Alternatives to the impact factor

Objective: To explore alternative bibliometric markers to the well-established journal impact factor. The bibliometric evolution of a leading ENT journal over a six year period is discussed with critical analysis of a predetermined set of bibliometric alternatives to the journal impact factor.

Design: Retrospective review of the bibliometric performance of Clinical Otolaryngology over a six year period.

Results: The results of the study reveal that Clinical Otolaryngology has made steady bibliometric progress when the impact factor (IF) is considered with a gradual increase in impact factor from 1.098 in 2006 to a peak of 2.393 in 2011. Self-citation rates reported by the Journal Citation Report (JCR) demonstrated a significant decline during 2007 with a reported self-citation rate of 0%. The SCImago Journal Rank (SJR) database however recorded a self-citation rate of 67. Independent evaluation demonstrated a 56 self-citations during this period. The percentage of review articles published remained stable during the period in question. A lagged association between the number of review manuscripts and the IF failed to demonstrate any significant correlation (r = -0.19). Comparison between the IF and the Eigen factor (EF) as well as the SJR yielded negative correlation (r = -0.46) and (r = -0.35) respectively. The Article Influence score (AIS) and Source Normalised Impact per Paper (SNIP) were the only bibliometric alternatives to demonstrate a positive correlation when compared to the IF (r = 0.94) and (r = 0.66) respectively.

Conclusions: The necessity of bibliometric markers cannot be called into question however the most widely employed of these, the journal impact factor has come under increased scrutiny of late. Despite some of the advantages offered by novel bibliometric markers, these do not necessarily compare favourably to the IF with regards to bibliometric performance. The only two markers to demonstrate a positive correlation when compared to the IF were the AI score and SNIP which would suggest that these are potential alternatives to the IF and have the added advantage that they are open access.