On the delay in publication of papers in scientific journals

This article was downloaded by: [University of Ioannina] On: 11 November 2014, At: 04:18 Publisher: Taylor & Francis Informa Ltd Registered in England...
Author: Baldric Stokes
1 downloads 0 Views 268KB Size
This article was downloaded by: [University of Ioannina] On: 11 November 2014, At: 04:18 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Applied Statistics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/cjas20

On the delay in publication of papers in scientific journals S. Loukas

a

a

Department of Mathematics , University of Ioannina , Ioannina, 45332, Greece Published online: 24 May 2006.

To cite this article: S. Loukas (1984) On the delay in publication of papers in scientific journals, Journal of Applied Statistics, 11:2, 163-169, DOI: 10.1080/02664768400000017 To link to this article: http://dx.doi.org/10.1080/02664768400000017

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Journal of Applied Statistics, Vol 11, No 2, 1984 ON THE DELAY

IN PUBLICATIM\I ff PAPERS IN SCIENTIFIC JOURNALS

S Loukas D e p a r t m e n t of Mathematics, University of Ioannina Ioannina 45332, G r e e c e Summary

Downloaded by [University of Ioannina] at 04:18 11 November 2014

The technique of multiple regression is used in a study of t h e f a c t o r s which influence t h e delay of publication of papers in scientific journals.

The d a t a used

were obtained from journals of various subjects and countries of origin over a f i v e year period. The main results a r e t h a t physics and biology journals publish papers significantly f a s t e r than m a t h e m a t i c s journals and t h a t journals behave in a n independent manner.

1.

Introduction

Scientific journals play a major role in t h e progress of modern science.

Their

importance lies in t h a t they a r e responsible for informing t h e scientific world and t h e public on t h e l a t e s t advances in t h e various fields on scientific endeavour. Quite understandably, t h e sooner a scientific work is published t h e b e t t e r i t is f o r t h e s a k e of all concerned; indeed i t is for this reason t h a t t h e scientific journal was c r e a t e d in t h e f i r s t place. I t would be ideal if t h e completion of a research paper could be followed by i t s i m m e d i a t e publication

- but this is not t h e case.

The f a c t s show t h a t t h e r e is a considerable delay between submission and t h e a c t u a l publication of a r e s e a r c h paper. The main reason for this is t h a t before t h e paper is a c c e p t e d f o r publication i t h a s t o be thoroughly examined for i t s originality and possible errors.

On t h e o t h e r hand research work deals with

problems which present a varying d e g r e e of subject difficulty during t h e i r examination and variable t i m e periods a r e t h e r e f o r e required in dealing with e a c h subject's specific problems. T h e refereeing process and t h e editing and publishing

process, differing f r o m one journal t o another, a r e f a c t o r s possibly affecting t h e delay in publication. One could also e x p e c t a journal's country of origin t o have a n e f f e c t on t h e delay of publication of a paper due t o t h e different socio-economic and educational s y s t e m s existing between t h e countries.

Both t h e refereeing

c o m m i t t e e and t h e publishing company, main f a c t o r s upon which t h e a c t u a l policy

Downloaded by [University of Ioannina] at 04:18 11 November 2014

of a journal depends, do not appreciably change from one year t o another and t h e r e f o r e o n e would e x p e c t t h e delay t o remain m o r e or less constant from year t o year.

Revised papers do not give information on t h e amount of t h e delay t h a t is

a t t r i b u t e d t o revision, so t h a t f o r revised papers, only t h e delay in publication of t h e resubmitted version should b e examined.

With s o many f a c t o r s causing delay in t h e publication of r e s e a r c h papers, it is only t o be e x p e c t e d t h a t variation should exist and, bearing in mind t h e significance of a f a s t publication, w e thought t h a t a study of t h e f a c t o r s t h a t influence t h e delay of publication of r e s e a r c h papers might be of interest.

D a t a w e r e obtained f r o m t h r e e journals taken f r o m e a c h of t h r e e subject m a t t e r a r e a s with a l l subjects t a k e n from e a c h of t h r e e countries over a five year period t o enable s a t i s f a c t o r y results t o b e given. To avoid d i f f e r e n c e s b e t w e e n obviously unrelated s u b j e c t s a s physics and history, t h e subjects m a t h e m a t i c s , physics and biology w e r e chosen. On t h e o t h e r hand England, USA and Germany w e r e t h e only countries t h a t could provide t h e required journals, although only two G e r m a n journals in physics w e r e found.

Taking into a c c o u n t results of s o m e preliminary t - t e s t analysis performed on d a t a obtained o v e r a five year period (1969-73) w e decided t o use t h i r t y observations per year f r o m t h r e e y e a r s (69, 71, 73) in t h e main analysis (delay did not vary significantly o v e r t h e years).

The most direct method of investigating t h e e f f e c t s of t h e proposed f a c t o r s and a satisfactory way t o obtain some idea of their predictive capacity is t o use multiple regression analysis.

The d a t a analysis is described in Section 2 below.

Conclusions and delay

Downloaded by [University of Ioannina] at 04:18 11 November 2014

predictions a r e given in Section 3. The 26 journals chosen and t h e d a t a obtained and used in this study a r e given in Loukas (1976). Calculations were done on t h e computer a t Sussex University, England by using t h e s t a t i s t i c a l package XDS3 (Multiple Regression Procedure 2).

2.

Statistical analysis and results

The types of delay variation t h a t could possibly exist, consist of variation between journals and within journals and a complete study thus requires both types of analysis.

The first one, called Model I analysis, considers t h e journal a s t h e experimental unit and seeks f a c t o r s explaining t h e differing behaviour of t h e journals and t h e other, called Model I1 analysis, considers t h e paper a s t h e experimental unit and seeks sources of delay variation of t h e published papers within t h e journal itself.

2.1

Model I analysis

In regression analysis, a n independent variable may be included, either a s an estimator of t h e dependent variable accounting for a considerable e x t e n t of t h e variation of t h e l a t t e r , or a s a correction t o t h e dependent and t h e other independent variables, s o a s t o strengthen t h e relationship found between them.

In this analysis, t h e general regression model consisted of t h e dependent variable Y, t h e average delay (in months) of t h e journal's published papers, and t h e independent variables X l i , denoting t h e f a c t o r s t h a t might have a f f e c t e d Y, a s follows:

X11, X12, dummy variables accounting f o r t h e country e f f e c t ; X13, X14, dummy variables accounting for t h e subject e f f e c t ; X15,

.....,X l 8 ,

dummy variables accounting f o r possible interaction between t h e

above t w o sets; X19, t h e p e r c e n t a g e of revised papers per journal; and

Downloaded by [University of Ioannina] at 04:18 11 November 2014

X110, t h e a v e r a g e number of pages of t h e journal's published papers.

Firstly, i t w a s considered whether t h e delay had been constant over t h e years in t h e analysis. If this were t h e c a s e then all d a t a could be used in a single analysis regardless of year, whereupon t h e variable values would be averages of 9U observations instead of 30, t h u s yielding more precise e s t i m a t e s of t h e underlying model.

Also, t h e conclusions derived from t h e analysis would be general and

applicable t o any year of interest.

An appropriate procedure is t h a t of testing

regression c o e f f i c i e n t s (parallelism) followed by testing differences in t h e position (coincidence) when regression lines a r e found t o be parallel [Seber, 1977).

The

calculated t e s t s t a t i s t i c in testing f o r parallelism of t h e regression lines f o r t h e years 69, 71, 7 3 ws F = 0.51 distributed a s ( - 1 F20,45, a value t h a t cannot r e j e c t t h e hypothesis t h a t t h e t h r e e regression lines w e r e parallel. s t a t i s t i c in t e s t i n g f o r coincidence w a s F = 2.00 value.

- F2,45

The calculated t e s t not a

Therefore t h e regression lines w e r e identical, in o t h e r words t h e delay was

found t o h a v e been c o n s t a n t over t h e years.

In t h e sequel, all d a t a w e r e incorporated in one analysis irrespective of year difference with t h e a v e r a g e s f o r t h e Y , X19 and X l l o values based on t h e 90 observations t h a t w e r e obtained per journal.

Next, interaction t e r m s w e r e t e s t e d for significance since t h e s e usually m a k e t h e model m o r e complicated and t h e i n t e r p r e t a t i o n of t h e results m o r e difficult. The calculated t e s t s t a t i s t i c w a s F = 1.16-F4,15

s o t h a t , since t h i s is not a significant

value, interaction t e r m s were not included in t h e regression models considered subsequently.

The average number of t h e published papers per journal did not account for a significant amount of t h e variation in t h e delay from journal t o journal (t = 0.26-

Downloaded by [University of Ioannina] at 04:18 11 November 2014

t15) while t h e percentage of revised papers per journal did account significantly ( a t 2.5%) in this variation (t = -2.53- t15).

There was found no significant variation in average delay per journal between countries, t h e calculated t e s t s t a t i s t i c for this being F = 0.99-F2,19.

In testing for subject effects, i t was found t h a t F = 10.67- F2,19, significant a t 1% indicating a wide variation in delay between subjects.

It was noted t h a t t h e regression equations were significant only when subject t e r m s were included in t h e model and t h e simplest (significant a t 1%)estimated regression equation useful for prediction was t h e following (delay in months)

Residual analysis showed no departures f r o m t h e normality and constant variance assumptions of t h e regression models.

2.2

Model I1 analysis

The regression model in this analysis consisted of t h e a c t u a l delay in months of t h e paper published a s t h e dependent variable Y and t h e independent variables X21 accounting for t h e e f f e c t of t h e number of pages of t h e paper on t h e delay and Xgz indicating whether t h e paper was a revised one or not.

The procedure of testing for difference of regressions (parallelism) was used in examining whether each journal was following its own policy and t h e calculated

test statistic was

F = 3.002- F46,2266, a significant value a t 1%, indicating t h a t

journals behave in an independent manner.

Hence, in order t o inquire whether

revision or length of papers contribute significantly t o t h e variation of delay within a journal, s e p a r a t e analyses were performed on t h e d a t a for each journal.

I t was found t h a t in most of t h e journals (17 out of 26) t h e length of a paper did not

Downloaded by [University of Ioannina] at 04:18 11 November 2014

a f f e c t significantly i t s publication delay, whilst for most of t h e journals (20 out of 2 6 ) there was a significant difference in delay between revised and non-revised

papers, with t h e revised papers published f a s t e r than t h e non-revised ones.

This

might have been expected since revised papers, having passed through a first examination, should be published f a s t e r than t h e non-revised ones.

3.

Conclusions

The results indicate t h a t delay in publication has been constant over t h e years and further, t h a t t h e r e is no significant variation in delay between countries. This l a t t e r finding is perhaps contrary t o what could be expected but t h e result is strongly in support of this contention and this could be attributed, amongst o t h e r things, to similar educational systems, similar cultural levels and t o t h e wide cooperation between t h e countries examined.

There is a wide variation in delay between subjects and it was found t h a t t h e r e is a significant variation in delay between t h e journals of mathematics and those of physics a s well a s between t h e journals of mathematics and those of biology but no significant variation between t h e journals of physics and those of biology was found. This could be supported by t h e view t h a t physics and biology a r e relatively similar subjects compared t o mathematics.

The average number of pages of t h e journal's papers does not contribute significantly t o t h e variation in delay between journals, whereas t h e percentage of

revised papers per journal does contribute i n the sense that the higher the number of revised papers, the less the average delay per journal.

Delay for papers submitted to a mathematical journal is estimated a t 11.5 months, whilst the delay for papers submitted to journals of physics and biology is

Downloaded by [University of Ioannina] at 04:18 11 November 2014

estimated a t 5.9 and 7.6 months respectively.

Finally, it was found that there are wide variations i n delay between journals and that journals vary i n the proportion of published papers referred back t o authors for revision. These variations are due to each journal's own policy.

Acknowledqements

The significant guidance offered t o the author i n connection w i t h this work by Professor J F Scott and Dr

J Haigh is hereby thankfully acknowledged.

References LOUKAS, S. (1976). "An investigation on the factors which influence the delay i n publication of papers i n various scientific journals". MSc Sussex University, England. SEBER, G.A.F. York.

(1977).

Linear regression analysis.

John Wiley and Sons Inc., New

Suggest Documents