A Wavelet Analysis of Scientific Papers and Time Dependent Factors: A Scientometric Look

Abstract

In this study, we use wavelet approach to study the scientific research papers received by two journals i.e., i) Journal of Informetrics and ii) Research Policy, by using monthly observations during the period of 1970:2 to 2010:2 from article received history of both publishers database. Instead of analyzing the time series at their original level, as it is usually done, this study decompose the two variables i.e., scientific papers and the  time dependent factors at various scales of resolution using wavelet decomposition and then we study the relationships among the decomposed series matched to their scales. A major finding of the study is that some variations in received articles explained by many other time dependent factors that are previously not captured due to uneven time lags, therefore, an estimate with a time trend is made in this study. Both variables are not constant over frequency bands as it depends on the time scales. The two series (i.e., Springer and time factors) are correlated at higher frequencies, but their correlations measures in time dimension do not last long. However, during two structural breaks i.e., 1990-1995 and 2007-2011, the two time series seems to be correlated for longer period with frequency band higher than 321days. The results further provides an evidence that the two time series are in phase and the Elsevier scientific papers are time dependent in the long run. Thus, the analysis of wavelet coherence reveal that the two series i.e., scientific papers and time dependent factors correlated during the sample period.