Scientometrics. 2015; 103(1): 213–228.

Published online 2015 Jan 22. doi: 10.1007/s11192-014-1524-z

PMCID: PMC4365275

# Power laws in citation distributions: evidence from Scopus

This article has been cited by other articles in PMC.

## Abstract

Modeling

distributions of citations to scientific papers is crucial for

understanding how science develops. However, there is a considerable

empirical controversy on which statistical model fits the citation

distributions best. This paper is concerned with rigorous empirical

detection of power-law behaviour in the distribution of citations

received by the most highly cited scientific papers. We have used a

large, novel data set on citations to scientific papers published

between 1998 and 2002 drawn from Scopus. The power-law model is compared

with a number of alternative models using a likelihood ratio test. We

have found that the power-law hypothesis is rejected for around half of

the Scopus fields of science. For these fields of science, the Yule,

power-law with exponential cut-off and log-normal distributions seem to

fit the data better than the pure power-law model. On the other hand,

when the power-law hypothesis is not rejected, it is usually empirically

indistinguishable from most of the alternative models. The pure

power-law model seems to be the best model only for the most highly

cited papers in “Physics and Astronomy”. Overall, our results seem to

support theories implying that the most highly cited scientific papers

follow the Yule, power-law with exponential cut-off or log-normal

distribution. Our findings suggest also that power laws in citation

distributions, when present, account only for a very small fraction of

the published papers (less than 1 % for most of science fields) and that

the power-law scaling parameter (exponent) is substantially higher

(from around 3.2 to around 4.7) than found in the older literature.

distributions of citations to scientific papers is crucial for

understanding how science develops. However, there is a considerable

empirical controversy on which statistical model fits the citation

distributions best. This paper is concerned with rigorous empirical

detection of power-law behaviour in the distribution of citations

received by the most highly cited scientific papers. We have used a

large, novel data set on citations to scientific papers published

between 1998 and 2002 drawn from Scopus. The power-law model is compared

with a number of alternative models using a likelihood ratio test. We

have found that the power-law hypothesis is rejected for around half of

the Scopus fields of science. For these fields of science, the Yule,

power-law with exponential cut-off and log-normal distributions seem to

fit the data better than the pure power-law model. On the other hand,

when the power-law hypothesis is not rejected, it is usually empirically

indistinguishable from most of the alternative models. The pure

power-law model seems to be the best model only for the most highly

cited papers in “Physics and Astronomy”. Overall, our results seem to

support theories implying that the most highly cited scientific papers

follow the Yule, power-law with exponential cut-off or log-normal

distribution. Our findings suggest also that power laws in citation

distributions, when present, account only for a very small fraction of

the published papers (less than 1 % for most of science fields) and that

the power-law scaling parameter (exponent) is substantially higher

(from around 3.2 to around 4.7) than found in the older literature.

**Keywords:**Citation distribution, Power law, Statistical modelling, Scopus

## Introduction

It

is often argued in scientometrics, social physics and other sciences

that distributions of some scientific “items” (e.g., articles,

citations) produced by some scientific sources (e.g., authors, journals)

have heavy tails that can be modelled using a power-law model. These

distributions are then said to conform to the Lotka’s law (Lotka 1926).

Examples of such distributions include author productivity, occurrence

of words, citations received by papers, nodes of social networks, number

of authors per paper, scattering of scientific literature in journals,

and many others (Egghe 2005).

In fact, power-law models are widely used in many sciences as physics,

biology, earth and planetary sciences, economics, finance, computer

science, and others (Newman 2005; Clauset et al. 2009). Models equivalent to Lotka’s law are known as Pareto’s law in economics (Gabaix 2009) and as Zipf’s law in linguistics (Baayen 2001).

Appropriate measuring and providing scientific explanations for power

laws plays an important role in understanding the behaviour of various

natural and social phenomena.

is often argued in scientometrics, social physics and other sciences

that distributions of some scientific “items” (e.g., articles,

citations) produced by some scientific sources (e.g., authors, journals)

have heavy tails that can be modelled using a power-law model. These

distributions are then said to conform to the Lotka’s law (Lotka 1926).

Examples of such distributions include author productivity, occurrence

of words, citations received by papers, nodes of social networks, number

of authors per paper, scattering of scientific literature in journals,

and many others (Egghe 2005).

In fact, power-law models are widely used in many sciences as physics,

biology, earth and planetary sciences, economics, finance, computer

science, and others (Newman 2005; Clauset et al. 2009). Models equivalent to Lotka’s law are known as Pareto’s law in economics (Gabaix 2009) and as Zipf’s law in linguistics (Baayen 2001).

Appropriate measuring and providing scientific explanations for power

laws plays an important role in understanding the behaviour of various

natural and social phenomena.

This paper is concerned

with empirical detection of power-law behaviour in the distribution of

citations received by scientific papers. The power-law distribution of

citations for the highly cited papers was first suggested by SollaPrice (1965), who also proposed a “cumulative advantage” mechanism that could generate the power-law distribution (SollaPrice 1976).

More recently, a growing literature has developed that aims at

measuring power laws in the right tails of citation distributions. In

particular, Redner (1998), Redner (2005)

found that the right tails of citation distributions for articles

published in Physical Review over a century and of articles published in

1981 in journals covered by Thomson Scientific’s Web of Science (WoS)

follow power laws. The latter data set was also modelled with power-law

techniques by Clauset et al. (2009) and Peterson et al. (2010).

The latter study also used data from 2007 list of the living highest

h-index chemists and from Physical Review D between 1975 and 1994.

VanRaan (2006)

observed that the top of the distribution of around 18,000 papers

published between 1991 and 1998 in the field of chemistry in Netherlands

follows a power law distribution. Power-law models were also fitted to

data from high energy physics (Lehmann et al. 2003), data for most cited physicists (Laherrère and Sornette 1998), data for all papers published in journals of the American Physical Society from 1983 to 2008 (Eom and Fortunato 2011), and to data for all physics papers published between 1980 and 1989 (Golosovsky and Solomon 2012).

with empirical detection of power-law behaviour in the distribution of

citations received by scientific papers. The power-law distribution of

citations for the highly cited papers was first suggested by SollaPrice (1965), who also proposed a “cumulative advantage” mechanism that could generate the power-law distribution (SollaPrice 1976).

More recently, a growing literature has developed that aims at

measuring power laws in the right tails of citation distributions. In

particular, Redner (1998), Redner (2005)

found that the right tails of citation distributions for articles

published in Physical Review over a century and of articles published in

1981 in journals covered by Thomson Scientific’s Web of Science (WoS)

follow power laws. The latter data set was also modelled with power-law

techniques by Clauset et al. (2009) and Peterson et al. (2010).

The latter study also used data from 2007 list of the living highest

h-index chemists and from Physical Review D between 1975 and 1994.

VanRaan (2006)

observed that the top of the distribution of around 18,000 papers

published between 1991 and 1998 in the field of chemistry in Netherlands

follows a power law distribution. Power-law models were also fitted to

data from high energy physics (Lehmann et al. 2003), data for most cited physicists (Laherrère and Sornette 1998), data for all papers published in journals of the American Physical Society from 1983 to 2008 (Eom and Fortunato 2011), and to data for all physics papers published between 1980 and 1989 (Golosovsky and Solomon 2012).

Recently, Albarrán and Ruiz-Castillo (2011)

tested for the power-law behavior using a large WoS dataset of 3.9

million articles published between 1998 and 2002 categorized in 22 WoS

research fields. The same dataset was also used to search for the power

laws in the right tail of citation distributions categorized in 219 WoS

scientific sub-fields (Albarrán et al. 2011a, b).

These studies offer the largest existing body of evidence on the

power-law behaviour of citation distributions. Three major conclusions

appear from them. First, the power-law behavior is not universal. The

existence of power law cannot be rejected in the WoS data for 17 out of

22 and for 140 out of 219 sub-fields studied in Albarrán and

Ruiz-Castillo (2011) and in Albarrán et al. (2011a, b),

respectively. Secondly, in opposition to previous studies, these papers

found that the scaling parameter (exponent) of the power-law

distribution is above 3.5 in most of the cases, while the older

literature suggested that the parameter value is between 2 and 3

(Albarrán et al. 2011).

Third, power laws in citation distributions are rather small—on average

they cover just about 2 % of the most highly cited articles in a given

WoS field of science and account for about 13.5 % of all citations in

the field.

tested for the power-law behavior using a large WoS dataset of 3.9

million articles published between 1998 and 2002 categorized in 22 WoS

research fields. The same dataset was also used to search for the power

laws in the right tail of citation distributions categorized in 219 WoS

scientific sub-fields (Albarrán et al. 2011a, b).

These studies offer the largest existing body of evidence on the

power-law behaviour of citation distributions. Three major conclusions

appear from them. First, the power-law behavior is not universal. The

existence of power law cannot be rejected in the WoS data for 17 out of

22 and for 140 out of 219 sub-fields studied in Albarrán and

Ruiz-Castillo (2011) and in Albarrán et al. (2011a, b),

respectively. Secondly, in opposition to previous studies, these papers

found that the scaling parameter (exponent) of the power-law

distribution is above 3.5 in most of the cases, while the older

literature suggested that the parameter value is between 2 and 3

(Albarrán et al. 2011).

Third, power laws in citation distributions are rather small—on average

they cover just about 2 % of the most highly cited articles in a given

WoS field of science and account for about 13.5 % of all citations in

the field.

The main aim of this paper is to use a

statistically rigorous approach to answer the empirical question of

whether the power-law model describes best the observed distribution of

highly cited papers. We use the statistical toolbox for detecting

power-law behaviour introduced by Clauset et al. (2009).

There are two major contributions of the present paper. First, we use a

very large, previously unused data set on the citation distributions of

the most highly cited papers in several fields of science. This data

set comes from Scopus, a bibliographic database introduced in 2004 by

Elsevier, and contains 2.2 million articles published between 1998 and

2002 and categorized in 27 Scopus major subject areas of science. Most

of the previous studies used rather small data sets, which were not

suitable for rigorous statistical detecting of the power-law behaviour.

In contrast, our sample is even bigger with respect to the most highly

cited papers than the large sample used in the recent contributions

based on WoS data (Albarrán and Ruiz-Castillo 2011; Albarrán et al. 2011a, b). This results from the fact that Scopus indexes about 70 % more sources compared to the WoS (López-Illescas et al. 2008; Chadegani et al. 2013) and therefore gives a more comprehensive coverage of citation distributions.

statistically rigorous approach to answer the empirical question of

whether the power-law model describes best the observed distribution of

highly cited papers. We use the statistical toolbox for detecting

power-law behaviour introduced by Clauset et al. (2009).

There are two major contributions of the present paper. First, we use a

very large, previously unused data set on the citation distributions of

the most highly cited papers in several fields of science. This data

set comes from Scopus, a bibliographic database introduced in 2004 by

Elsevier, and contains 2.2 million articles published between 1998 and

2002 and categorized in 27 Scopus major subject areas of science. Most

of the previous studies used rather small data sets, which were not

suitable for rigorous statistical detecting of the power-law behaviour.

In contrast, our sample is even bigger with respect to the most highly

cited papers than the large sample used in the recent contributions

based on WoS data (Albarrán and Ruiz-Castillo 2011; Albarrán et al. 2011a, b). This results from the fact that Scopus indexes about 70 % more sources compared to the WoS (López-Illescas et al. 2008; Chadegani et al. 2013) and therefore gives a more comprehensive coverage of citation distributions.

^{1}The

second major contribution of the paper is to provide a rigorous

statistical comparison of the power-law model and a number of

alternative models with respect to the problem which theoretical

distribution fits better empirical data on citations. This problem of

model selection has been previously studied in some contributions to the

literature. It has been argued that models like stretched exponential

(Laherrère and Sornette 1998), Yule (SollaPrice 1976), log-normal (Redner 2005; Stringer et al. 2008; Radicchi et al. 2008), Tsallis (Tsallis and deAlbuquerque 2000; Anastasiadis et al. 2010; Wallace et al. 2009) or shifted power law (Eom and Fortunato 2011)

fit citation distributions equally well or better than the pure

power-law model. However, previous papers have either focused on a

single alternative distribution or used only visual methods to choose

between the competing models. The present paper fills the gap by

providing a systematic and statistically rigorous comparison of the

power-law distribution with such alternative models as the log-normal,

exponential, stretched exponential (Weibull), Tsallis, Yule and

power-law with exponential cut-off. The comparison between models was

performed using a likelihood ratio test (Vuong 1989; Clauset et al. 2009).

second major contribution of the paper is to provide a rigorous

statistical comparison of the power-law model and a number of

alternative models with respect to the problem which theoretical

distribution fits better empirical data on citations. This problem of

model selection has been previously studied in some contributions to the

literature. It has been argued that models like stretched exponential

(Laherrère and Sornette 1998), Yule (SollaPrice 1976), log-normal (Redner 2005; Stringer et al. 2008; Radicchi et al. 2008), Tsallis (Tsallis and deAlbuquerque 2000; Anastasiadis et al. 2010; Wallace et al. 2009) or shifted power law (Eom and Fortunato 2011)

fit citation distributions equally well or better than the pure

power-law model. However, previous papers have either focused on a

single alternative distribution or used only visual methods to choose

between the competing models. The present paper fills the gap by

providing a systematic and statistically rigorous comparison of the

power-law distribution with such alternative models as the log-normal,

exponential, stretched exponential (Weibull), Tsallis, Yule and

power-law with exponential cut-off. The comparison between models was

performed using a likelihood ratio test (Vuong 1989; Clauset et al. 2009).

## Materials and methods

### Fitting power-law model to citation data

We follow Clauset et al. (2009)

in choosing methods for fitting power laws to citation distributions.

These authors carefully show that, in general, the appropriate methods

depend on whether the data are continuous or discrete. In our case, the

latter is true as citations are non-negative integers. Let

be the number of citations received by an article in a given field of

science. The probability density function (pdf) of the discrete

power-law model is defined as

p(x)=x−αζ(α,x0), where

is a shape parameter of the power-law distribution, known as the

power-law exponent or scaling parameter. The power-law behaviour is

usually found only for values greater than some minimum, denoted by

In case of citation distributions, the power-law behaviour has been

found on average only in the top 2 % of all articles published in a

field of science (Albarrán et al. 2011a, b).

in choosing methods for fitting power laws to citation distributions.

These authors carefully show that, in general, the appropriate methods

depend on whether the data are continuous or discrete. In our case, the

latter is true as citations are non-negative integers. Let

*x*be the number of citations received by an article in a given field of

science. The probability density function (pdf) of the discrete

power-law model is defined as

1

*ζ*(*α*,*x*_{0}) is the generalized or Hurwitz zeta function. The*α*is a shape parameter of the power-law distribution, known as the

power-law exponent or scaling parameter. The power-law behaviour is

usually found only for values greater than some minimum, denoted by

*x*_{0}.In case of citation distributions, the power-law behaviour has been

found on average only in the top 2 % of all articles published in a

field of science (Albarrán et al. 2011a, b).

The lower bound on the power-law behaviour,

should be therefore estimated if we want to measure precisely in which

part of a citation distribution the model applies. Moreover, we need an

estimate of

*x*_{0},should be therefore estimated if we want to measure precisely in which

part of a citation distribution the model applies. Moreover, we need an

estimate of

*x*_{0}if we want to obtain an unbiased estimate of the power-law exponent,*α*.We estimate

L(α)=−nlnζ(α,x0)−α∑ni=1lnxi, where

*α*using the maximum likelihood (ML) estimation. The log-likelihood function corresponding to (1) is2

*x*_{i}is the number of citations received by the*i*th paper (*i*= 1, ⋯ ,*n*).The ML estimate for

*α*is found by numerical maximization of (2).^{2}Following Clauset et al. (2009), we use the following procedure to estimate the lower bound on the power-law behaviour, α^ ,

and then we compute the well-known Kolmogorov–Smirnov (KS) statistic

for the data and the fitted model. The KS statistic is defined as

KS=maxx⩾x0|S(x)−P(x;α^)|, where P(x,α^) is the cdf for the fitted power-law model to observations for which x^0 is then chosen as a value of α^ and x^0 , are computed with standard bootstrap methods with 1,000 replications.

*x*_{0}. For each*x*⩾*x*_{min}, we calculate the ML estimate of the power-law exponent,and then we compute the well-known Kolmogorov–Smirnov (KS) statistic

for the data and the fitted model. The KS statistic is defined as

3

*S*(*x*) is the cumulative distribution function (cdf) for the observations with value at least*x*_{0}, and*x*⩾*x*_{0}. The estimate*x*_{0}for which the KS statistic is the smallest. The standard errors for both estimated parameters,### Goodness-of-fit and model selection tests

The

next step in measuring power laws involves testing goodness of fit. A

positive result of such a test allows to conclude that a power-law model

is consistent with data. Following Clauset et al. (2009) again, we use a test based on a semi-parametric bootstrap approach.

Next, a large number of synthetic data sets is generated that follow

the originally fitted power-law model above the estimatedx^0 .

Then, a power-law model is fitted to each of the generated data sets

using the same methods as for the original data set, and the KS

statistics are calculated. The fraction of data sets for which their own

KS statistic is larger than

value of the test. It represents a probability that the KS statistics

computed for data drawn from the power-law model fitted to the original

data is at least as large as

next step in measuring power laws involves testing goodness of fit. A

positive result of such a test allows to conclude that a power-law model

is consistent with data. Following Clauset et al. (2009) again, we use a test based on a semi-parametric bootstrap approach.

^{3}The procedure starts with fitting a power-law model to data and calculating a KS statistic (see Eq. 3) for this fit, denoted by*k*.Next, a large number of synthetic data sets is generated that follow

the originally fitted power-law model above the estimated

*x*_{0}and have the same non-power-law distribution as the original data set belowThen, a power-law model is fitted to each of the generated data sets

using the same methods as for the original data set, and the KS

statistics are calculated. The fraction of data sets for which their own

KS statistic is larger than

*k*is the*p*value of the test. It represents a probability that the KS statistics

computed for data drawn from the power-law model fitted to the original

data is at least as large as

*k*. The power-law hypothesis is rejected if the*p*value is smaller than some chosen threshold.^{4}Following Clauset et al. (2009), we rule out the power-law model if the estimated*p*value for this test is smaller than 0.1. In the present paper, we use 1,000 generated data sets.If

the goodness-of-fit test rejects the power-law hypothesis, we may

conclude that the power law has not been found. However, if a data set

is fitted well by a power law, the question remains if there is an

alternative distribution, which is an equally good or better fit to this

data set. We need, therefore, to fit some rival distributions and

evaluate which distribution gives a better fit. To this aim, we use the

likelihood ratio test, which tests if the compared models are equally

close to the true model against the alternative that one is closer. The

test computes the logarithm of the ratio of the likelihoods of the data

under two competing distributions, LR, which is negative or positive

depending on which model fits data better. Specifically, let us consider

two distributions with pdfs denoted by

LR=∑ni=1[lnp1(xi)−lnp2(xi)]. A positive value of the LR suggests that model

fits the data better. However, the sign of the LR can be used to

determine which model should be favored only if the LR is significantly

different from zero. Vuong (1989) showed that in the case of non-nested models the normalized log-likelihood ratio NLR =

value is small (for example, smaller than 0.1), then the sign of the LR

can probably be trusted as an indicator of which model is preferred.

However, if the

the goodness-of-fit test rejects the power-law hypothesis, we may

conclude that the power law has not been found. However, if a data set

is fitted well by a power law, the question remains if there is an

alternative distribution, which is an equally good or better fit to this

data set. We need, therefore, to fit some rival distributions and

evaluate which distribution gives a better fit. To this aim, we use the

likelihood ratio test, which tests if the compared models are equally

close to the true model against the alternative that one is closer. The

test computes the logarithm of the ratio of the likelihoods of the data

under two competing distributions, LR, which is negative or positive

depending on which model fits data better. Specifically, let us consider

two distributions with pdfs denoted by

*p*_{1}(*x*) and*p*_{2}(*x*). The LR is defined as:4

*p*_{1}(*x*)fits the data better. However, the sign of the LR can be used to

determine which model should be favored only if the LR is significantly

different from zero. Vuong (1989) showed that in the case of non-nested models the normalized log-likelihood ratio NLR =

*n*^{-1/2}LR/*σ*, where*σ*is the estimated standard deviation of LR, has a limit standard normal distribution.^{5}This result can be used to compute a*p*value for the test discriminating between the competing models. If the*p*value is small (for example, smaller than 0.1), then the sign of the LR

can probably be trusted as an indicator of which model is preferred.

However, if the

*p*value is large, then the test is unable to choose between the compared distributions.We have followed Clauset et al. (2009)

in choosing the following alternative discrete distributions:

exponential, stretched exponential (Weibull), log-normal, Yule and the

power law with exponential cut-off.

Most of these models have been considered in previous literature on

modeling citation distribution. As another alternative, we also use the

Tsallis distribution, which has been also proposed as a model for

citation distributions (Wallace et al. 2009; Anastasiadis et al. 2010).

Finally, we also consider a “digamma” model using exponential functions

of a digamma function, which was recently introduced for distributions

with heavy tails in a statistical physics framework based on the

principle of maximum entropy (Peterson et al. 2013).

in choosing the following alternative discrete distributions:

exponential, stretched exponential (Weibull), log-normal, Yule and the

power law with exponential cut-off.

^{6}Most of these models have been considered in previous literature on

modeling citation distribution. As another alternative, we also use the

Tsallis distribution, which has been also proposed as a model for

citation distributions (Wallace et al. 2009; Anastasiadis et al. 2010).

Finally, we also consider a “digamma” model using exponential functions

of a digamma function, which was recently introduced for distributions

with heavy tails in a statistical physics framework based on the

principle of maximum entropy (Peterson et al. 2013).

^{7}The definitions of our alternative distributions are given in Table 1.

### Data

We

use citation data from Scopus, a bibliographic database introduced in

2004 by Elsevier. Scopus is a major competitor to the most-widely used

data source in the literature on modeling citation distributions—Web of

Science (WoS) from Thomson Reuters. Scopus covers 29 million records

with references going back to 1996 and 21 million pre-1996 records going

back as far as 1823. An important limitation of the database is that it

does not cover cited references for pre-1996 articles. Scopus contains

21,000 peer-reviewed journals from more than 5,000 international

publishers. It covers about 70 % more sources compared to the WoS

(López-Illescas et al. 2008),

but a large part of the additional sources are low-impact journals. A

recent literature review has found that the quite extensive literature

that compares WoS and Scopus from the perspective of citation analysis

offers mixed results (Chadegani et al. 2013).

However, most of the studies suggest that, at least for the period from

1996 on, the number of citations in both databases is either roughly

similar or higher in Scopus than in WoS. Therefore, is seems that Scopus

constitutes a useful alternative to WoS from the perspective of

modeling citation distributions.

use citation data from Scopus, a bibliographic database introduced in

2004 by Elsevier. Scopus is a major competitor to the most-widely used

data source in the literature on modeling citation distributions—Web of

Science (WoS) from Thomson Reuters. Scopus covers 29 million records

with references going back to 1996 and 21 million pre-1996 records going

back as far as 1823. An important limitation of the database is that it

does not cover cited references for pre-1996 articles. Scopus contains

21,000 peer-reviewed journals from more than 5,000 international

publishers. It covers about 70 % more sources compared to the WoS

(López-Illescas et al. 2008),

but a large part of the additional sources are low-impact journals. A

recent literature review has found that the quite extensive literature

that compares WoS and Scopus from the perspective of citation analysis

offers mixed results (Chadegani et al. 2013).

However, most of the studies suggest that, at least for the period from

1996 on, the number of citations in both databases is either roughly

similar or higher in Scopus than in WoS. Therefore, is seems that Scopus

constitutes a useful alternative to WoS from the perspective of

modeling citation distributions.

Journals in Scopus are

classified under four main subject areas: life sciences (4,200

journals), health sciences (6,500 journals), physical sciences (7,100

journals) and social sciences including arts and humanities (7,000

journals). The four main subject areas are further divided into 27 major

subject areas and more than 300 minor subject areas. Journals may be

classified under more than one subject area.

classified under four main subject areas: life sciences (4,200

journals), health sciences (6,500 journals), physical sciences (7,100

journals) and social sciences including arts and humanities (7,000

journals). The four main subject areas are further divided into 27 major

subject areas and more than 300 minor subject areas. Journals may be

classified under more than one subject area.

The analysis in this paper was performed on the level of 27 Scopus major subject areas of science.

From the various document types contained in Scopus, we have selected

only articles. For the purpose of comparability with the recent

WoS-based studies (Albarrán and Ruiz-Castillo 2011; Albarrán et al. 2011a),

only the articles published between 1998 and 2002 were considered.

Following previous literature, we have chosen a common 5-year citation

window for all articles published in 1998–2002.

^{8}From the various document types contained in Scopus, we have selected

only articles. For the purpose of comparability with the recent

WoS-based studies (Albarrán and Ruiz-Castillo 2011; Albarrán et al. 2011a),

only the articles published between 1998 and 2002 were considered.

Following previous literature, we have chosen a common 5-year citation

window for all articles published in 1998–2002.

^{9}See Albarrán and Ruiz-Castillo (2011) for a justification of choosing the 5-year citation window common for all fields of science.In

order to measure the power-law behaviour of citations, we need data on

the right tails of citation distributions. To this end, we have used the

Scopus Citation Tracker to collect citations formin(100,000;x) of the highest cited articles, where

is the actual number of articles published in a given field of science

during 1998–2002. This analysis was performed separately for each of the

27 science fields categorized by Scopus.

order to measure the power-law behaviour of citations, we need data on

the right tails of citation distributions. To this end, we have used the

Scopus Citation Tracker to collect citations for

*x*is the actual number of articles published in a given field of science

during 1998–2002. This analysis was performed separately for each of the

27 science fields categorized by Scopus.

Descriptive statistics for our data sets are presented in Table 2.

In

some cases, there was less than 100,000 articles published in a field

of science during 1998–2002 and we were able to obtain complete or

almost complete distributions of citations (see columns 2–4 of Table 2).

In other cases, we have obtained only a part of the relevant

distribution encompassing the right tail and some part of the middle of

the distribution. The smallest portions of citation distributions were

obtained for Medicine (8.4 % of total papers), Biochemistry, Genetics

and Molecular Biology (15.7 %) and Physics and Astronomy (18.4 %).

However, using the WoS data for 22 science categories, Albarrán and

Ruiz-Castillo (2011)

found that power laws account usually only for less than 2 % of the

highest-cited articles. Therefore, it seems that the coverage of the

right tails of citation distributions in our samples is satisfactory for

our purposes.

some cases, there was less than 100,000 articles published in a field

of science during 1998–2002 and we were able to obtain complete or

almost complete distributions of citations (see columns 2–4 of Table 2).

^{10}In other cases, we have obtained only a part of the relevant

distribution encompassing the right tail and some part of the middle of

the distribution. The smallest portions of citation distributions were

obtained for Medicine (8.4 % of total papers), Biochemistry, Genetics

and Molecular Biology (15.7 %) and Physics and Astronomy (18.4 %).

However, using the WoS data for 22 science categories, Albarrán and

Ruiz-Castillo (2011)

found that power laws account usually only for less than 2 % of the

highest-cited articles. Therefore, it seems that the coverage of the

right tails of citation distributions in our samples is satisfactory for

our purposes.

## Results and discussion

Table 3

presents results of fitting the discrete power-law model to our data

sets consisting of citations to scientific articles published over

1998–2002 (with a common 5-year citation window), separately for each of

the 27 Scopus major subject areas of science. The last row gives also

results for all subject areas combined (“All sciences”). Beside

estimates of the power-law exponent(α^) and the lower bound on the power-law behaviour (x^0) , the table gives also the estimated number and the percentage of power-law distributed papers, as well as the

presents results of fitting the discrete power-law model to our data

sets consisting of citations to scientific articles published over

1998–2002 (with a common 5-year citation window), separately for each of

the 27 Scopus major subject areas of science. The last row gives also

results for all subject areas combined (“All sciences”). Beside

estimates of the power-law exponent

*p*value for our goodness-of-fit test.Results

with respect to the goodness-of-fit suggest that the power-law

hypothesis cannot be rejected for the following 14 Scopus science

fields: “Agricultural and Biological Sciences”, “Biochemistry, Genetics

and Molecular Biology”, “Chemical Engineering”, “Chemistry”, “Energy”,

“Environmental Science”, “Materials Science”, “Neuroscience”, “Nursing”,

“Pharmacology, Toxicology and Pharmaceutics”, “Physics and Astronomy”,

“Psychology”, “Health Professions”, and “Multidisciplinary”. The

remaining 13 Scopus fields of science for which the power-law model is

rejected include humanities and social sciences (“Arts and Humanities”,

“Business, Management and Accounting”, “Economics, Econometrics and

Finance”, “Social Sciences”), but also formal sciences (“Computer

Science”, “Decision Sciences”, “Mathematics”), life sciences

(“Immunology and Microbiology”, “Medicine”, “Veterinary”, “Dentistry”),

as well as “Earth and Planetary Sciences” and “Engineering”. The best

power-law fits for these fields of science are shown on

Fig. 1.

with respect to the goodness-of-fit suggest that the power-law

hypothesis cannot be rejected for the following 14 Scopus science

fields: “Agricultural and Biological Sciences”, “Biochemistry, Genetics

and Molecular Biology”, “Chemical Engineering”, “Chemistry”, “Energy”,

“Environmental Science”, “Materials Science”, “Neuroscience”, “Nursing”,

“Pharmacology, Toxicology and Pharmaceutics”, “Physics and Astronomy”,

“Psychology”, “Health Professions”, and “Multidisciplinary”. The

remaining 13 Scopus fields of science for which the power-law model is

rejected include humanities and social sciences (“Arts and Humanities”,

“Business, Management and Accounting”, “Economics, Econometrics and

Finance”, “Social Sciences”), but also formal sciences (“Computer

Science”, “Decision Sciences”, “Mathematics”), life sciences

(“Immunology and Microbiology”, “Medicine”, “Veterinary”, “Dentistry”),

as well as “Earth and Planetary Sciences” and “Engineering”. The best

power-law fits for these fields of science are shown on

Fig. 1.

The complementary cumulative distribution functions (

*blue circles*) and best power-law fits (*dashed black line*) for citation distributions that did not pass the goodness-of-fit test, Scopus, 1998–2002, 5-year citation windowFor most of the distributions shown on Fig. 1,

it can be clearly seen that their right tails decay faster than the

pure power-law model indicates. This suggest that the largest

observations for these distributions should be rather modeled with a

distribution having a lighter tail than the pure power-law model like

the log-normal or power-law with exponential cut-off models.

it can be clearly seen that their right tails decay faster than the

pure power-law model indicates. This suggest that the largest

observations for these distributions should be rather modeled with a

distribution having a lighter tail than the pure power-law model like

the log-normal or power-law with exponential cut-off models.

The

value for our goodness-of-fit test in case of “All Sciences” is 0.076,

which is below our acceptance threshold of 0.1. However, this

values for most of the 13 Scopus fields of science for which we reject

the power-law hypothesis. For this reason, we conclude that the evidence

is not conclusive in this case. Our result for “All Sciences” is,

however, in a stark contrast with that of Albarrán and Ruiz-Castillo (2011), who using the WoS data found that the fit for a corresponding data set was very good (with a

*p*value for our goodness-of-fit test in case of “All Sciences” is 0.076,

which is below our acceptance threshold of 0.1. However, this

*p*value is non-negligible and significantly higher than*p*values for most of the 13 Scopus fields of science for which we reject

the power-law hypothesis. For this reason, we conclude that the evidence

is not conclusive in this case. Our result for “All Sciences” is,

however, in a stark contrast with that of Albarrán and Ruiz-Castillo (2011), who using the WoS data found that the fit for a corresponding data set was very good (with a

*p*value of 0.85).^{11}The

estimates of the power-law exponent for the 14 Scopus science fields

for which the power law seems to be a plausible hypothesis range from

3.24 to 4.69. This is in a good agreement with Albarrán and

Ruiz-Castillo (2011)

and confirms their assessment that the true value of this parameter is

substantially higher than found in the earlier literature (Redner 1998; Lehmann et al. 2003; Tsallis and deAlbuquerque 2000), which offered estimates ranging from around 2.3 to around 3. We also confirm the observation of Albarrán and Ruiz-Castillo (2011)

that power laws in citation distributions are rather small—they account

usually for less than 1 % of total articles published in a field of

science. The only two fields in our study with slightly “bigger” power

laws are “Chemistry” (2 %) and “Multidisciplinary” (2.8 %).

estimates of the power-law exponent for the 14 Scopus science fields

for which the power law seems to be a plausible hypothesis range from

3.24 to 4.69. This is in a good agreement with Albarrán and

Ruiz-Castillo (2011)

and confirms their assessment that the true value of this parameter is

substantially higher than found in the earlier literature (Redner 1998; Lehmann et al. 2003; Tsallis and deAlbuquerque 2000), which offered estimates ranging from around 2.3 to around 3. We also confirm the observation of Albarrán and Ruiz-Castillo (2011)

that power laws in citation distributions are rather small—they account

usually for less than 1 % of total articles published in a field of

science. The only two fields in our study with slightly “bigger” power

laws are “Chemistry” (2 %) and “Multidisciplinary” (2.8 %).

The comparison between the power-law hypothesis and alternatives using the Vuong’s test is presented in Table 4.

It can be observed that the exponential model can be ruled out in most

of the cases. We discuss other results first for the 13 Scopus fields of

science that did not pass our goodness-of-fit test. For all of these

fields, except for “Veterinary”, the Yule and power-law with exponential

cut-off models fit the data better than the pure power-law model in a

statistically significant way. The log-normal model is better than the

pure power-law model in 10 of the discussed fields; the same holds for

the Weibull distribution in case of 5 fields and for the digamma

distribution in case of 4 fields. However, these results do not imply

that the distributions, which give a better fit to the non-power-law

distributed data than the pure power-law model are plausible hypotheses

for these data sets. This issue should be further studied using

appropriate goodness-of-fit tests.

It can be observed that the exponential model can be ruled out in most

of the cases. We discuss other results first for the 13 Scopus fields of

science that did not pass our goodness-of-fit test. For all of these

fields, except for “Veterinary”, the Yule and power-law with exponential

cut-off models fit the data better than the pure power-law model in a

statistically significant way. The log-normal model is better than the

pure power-law model in 10 of the discussed fields; the same holds for

the Weibull distribution in case of 5 fields and for the digamma

distribution in case of 4 fields. However, these results do not imply

that the distributions, which give a better fit to the non-power-law

distributed data than the pure power-law model are plausible hypotheses

for these data sets. This issue should be further studied using

appropriate goodness-of-fit tests.

We

now turn to results for the remaining Scopus fields of science that

were not rejected by our goodness-of-fit test. The power-law hypothesis

seems to be the best model only for “Physics and Astronomy”. In this

case, the test statistics is always non-negative implying that the

power-law model fits the data as good as or better than each of the

alternatives. For the remaining 13 fields of science, the log-normal,

Yule and power-law with exponential cut-off models have always higher

log-likelihoods suggesting that these models may fit the data better

than the pure power-law distribution. However, only in a few cases the

differences between models are statistically significant. For

“Chemistry” and “Multidisciplinary” both the Yule and power-law with

exponential cut-off models are favoured over the pure power-law model.

The power-law with exponential cut-off is also favoured in case of

“Health Professions”. In other cases, the

likelihood ratio test are large, which implies that there is no

conclusive evidence that would allow to distinguish between the pure

power-law, log-normal, Yule and power-law with exponential cut-off

distributions. Comparing the power-law distribution with the Weibull and

Tsallis distributions, we observe that the sign of the test statistic

is positive in roughly half of the cases, but the

always large and neither model can be ruled out. For the considered 13

fields of science, the digamma model is never better than the power law,

judging by the sign of the test statistic. Our likelihood ratio tests

suggest therefore that when the power law is a plausible hypothesis

according to our goodness-of-fit test it is often indistinguishable from

some alternative models.

now turn to results for the remaining Scopus fields of science that

were not rejected by our goodness-of-fit test. The power-law hypothesis

seems to be the best model only for “Physics and Astronomy”. In this

case, the test statistics is always non-negative implying that the

power-law model fits the data as good as or better than each of the

alternatives. For the remaining 13 fields of science, the log-normal,

Yule and power-law with exponential cut-off models have always higher

log-likelihoods suggesting that these models may fit the data better

than the pure power-law distribution. However, only in a few cases the

differences between models are statistically significant. For

“Chemistry” and “Multidisciplinary” both the Yule and power-law with

exponential cut-off models are favoured over the pure power-law model.

The power-law with exponential cut-off is also favoured in case of

“Health Professions”. In other cases, the

*p*values for thelikelihood ratio test are large, which implies that there is no

conclusive evidence that would allow to distinguish between the pure

power-law, log-normal, Yule and power-law with exponential cut-off

distributions. Comparing the power-law distribution with the Weibull and

Tsallis distributions, we observe that the sign of the test statistic

is positive in roughly half of the cases, but the

*p*values arealways large and neither model can be ruled out. For the considered 13

fields of science, the digamma model is never better than the power law,

judging by the sign of the test statistic. Our likelihood ratio tests

suggest therefore that when the power law is a plausible hypothesis

according to our goodness-of-fit test it is often indistinguishable from

some alternative models.

Overall, our

results show that the evidence in favour of the power-law behaviour of

the right-tails of citation distributions is rather weak. For roughly

half of the Scopus fields of science studied, the power-law hypothesis

is rejected. Other distributions, especially the Yule, power-law with

exponential cut-off and log-normal seem to fit the data from these

fields of science better than the pure power-law model. On the other

hand, when the power-law hypothesis is not rejected, it is usually

empirically indistinguishable from all alternatives with the exception

of the exponential distribution. The pure power-law model seems to be

favoured over alternative models only for the most highly cited papers

in “Physics and Astronomy”. Our results suggest that theories implying

that the most highly cited scientific papers follow the Yule, power-law

with exponential cut-off or log-normal distribution may have slightly

more support in data than theories predicting the pure power-law

behaviour.

results show that the evidence in favour of the power-law behaviour of

the right-tails of citation distributions is rather weak. For roughly

half of the Scopus fields of science studied, the power-law hypothesis

is rejected. Other distributions, especially the Yule, power-law with

exponential cut-off and log-normal seem to fit the data from these

fields of science better than the pure power-law model. On the other

hand, when the power-law hypothesis is not rejected, it is usually

empirically indistinguishable from all alternatives with the exception

of the exponential distribution. The pure power-law model seems to be

favoured over alternative models only for the most highly cited papers

in “Physics and Astronomy”. Our results suggest that theories implying

that the most highly cited scientific papers follow the Yule, power-law

with exponential cut-off or log-normal distribution may have slightly

more support in data than theories predicting the pure power-law

behaviour.

## Conclusions

We

have used a large, novel data set on citations to scientific papers

published between 1998 and 2002 drawn from Scopus to test empirically

for the power-law behaviour of the right-tails of citation

distributions. We have found that the power-law hypothesis is rejected

for around half of the Scopus fields of science. For the remaining

fields of science, the power-law distribution is a plausible model, but

the differences between the power law and alternative models are usually

statistically insignificant. The paper also confirmed recent findings

of Albarrán and Ruiz-Castillo (2011)

that power laws in citation distributions, when they are a plausible,

account only for a very small fraction of the published papers (less

than 1 % for most of science fields) and that the power-law exponent is

substantially higher than found in the older literature.

have used a large, novel data set on citations to scientific papers

published between 1998 and 2002 drawn from Scopus to test empirically

for the power-law behaviour of the right-tails of citation

distributions. We have found that the power-law hypothesis is rejected

for around half of the Scopus fields of science. For the remaining

fields of science, the power-law distribution is a plausible model, but

the differences between the power law and alternative models are usually

statistically insignificant. The paper also confirmed recent findings

of Albarrán and Ruiz-Castillo (2011)

that power laws in citation distributions, when they are a plausible,

account only for a very small fraction of the published papers (less

than 1 % for most of science fields) and that the power-law exponent is

substantially higher than found in the older literature.

## Acknowledgments

I

would like to thank two anonymous referees for helpful comments and

suggestions that improved this paper. The use of Matlab and R software

accompanying the papers by Clauset et al. (2009), Shalizi (2007) and Peterson et al. (2013) is gratefully acknowledged. Any remaining errors are my responsibility.

would like to thank two anonymous referees for helpful comments and

suggestions that improved this paper. The use of Matlab and R software

accompanying the papers by Clauset et al. (2009), Shalizi (2007) and Peterson et al. (2013) is gratefully acknowledged. Any remaining errors are my responsibility.

## Footnotes

^{1}From

the perspective of measuring power laws in citation distributions, the

most important part of the distribution is the right tail. It seems that

the database used in this paper has a better coverage of the right tail

of citation distributions. The most highly cited paper in our database

has received 5,187 citations (see Table 2), while the corresponding number for the database based on WoS is 4,461 (Li and Ruiz-Castillo 2013). Our database is further described in “Materials and methods” section.

^{2}

Clauset et al. (2009) provide also an approximate method of estimating

*α*

for the discrete power-law model by assuming that continuous power-law

distributed reals are rounded to the nearest integers. However, it this

paper we use an exact approach based on maximizing (2).

^{3}If

our data were drawn from a given model, then we could use the KS

statistic in testing goodness of fit, because the distribution of the KS

statistic is known in such a case. However, when the underlying model

is not known or when its parameters are estimated from the data, which

is our case, the distribution of the KS statistic must be obtained by

simulation.

^{4}In this

goodness-of-fit test we are interested in verifying if the power-law

model is a plausible hypothesis for our data sets. Hence, high

*p*

values suggest that the power law is not ruled out. This approach is to

be contrasted with the usual approach, which for a given null

hypothesis interprets low

*p*values as evidence in favor of the alternative hypothesis. See Clauset et al. (2009) for a more detailed discussion of these interpretations of

*p*values.

^{6}The power-law with exponential cut-off behaves like the pure power-law model for smaller values of

*x*,

*x*⩾

*x*

_{0},while for larger values of

*x*

it behaves like an exponential distribution. The pure power-law model

is nested within the power-law with exponential cut-off, and for this

reason the latter always provides a fit at least as good as the former.

^{7}I

would like to thank an anonymous referee for suggesting the inclusion

of this distribution in our comparison of alternative models.

^{8}See Table 2 for a list of the analyzed Scopus areas of science.

^{9}For

example, for articles published in 1998 we have analyzed citations

received during 1998–2002, while for articles published in 2002, those

received during 2002–2006.

^{10}For

all fields of science analyzed, there were some articles with missing

information on citations. These articles were removed from our samples.

However, this has usually affected only about 0.1 % of our samples.

^{11}In Albarrán and Ruiz-Castillo (2011),

the power-law hypothesis is found plausible for 17 out of 22 WoS fields

of science. It is rejected for “Pharmacology and Toxicology”,

“Physics”, “Agricultural Sciences”, “Engineering”, and “Social Sciences,

General”. These results are not directly comparable with those of the

present paper as Scopus and WoS use different classification systems to

categorize journals.

## References

- Albarrán P, Crespo JA, Ortuño I, Ruiz-Castillo J. The skewness of science in 219 sub-fields and a number of aggregates. Scientometrics. 2011;88(2):385–397. doi: 10.1007/s11192-011-0407-9. [Cross Ref]
- Albarrán,

P., Crespo, J.A., Ortuño, I., & Ruiz-Castillo, J. (2011). The

skewness of science in 219 sub-fields and a number of aggregates.

Working paper 11–09, Universidad Carlos III. - Albarrán P, Ruiz-Castillo J. References made and citations received by scientific articles. Journal of the American Society for Information Science and Technology. 2011;62(1):40–49. doi: 10.1002/asi.21448. [Cross Ref]
- Anastasiadis

AD, deAlbuquerque MP, deAlbuquerque MP, Mussi DB. Tsallis q-exponential

describes the distribution of scientific citations—A new

characterization of the impact. Scientometrics. 2010;83(1):205–218. doi: 10.1007/s11192-009-0023-0. [Cross Ref] - Baayen RH. Word frequency distributions. Dordrecht: Kluwer; 2001.
- Chadegani

AA, Salehi H, Md Yunus M, Farhadi H, Fooladi M, Farhadi M, Ebrahim NA. A

comparison between two main academic literature collections: Web of

Science and Scopus databases. Asian Social Science. 2013;9(5):18–26. doi: 10.5539/ass.v9n5p18. [Cross Ref] - Clauset A, Shalizi CR, Newman ME. Power-law distributions in empirical data. SIAM Review. 2009;51(4):661–703. doi: 10.1137/070710111. [Cross Ref]
- de Solla Price D. Networks of scientific papers. Science. 1965;149:510–515. doi: 10.1126/science.149.3683.510. [PubMed] [Cross Ref]
- de Solla Price D. A general theory of bibliometric and other cumulative advantage processes. Journal of the American Society for Information Science. 1976;27(5):292–306. doi: 10.1002/asi.4630270505. [Cross Ref]
- Egghe L. Power laws in the information production process: Lotkaian informetrics. Oxford: Elsevier; 2005.
- Eom YH, Fortunato S. Characterizing and modeling citation dynamics. PloS One. 2011;6(9):e24,926. doi: 10.1371/journal.pone.0024926. [PMC free article] [PubMed] [Cross Ref]
- Gabaix X. Power laws in economics and finance. Annual Review of Economics. 2009;1(1):255–294. doi: 10.1146/annurev.economics.050708.142940. [Cross Ref]
- Golosovsky M, Solomon S. Runaway events dominate the heavy tail of citation distributions. The European Physical Journal Special Topics. 2012;205(1):303–311. doi: 10.1140/epjst/e2012-01576-4. [Cross Ref]
- Laherrère J, Sornette D. Stretched exponential distributions in nature and economy:“Fat tails” with characteristic scales. The European Physical Journal B. 1998;2(4):525–539. doi: 10.1007/s100510050276. [Cross Ref]
- Lehmann S, Lautrup B, Jackson A. Citation networks in high energy physics. Physical Review E. 2003;68(2):026,113. doi: 10.1103/PhysRevE.68.026113. [PubMed] [Cross Ref]
- Li,

Y., & Ruiz-Castillo, J. (2013). The impact of extreme observations

in citation distributions. Tech. rep., Universidad Carlos III,

Departamento de Economía. - López-Illescas

C, deMoya-Anegón F, Moed HF. Coverage and citation impact of

oncological journals in the Web of Science and Scopus. Journal of Informetrics. 2008;2(4):304–316. doi: 10.1016/j.joi.2008.08.001. [Cross Ref] - Lotka A. The frequency distribution of scientific productivity. Journal of Washington Academy Sciences. 1926;16(12):317–323.
- Newman ME. Power laws, Pareto distributions and Zipf’s law. Contemporary Physics. 2005;46(5):323–351. doi: 10.1080/00107510500052444. [Cross Ref]
- Peterson GJ, Pressé S, Dill KA. Nonuniversal power law scaling in the probability distribution of scientific citations. Proceedings of the National Academy of Sciences. 2010;107(37):16,023–16,027. doi: 10.1073/pnas.1010757107. [PMC free article] [PubMed] [Cross Ref]
- Peterson J, Dixit PD, Dill KA. A maximum entropy framework for nonexponential distributions. Proceedings of the National Academy of Sciences. 2013;110(51):20,380–20,385. doi: 10.1073/pnas.1320578110. [PMC free article] [PubMed] [Cross Ref]
- Radicchi

F, Fortunato S, Castellano C. Universality of citation distributions:

Toward an objective measure of scientific impact. Proceedings of the National Academy of Sciences. 2008;105(45):17,268–17,272. doi: 10.1073/pnas.0806977105. [PMC free article] [PubMed] [Cross Ref] - Redner S. How popular is your paper? An empirical study of the citation distribution. The European Physical Journal B. 1998;4(2):131–134. doi: 10.1007/s100510050359. [Cross Ref]
- Redner S. Citation statistics from 110 years of Physical Review. Physics Today. 2005;58:49–54. doi: 10.1063/1.1996475. [Cross Ref]
- Shalizi, C.R. (2007). Maximum likelihood estimation for q-exponential (Tsallis) distributions. Tech. rep., arXiv preprint math/0701854.
- Stringer MJ, Sales-Pardo M, Amaral LAN. Effectiveness of journal ranking schemes as a tool for locating information. PLoS One. 2008;3(2):e1683. doi: 10.1371/journal.pone.0001683. [PMC free article] [PubMed] [Cross Ref]
- Tsallis C, deAlbuquerque MP. Are citations of scientific papers a case of nonextensivity? The European Physical Journal B. 2000;13(4):777–780. doi: 10.1007/s100510050097. [Cross Ref]
- VanRaan AFJ. Statistical properties of bibliometric indicators: Research group indicator distributions and correlations. Journal of the American Society for Information Science and Technology. 2006;57(3):408–430. doi: 10.1002/asi.20284. [Cross Ref]
- Vuong QH. Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica. 1989;57(2):307–333. doi: 10.2307/1912557. [Cross Ref]
- Wallace ML, Larivière V, Gingras Y. Modeling a century of citation distributions. Journal of Informetrics. 2009;3(4):296–303. doi: 10.1016/j.joi.2009.03.010. [Cross Ref]

Power laws in citation distributions: evidence from Scopus

## No comments:

## Post a comment