Requirements for a cocitation similarity measure, with special reference to Pearson's correlation coefficient

Abstract

Author cocitation analysis (ACA), a special type of cocitation analysis, was introduced by White and Griffith in 1981.This technique is used to analyze the intellectual structure of a given scientific field. In 1990, McCain published a technical overview that has been largely adopted as a standard. Here, McCain notes that Pearson's correlation coefficient (Pearson's r) is often used as a similarity measure in ACA and presents some advantages of its use. The present article criticizes the use of Pearson's r in ACA and sets forth two natural requirements that a similarity measure applied in ACA should satisfy. It is shown that Pearson's r does not satisfy these requirements. Real and hypothetical data are used in order to obtain counterexamples to both requirements. It is concluded that Pearson's r is probably not an optimal choice of a similarity measure in ACA. Still, further empirical research is needed to show if, and in that case to what extent, the use of similarity measures in ACA that fulfill these requirements would lead to objectively better results in full-scale studies. Further, problems related to incomplete cocitation matrices are discussed.

On the normalization and visualization of author co‐citation data: Salton's Cosine versus the Jaccard index

“…The argument is illustrated with an analysis using the same data as in Ahlgren et al (2003). This dataset (provided in Table 1) is extremely structured: It contains exclusively positive correlations within both groups and negative correlations between the two groups.…”

“…In principle, one can normalize both symmetrical and asymmetrical matrices with the various measures. Ahlgren et al (2003) provided arguments for using the cosine instead of the Pearson correlation coefficient, particularly if one aims at visualization of the structure as in the case of social network analysis or multidimensional scaling (MDS). Bensman (2004) provided arguments regarding why one might nevertheless prefer the Pearson correlation coefficient when the purpose of the study is a statistical (e.g., multivariate) analysis.…”

Entrepreneurship, structural change, and economic growth

“…Since using different methods to calculate similarity measures might yield different results, we compare two different measures: a correlation coefficient similarity measure and a cosine similarity measure. The appropriateness of using the correlation coefficient similarity measure is questioned by Ahlgren et al (2003) in a paper that analyzes its aptness for measuring authors' co-citation profiles. The authors argue that this measure is sensitive to zeros.…”

Similarity measures, author cocitation analysis, and information theory

“…Like White (2003), I applied the above algorithm to the dataset which was published by Ahlgren et al (2003) in Table 7, at p. 555. Since the algorithm proposed above is not included in standard software packages I had to construct a visualization in Excel based on the numeric output.…”

“…In a provocative study, Ahlgren et al (2003) questioned the use of Pearson's Correlation Coefficient as a similarity measure in author cocitation analysis (ACA) with the argument that this measure is sensitive for zeros. Analytically, the addition of zeros to two variables should add to their similarity, but the authors show with empirical examples that this addition can depress the correlation coefficient between these variables.…”