Characterizing the effect of retractions on publishing careers

TL;DR

The findings suggest that retractions may impose a disproportionate impact on early career authors, and retracted authors generally retain less senior and less productive co-authors, but gain more impactful co-authors post-retraction.

Abstract

Retracting academic papers is a fundamental tool of quality control, but it may have far-reaching consequences for retracted authors and their careers. Previous studies have highlighted the adverse effects of retractions on citation counts and coauthors' citations; however, the broader impacts beyond these have not been fully explored. We address this gap leveraging Retraction Watch, the most extensive data set on retractions and link it to Microsoft Academic Graph and Altmetric. Retracted authors, particularly those with less experience, often leave scientific publishing in the aftermath of retraction, especially if their retractions attract widespread attention. However, retracted authors who remain active in publishing maintain and establish more collaborations compared to their similar non-retracted counterparts. Nevertheless, retracted authors generally retain less senior and less productive coauthors, but gain more impactful coauthors post-retraction. Our findings suggest that retractions may impose a disproportionate impact on early-career authors.

Figures (21)

• Figure 1: Characteristics of retracted authors.(a) shows percentage of retracted authors who left (blue) versus those who did not (red). (b) shows comparisons across different author-level characteristics among retracted authors who left scientific publishing at the time of retraction, and those who have not with $N=12,742$. (c) similar to (b) but for paper-level characteristics with $N=4,267$. The boxes extend from the lower to upper quartile values of the data, with a line at the median; whiskers extend to the minimum and maximum values within 1.5 times the interquartile range (IQR) of the lower and upper quartiles, respectively; means are shown as additional square markers. Normality and homogeneity of variances were tested and not met; thus, $p$-values were calculated using a non-parametric two-sided Mann-Whitney U test. Outliers are removed from box plots for presentation purposes. *** $p<.001$
• Figure 2: Retraction and attrition, and the role of attention(a) Cumulative visualization of the difference in publishing career length between retracted authors with known attrition years and their matched non-retracted counterparts. (b) Raincloud plot allen2019raincloud showing the distribution of logged Altmetric score 6 months pre- and post-retraction. The x-axis represents monthly windows between the retraction and attention, with 0 being the day of retraction (not displayed), -1 the month right before, and 1 is the month after. The y-axis shows the logged Altmetric score for a paper in the given month. Note, that Altmetric scores [0, 1] are frequent, e.g., 1 tweet results in a score of 0.25. The boxes extend from the lower to upper quartile values of the data, with a line at the median; whiskers extend to the minimum and maximum values within 1.5 times the interquartile range (IQR) of the lower and upper quartiles, respectively. The black trend line represents the average logged Altmetric score. Total number of papers being plotted is 6,507. Papers with no attention within the 12-month window are excluded (based on abhari2022twitter). Comparison across months shows retracted papers receive the most attention within one month of retraction. (c) Gap between the cumulative proportion of high attention retracted authors who left publishing and their matched counterparts, shown at different cutoffs of attention as measured by the Altmetric score.
• Figure 3: Analyzing collaborator retention, gain, and triadic closure among retracted authors who stayed in scientific publishing post-retraction. The figure shows the difference between (a) the numbers of collaborators retained, (b) the numbers of collaborators gained, and (c) the proportions of triads closed 5 years post-retraction for the ($N=2,348$) authors who were retracted (red circle), and their matched non-retracted pairs (green square). These are further stratified by gender, year of retraction, academic age, author order, reason of retraction, type of retraction, and discipline. Data are presented as mean values. Solid line represents statistically significant difference using two-sided Welch's t-test (assuming unequal variances). Supplementary Tables 6–8 present the 95% confidence intervals ($CI_{95\%}$), as well as results from additional non-parametric tests.
• Figure 4: Comparison of characteristics of collaborators retained and gained by retracted and non-retracted authors, stratified by seniority ($N=1,187$) and by reasons of retraction ($N=1,143$).(a-c) display the comparison based on academic age, number of papers, and number of citations for retained collaborators of retracted and non-retracted authors of different age groups. (d-f) show the comparison for collaborators gained. (g-i) illustrate the results of the difference-in-difference analysis comparing the difference of collaborators retained and collaborators lost. (j-r) show similar comparison for collaborators of retracted and non-retracted authors stratified by reasons of retraction.
• Figure Supplementary Figure 1: Distribution of the longest gap found in the publication career of all STEM scientists in Microsoft Academic Graph up to the 2010 cohort.