SJR Calculation: The SJR of a journal is calculated by averaging the citation impact of all articles
published in the journal over a specific period (typically three years).
Key Advantages of SJR:
•Contextualized Impact: SJR considers both the quantity and quality of citations, providing a more
nuanced measure of journal impact.
•Fairer Comparisons: It allows for fairer comparisons between journals in different fields, as it
accounts for differences in citation practices.
•Improved Decision-Making: Researchers and institutions can use SJR to make informed decisions
about journal selection, research evaluation, and overall scholarly impact.
Journal Level Metrics
JOURNAL LEVEL METRICS
Eigenfactor: The Eigenfactor is a metric used to assess the importance of a scientific
journal within the scholarly community. It considers the number and quality of citations
received by articles published in the journal, giving more weight to citations from highly
ranked journals.
Example: Let's consider two journals:
Journal A: A well-established, highly-cited journal in a specific field.
Journal B: A newer journal with fewer publications but gaining recognition.

Scenario:
Journal A publishes an article that is cited by many articles in other journals, including
some from highly-regarded journals.
Journal B publishes an article that is cited by a few articles, but these citations come
from highly-regarded journals in the field.

Eigenfactor Calculation: The Eigenfactor algorithm would assign more weight to the
citations received by the article from Journal B, even though it has fewer citations
overall. This is because the citations are coming from influential sources, which
indicates a higher impact.
Interpretation: In this scenario, even though Journal A may have a higher overall
number of citations, Journal B might have a higher Eigenfactor score due to the quality
of the citations it receives. This suggests that Journal B is becoming increasingly
JOURNAL LEVEL METRICS
Article Influence Score: The Article Influence Score is a metric that measures
the relative importance of a journal on a per-article basis. It's calculated by
dividing the journal's Eigenfactor Score by the fraction of articles published by that
journal.
Key points about Article Influence Score:
1. Relative Importance: It measures the average influence of an article published
in a particular journal.
2. Comparison to Average: A score greater than 1.00 indicates that each article in
the journal has above-average influence, while a score less than 1.00 indicates
below-average influence.
3. Based on Eigenfactor Score: The calculation relies on the Eigenfactor Score,
which measures the overall impact of a journal within the scientific community.
4. Normalization: The fraction of articles published by the journal is normalized to
ensure fair comparisons across different journals.

How to Interpret Article Influence Score:

High Score: A high score suggests that the journal is highly influential and that
each article published in the journal tends to have a significant impact on the
field.
 AUTHOR LEVEL METRICS
H-Index: The h-index is a widely used metric that measures both the productivity and impact of a
researcher's published work. It's a single number that represents the number of papers (h) that the
author has published, each of which has been cited at least h times.
Here's a breakdown of how it works:
1.Sort Publications: List all of the author's publications in descending order of citations.
2.Identify the H-Point: Find the point where the number of publications (n) is equal to the number of
citations (c) for that publication.
3.Determine the H-Index: The h-index is the value of n (or c) at that point.
Example:
 AUTHOR LEVEL METRICS
I10-index: The i10-index is a simple metric introduced by Google Scholar to assess a researcher's
impact. It represents the number of publications the researcher has written that have received at
least 10 citations.
For example, if a researcher has 25 publications, and 15 of them have been cited at least 10
times, their i10-index is 15.

Key points about the i10-index:

1. Focus on highly cited papers: It emphasizes the number of papers with substantial impact,
rather than the total number of citations.
2. Easy to understand: It's a straightforward metric that can be quickly calculated.
3. Complements the h-index: While the h-index considers the number of papers with a
certain citation threshold, the i10-index focuses specifically on papers with at least 10
citations.
 AUTHOR LEVEL METRICS
g-Index: The g-index is a metric used to assess the productivity and impact of a researcher's work. It is
similar to the h-index but gives more weight to highly-cited papers.

Calculation Process:
1. Rank Publications: List all of the researcher's publications in descending order of citations
received.
2. Identify the g-index: The g-index is the largest number g such that the top g publications have at
least g² citations combined.

Example:
• The top 1 publication has 100 citations, which is greater than 1².
• The top 2 publications have 50 citations, which is greater than 2².
• The top 3 publications have 30 citations, which is greater than 3².
• The top 4 publications have 20 citations, which is greater than 4².
• The top 5 publications have 15 citations, which is less than 5².
• Therefore, the g-index for this researcher is 4.
 AUTHOR LEVEL METRICS
m-index: The m-index is a metric used to assess the productivity and impact of a researcher's
work, taking into account their career length. It is calculated by dividing the h-index by the
number of years since the researcher's first publication.
Formula:
m-index = h-index / n
where:
h-index: The researcher's h-index
n: The number of years since the researcher's first publication

Example: Let's say a researcher has an h-index of 20 and has been publishing papers for 10
years. Their m-index would be:
m-index = 20 / 10 = 2

The m-index provides a normalized measure of a researcher's productivity and impact, allowing
for comparisons between researchers at different stages of their careers. A higher m-index
indicates a more productive and impactful researcher, even if they have been publishing for a
shorter period.
 AUTHOR LEVEL METRICS
Advantages of the m-index:
•Accounts for career length: It considers the time a researcher has been active, making it fairer
for younger researchers.
•Allows for comparison: It enables comparisons between researchers at different career stages.
•Identifies early-stage talent: It can help identify young researchers with high potential.

Limitations of the m-index:

•Assumes continuous research activity: It may not be accurate for researchers who have taken
career breaks or had periods of reduced productivity.
•Doesn't account for field-specific differences: It may not be appropriate for all fields, as
publication rates and citation practices vary.

While the m-index is a useful tool for evaluating researcher performance, it should be used in
conjunction with other metrics and qualitative assessments to get a comprehensive picture of a
researcher's impact.
 ARTICLE LEVEL METRICS
1. Citations:
•Traditional Citation Count: The number of times an article has been cited in other scholarly works.
•Field-Weighted Citation Impact (FWCI): This metric accounts for the impact of the journals where citations
originate, giving more weight to citations from high-impact journals.

2. Downloads and Views:

•PDF Downloads: The number of times the article's PDF has been downloaded.
•HTML Views: The number of times the article's HTML version has been viewed.
•Abstract Views: The number of times the article's abstract has been viewed.

3. Social Media Mentions:

•Twitter Mentions: The number of times the article or its authors have been mentioned on Twitter.
•Facebook Shares: The number of times the article has been shared on Facebook.
•Reddit Discussions: The number of discussions on Reddit that reference the article.
•Bookmarks and Saves:
•ResearchGate Saves: The number of times the article has been saved on ResearchGate.
•Mendeley Bookmarks: The number of times the article has been bookmarked on Mendeley.

4. Altmetric Score:
•A composite score that combines various ALMs, providing a single metric to assess an article's overall impact.
THANK YOU