18
Suicide and Life-Threatening Behavior 45 (1) February 2015 © 2014 The American Association of Suicidology DOI: 10.1111/sltb.12106
Predictors of Article Impact in Suicidology: The Bereavement Literature, A Research Note KARL ANDRIESSEN, MSUICIDOLOGY, KAROLINA KRYSINSKA, PHD, AND STEVEN STACK, PHD
Citation analysis has been neglected in suicidology. The present note applies a mixed-methods approach to both test and suggest hypotheses for the variation in article impact in the bereavement literature. One hundred three articles from three core suicidology journals met the criteria for inclusion in the investigation. Citations to the articles were obtained from the Web of Science. Predictor variables included structural characteristics of the author (e.g., gender) and the article itself (e.g., years since publication). A multivariate regression analysis determined that, controlling for the other variables, the most important predictor of citations was the review article (b = .461), followed by year of publication (b = .414), the multiauthored article (b = .302), publication in Suicide and Life-Threatening Behavior (SLTB) (b = .161), and male gender (b = .156). The 12 most cited articles were published between 1979 and 2004 in SLTB. The majority of these papers was written by males, were U.S. authors, and had more than one author. Four of the most cited articles were reviews. The study concludes that structural characteristics of articles and authors explained 41% of the variance in citations. The qualitative analysis determined that review papers, and papers on characteristics of suicide bereavement and psychological autopsies have been most frequently cited. Replication studies are needed for other subfields of suicidology.
Citation analysis helps to track the leading topics and the researchers who are having the greatest impact on a field. Citation analysis has blossomed in many scientific fields including anesthesiology, chemistry, fertility and sterility, deviant behavior, general medicine, pediatrics, public health, rehabilitation, KARL ANDRIESSEN, Faculty of Psychology and Educational Sciences, KU Leuven – University of Leuven, Leuven, Belgium; KAROLINA KRYSINSKA, Clinical Psychology, KU Leuven – University of Leuven, Leuven, Belgium; STEVEN STACK, Department of Psychiatry & Department of Criminology, Wayne State University, Detroit, MI, USA. Address correspondence to Karl Andriessen, Faculty of Psychology and Educational Sciences, KU Leuven – University of Leuven, Vanderkelenstraat 32, 3000 Leuven, Belgium; E-mail: [email protected]
and trauma research (Callaham, Wears, & Weber, 2002; Garfield, 1987; Hall, 1998; Holden, Rosenberg, & Barker, 2005; Key, 1988; Mehlman & Wenger, 2006; Ollerton & Sugrue, 2005; Seiber, 2010; Smith & Leggat, 2008; Stack, 2013; Yang & Pan, 2006). While there is a growing literature on such analyses, there is scant attention paid to these issues in suicidology. Only two article impact studies are available in suicide studies. Both of these are mainly descriptive works that provide lists of the most cited articles in Suicide and Life-Threatening Behavior (Cardinal, 2008; Stack, 2012). The present note contributes to the literature by applying a mixed-methods approach to citation analysis in three core suicidology journals. It goes beyond descriptive work in the previous studies
ANDRIESSEN
ET AL.
(Cardinal, 2008; Stack, 2012) by testing a series of structural hypotheses to explain the variation in citation impact in a sample of articles on bereavement and suicide, including psychological autopsy studies specifically discussing survivor issues in this type of research (Andriessen, 2014). These hypotheses involve structural characteristics of the article (e.g., year of publication, number of authors), characteristics of the author (e.g., gender, nationality), and the journal (e.g., impact factor). Second, it provides a qualitative analysis of the articles to assist in further explaining the variation in article impact on the field.
19 2005 to 2013, and for Archives of Suicide Research from 2006 to 2013. Suicidology Online was not cited in WOS. The sample contains 103 articles published during 1976 through 2013. Citations Citations to each of the bereavement articles were obtained from the online version of the WOS. The WOS has been the central source for citations to academic research since the 1970s (Garfield, 1987). Citations to articles include all those that were contained in the several thousand source journals in the WOS. Citations were gathered during January 2014.
METHODS
Predictor Variables A search for articles on suicide bereavement was undertaken by two of the authors with a background in bereavement research. Four source journals were searched: Suicide and Life-Threatening Behavior (SLTB), Crisis, Archives of Suicide Research, and Suicidology Online. The online database of each journal was searched with the search words: “bereavement,” “postvention,” and “survivor.” In addition, the tables of contents were hand searched to exclude articles from the original search that were not on suicide bereavement or postvention, and to include articles that were not identified by the original search. All published papers were included in the study, whereas announcements, errata, short comments on previously published articles, in memoria, and book reviews were excluded. Psychological autopsy studies were included when the study specifically addressed survivor’s bereavement issues in this type of research. Psychological autopsy studies that focused on and discussed merely the detection of risk factors were not included in the study. The years of the journals to be searched had to be matched with the availability of citation data in the source for citations: Thomson-Reuter’s the Web of Science (WOS). WOS citation coverage was available for SLTB from 1976 to 2013, for Crisis from
Predictor variables are categorized into three groups: characteristics of the article, the author, and the journal. A distinction is made between multiauthored articles and sole-authored articles. There is some evidence that the former receive more citations than the latter (Holden et al., 2005). Articles with two or more authors have access to larger researcher networks than articles with but one author. All else being equal, the years since publication of an article should influence citations. Articles that have been in print longer would be expected to garner more cites than recently published works. Finally, following previous work, it is assumed that review articles will receive more citations than nonreview articles (Stack, 2012). Some research has indicated a gender bias, favoring males, in publishing and receiving citations (Xie & Shauman, 2003). We include a control for the gender of the lead author of each article. Following Stack (2012), we also include a control for the nationality of the lead author. A distinction is made between U.S. authors (1) versus authors of all other nations (0). Finally, a characteristic of the journal publishing the article is controlled. Following previous work (Callaham et al., 2002;
20
PREDICTORS
Garfield, 2006), prestige is captured by the 5-year impact factor (IF) for each journal: the average number of citations an article receives during the 5 years after its publication. IFs are taken from data in the WOS. Herein, a dichotomous variable is employed where 1 = article published in SLTB and 0 = article published in Crisis or Archives of Suicide Research. SLTB had the highest IF. As there were very few articles meeting the criteria for inclusion in Archives of Suicide Research, they were combined with the articles from Crisis. In results not fully reported here, the omission of these few articles from the analysis had no significant bearing on the results.
RESULTS
The 103 articles received a total of 1,495 citations. The range in citations went from 0 to 139. The average article received 14.51 citations. There was substantial variability in article characteristics. The average number of authors per article was 2.76, with a range from 1 to 10. Most articles were multiauthored (70.9%), with 29.1% being sole authored. Thirteen papers or 12.6% were review articles. Year of publication ranged from 1976 through 2013, and 52.41 percent of the lead authors were males. A majority of lead authors (54.4%) were from the United States and the remainder were from other nations. Quantitative Analysis The results of the multivariate analysis are presented in Table 1. Controlling for the other variables, several characteristics of the article predicted citation impact. Review articles, in particular, were predictive of citations received. From the regression coefficient, review articles received 28.8 more citations than nonreview articles. The coefficient is 5.8 times its standard error. In addition, as anticipated, multiauthored articles received more citations than their sole-authored counterparts. On aver-
OF
ARTICLE IMPACT
IN
SUICIDOLOGY
age, coauthored articles received 13.8 more citations than sole-authored ones. As anticipated, newer articles received fewer citations than older ones. For example, from the regression coefficient applied to the years 2000 and 2001, all else being equal, an article published in 2001 would receive .86 fewer citations than an article published in 2000. Journal prestige also influenced article impact. Articles in SLTB received more citations (7.2 more cites) than articles in the other journals. The findings on author characteristics were split. Male gender predicted citations, with articles with male lead authors receiving, on average, 6.5 more citations than those authored by females. However, articles with U.S. lead authors did not differ from non-U.S.-authored articles in citations received. From the standardized beta coefficients, the review article was the most important predictor of citations (b = .461), followed by year of publication (b = .414) and the multiauthored article (b = .302). While significant, the remaining predictors had less of an effect on citations: male gender (b = .156) and publication in SLTB (b = .161). The model provided a reasonably good fit for explaining citations, given the significant F-ratio and size of the r-squared statistic. From the latter, the model explains 40.8% of the variance in citations. Qualitative Analysis A list of the 12 most-cited articles is presented in Table 2. The structural variables from Table 1 help to provide an explanation for why these particular articles received a large number of citations. Four of the articles are reviews (e.g., McIntosh, 1993) and the majority (8/12) have several authors. However, while soleauthored articles were at a disadvantage for being cited relative to multiauthored articles, four sole-authored articles made it into the most-cited list. These included two by Reed (1993, 1998) and one each by McIntosh (1993) and Jordan (2001). In only
ANDRIESSEN
ET AL.
21
TABLE 1
The Effects of Article Characteristics, Author Characteristics, and Citing Journal on the Scholarly Impact (Citations) of 103 Articles on Suicide Bereavement Predictor variable (A) Article characteristics Multiauthor Year of publication Review article (B) Lead author characteristics Gender: Male North American (i.e., United States) (C) Journal Suicide and Life-Threatening Behavior Constant F statistic R-Squared
Unstandardized coefficient
Standard error
t statistic
Standardized coefficient
13.80* 0.86* 28.82*
4.22 0.22 4.98
3.26 3.91 5.788
.302 .414 .461
6.49* 0.56
3.32 3.83
1.95 .88
.156 .014
7.29* 1,730.86* 11.02* 0.408
4.37 444.86 – –
1.67 3.89 – –
.161 – – –
*p < .05
two cases, the first authors are females (Kolodny, Binder, Bronstein, & Friend, 1979; Seguin, Lesage, & Kiely, 1995). The most frequently cited paper (Beskow, Runeson, & Asg ard, 1990) reviews the methodology of psychological autopsy studies with emphasis on the ethical considerations in the work with suicide survivors in this type of research. The extent to which suicide bereavement is different from nonsuicide bereavement is the subject of four papers, including one review of the literature (Bailley, Kral, & Dunham, 1999; Jordan, 2001; McIntosh, 1993; Seguin et al., 1995). The study of psycho social variables affecting the process of suicide bereavement was the focus of four other studies (Reed, 1993, 1998; Reed & Greenwald, 1991), including a longitudinal follow-up study (Brent, Moritz, Bridge, Perper, & Canobbio, 1996). One paper (Kolodny et al., 1979) deals with the grief process of therapists who lost a patient by suicide, and emphasizes the need for training and support for surviving clinicians. An overview of peer and professional support available to suicide survivors, including a review of the effectiveness of such interventions is a subject of one paper
(Jordan & McMenamy, 2004). One paper (Crosby & Sacks, 2002) presents results of a community survey regarding the prevalence of exposure to suicide in a general U.S. community sample and at the estimated impact of suicide exposure on suicidal ideation and behavior of respondents.
CONCLUSIONS
Analyses of citations to articles are useful in efforts at taking stock of a field. Such analyses can present preliminary answers to such questions as which are the most cited articles; who are the highest cited scholars; and what new developments, reviews, and debates are apparently receiving the most attention by the community of researchers in the field. While there has been much relevant work in a wide variety of scientific fields (Callaham et al., 2002; Garfield, 1987; Hall, 1998; Holden et al., 2005; Key, 1988; Mehlman & Wenger, 2006; Ollerton & Sugrue, 2005; Seiber, 2010; Smith & Leggat, 2008; Stack, 2013; Yang & Pan, 2006), there have been only two descriptive investigations in suicidology (Cardinal, 2008; Stack, 2012). The present study first contributes to the literature
22
PREDICTORS
OF
ARTICLE IMPACT
IN
SUICIDOLOGY
TABLE 2
The 12 Most Cited Articles on Bereavement and Suicide by Citations Received, Author, Year Published, Title, Journal Cites
Lead author (year of publication)
36
McIntosh, John L. (1993)
36
Crosby, Alex E. (2002)
39
Kolodny, Susan (1979)
39
Reed, Mark D. (1993)
39
Reed, Mark D. (1998)
43
Reed, Mark D. (1991)
43
Bailey, Steve (1999)
48
Jordan, John R. (2004)
53
Seguin, Monique (1995)
54
Brent, David A. (1996)
108
Jordan, John R. (2001)
139
Beskow, Jan (1990)
Article Title, Journal, Volume, Page Range Control group studies of suicide survivors: A review and critique. Suicide and Life-Threatening Behavior, 23(2), 146–161 Exposure to suicide: Incidence and association with suicidal ideation and behavior: United States, 1994. Suicide and Life-Threatening Behavior, 32(3), 321–328 The working through of patients’ suicides by four therapists. Suicide and Life-Threatening Behavior, 9(1), 33–46 Sudden death and bereavement outcomes: The impact of resources on grief symptomatology and detachment. Suicide and Life-Threatening Behavior, 23(3), 204–220 Predicting grief symptomatology among the suddenly bereaved. Suicide and Life-Threatening Behavior, 28(3), 285–301 Survivor-victim status, attachment, and sudden death bereavement. Suicide and Life-Threatening Behavior, 21(4), 385–401 Survivors of suicide do grieve differently: Empirical support for a common sense proposition. Suicide and Life-Threatening Behavior, 29(3), 256–271 Interventions for suicide survivors: A review of the literature. Suicide and Life-Threatening Behavior, 34(4), 337–349 Parental bereavement after suicide and accident: A comparative study. Suicide and Life-Threatening Behavior, 25(4), 489–498 The impact of adolescent suicide on siblings and parents: A longitudinal follow-up. Suicide and Life-Threatening Behavior, 26 (3), 253–259 Is suicide bereavement different? A reassessment of the literature. Suicide and Life-Threatening Behavior, 31(1), 91–102 Psychological autopsies: Methods and ethics. Suicide and Life-Threatening Behavior, 20(4), 307–320
through a hypothesis testing, quantitative analysis of a subfield of suicide studies, bereavement. A multiple regression analysis was able to explain 41% of the variance in citations received among 103 bereavement articles. The most important predictors were the structural characteristics of the articles themselves. Review articles received relatively more citations than other articles, a finding consistent with many of the previous studies limited to citation classics (e.g., Stack, 2012, 2013). Years since publication and the multiauthored article type were also predictive of citations irrespective of their qualitative content.
The qualitative analysis added additional considerations to explaining why some articles have more usefulness or impact than other articles to the research community. The highest number of citations reported for the review paper on methodology of psychological autopsy studies and the ethical considerations regarding involvement of survivors in this type of research (Beskow et al., 1990) seems not surprising in the context of the continuing popularity of this research method (e.g., Pouliot & De Leo, 2006) as well as the recent discussion regarding its scientific soundness (Hjelmeland, Dieserud, Dyregrov, Knizek, & Leenaars, 2012).
ANDRIESSEN
ET AL.
While coauthored articles have a much greater chance than sole-authored ones to have a high impact, nevertheless, certain sole-authored ones do have a substantial impact. This seems to be the case especially for high quality reviews of the literature, such as papers by Jordan (2001) and McIntosh (1993), which not only summarize the evidence-based knowledge, but also provide a critical review of the methodology and recommendations for future studies. The extent to which suicide bereavement is different from bereavement after other modes of death and the study of psychological and social factors affecting the process and the outcome of suicide bereavement may have appeal to a larger audience of scholars of generalized bereavement
23 processes, resulting in more citations. In addition, the question whether suicide survivors “do grieve differently” remains a challenge and inspires new studies and theoretical discussions (Andriessen & Krysinska, 2012; Jordan & McIntosh, 2011). Of interest, almost all studies looking at the process and outcome of grief after suicide were based on samples of close (i.e., firstdegree) relatives of suicide victims. These findings are in line with a recent review highlighting the lack of studies with other groups of survivors or with general population samples (Andriessen, 2014). Future research is needed to see whether the patterns uncovered in the bereavement literature will be replicated for other subfields in suicidology such as sociology and clinical research.
REFERENCES ANDRIESSEN, K. (2014). Suicide bereavement and postvention in major suicidology journals: Lessons learned for the future of postvention. Crisis, 35, 338–348. ANDRIESSEN, K., & KRYSINSKA, K. (2012). Essential questions on suicide bereavement and postvention. International Journal of Environmental Research and Public Health, 9, 24–32. BAILLEY, S. E., KRAL, M. J., & DUNHAM, K. (1999). Survivors of suicide do grieve differently: Empirical support for a common sense proposition. Suicide and Life-Threatening Behavior, 29, 256–271. BESKOW, J., RUNESON, B., & A SG ARD, U. (1990). Psychological autopsies: Methods and ethics. Suicide and Life-Threatening Behavior, 20, 307–320. BRENT, D. A., MORITZ, G., BRIDGE, J., PERPER, J., & CANOBBIO, R. (1996). The impact of adolescent suicide on siblings and parents: A longitudinal follow-up. Suicide and Life-Threatening Behavior, 26, 253–259. CALLAHAM, M., WEARS, R. L., & WEBER, E. (2002). Journal prestige, publication bias, and other characteristics associated with citation of published articles. Journal of the American Medical Association, 287, 2847–2850. CARDINAL, C. (2008). Three decades of Suicide and Life-Threatening Behavior: A bibliometric study. Suicide and Life-Threatening Behavior, 38, 260–273. CROSBY, A. E., & SACKS, J. J. (2002). Exposure to suicide: Incidence and association
with suicidal ideation and behavior: United States, 1994. Suicide and Life-Threatening Behavior, 32, 321–328. GARFIELD, E. (1987). 100 citation classics from the Journal of the American Medical Association. Journal of the American Medical Association, 257, 52–59. GARFIELD, E. (2006). The history and meaning of the journal impact factor. Journal of the American Medical Association, 295, 90–93. HALL, G. M. (1998). BJA citation classics 1945–1992. British Journal of Anesthesia, 80, 4–6. HJELMELAND, H., DIESERUD, G., DYREGROV, K., KNIZEK, B. L., & LEENAARS, A. A. (2012). Psychological autopsy studies as diagnostic tools: Are they methodologically flawed? Death Studies, 36, 605–626. HOLDEN, G., ROSENBERG, G., & BARKER, K. (2005). Bibliometrics: A potential decision making aid in hiring, reappointment, tenure, and promotion decisions. Social Work in Health Care, 41, 67–92. JORDAN, J. R. (2001). Is suicide bereavement different? A reassessment of the literature. Suicide and Life-Threatening Behavior, 31, 91–102. JORDAN, J. R., & MCINTOSH, J. L. (2011). Is suicide bereavement different? A framework for rethinking the question. In J. R. Jordan & J. L. McIntosh (Eds.), Grief after suicide (pp. 19– 42). New York: Routledge. JORDAN, J. R., & MCMENAMY, J. (2004). Interventions for suicide survivors: A review of the
24 literature. Suicide and Life-Threatening Behavior, 34, 337–349. KEY, J. D. (1988). Citation classics: Most cited articles from the archives of PM&R. Archives of Physical Medicine & Rehabilitation, 69, 1058–1059. KOLODNY, S., BINDER, R. L., BRONSTEIN, A. A., & FRIEND, R. L. (1979). The working through of patients’ suicides by four therapists. Suicide and Life-Threatening Behavior, 9, 33–46. MCINTOSH, J. L. (1993). Control group studies of suicide survivors: A review and critique. Suicide and Life-Threatening Behavior, 23, 146–161. MEHLMAN, C. T., & WENGER, D. R. (2006). The top 25 at 25: Citation classics in the Journal of Pediatric Orthopedics. Journal of Pediatric Orthopedics, 26, 691–694. OLLERTON, J. E., & SUGRUE, M. (2005). Citation classics in Trauma. The Journal of Trauma, 58, 364–369. POULIOT, L., & DE LEO, D. (2006). Critical issues in psychological autopsy studies. Suicide and Life-Threatening Behavior, 36, 491–510. REED, M. D. (1993). Sudden death and bereavement outcomes: The impact of resources on grief symptomatology and detachment. Suicide and Life-Threatening Behavior, 23, 204–220. REED, M. D. (1998). Predicting grief symptomatology among the suddenly bereaved. Suicide and Life-Threatening Behavior, 28, 285– 301.
PREDICTORS
OF
ARTICLE IMPACT
IN
SUICIDOLOGY
REED, M. D., & GREENWALD, J. Y. (1991). Survivor-victim status, attachment, and sudden death bereavement. Suicide and Life-Threatening Behavior, 21, 385–401. SEGUIN , M., LESAGE, A., & KIELY, M. (1995). Parental bereavement after suicide and accident: A comparative study. Suicide and LifeThreatening Behavior, 25, 489–498. SEIBER, J. N. (2010). Citation classics and classic citations in JAFC. Journal of Agricultural and Food Chemistry, 58, 1–3. SMITH, D. R., & LEGGAT, P. A. (2008). Ten citation classics from the Australian and New Zealand Journal of Public Health. Australian and New Zealand Journal of Public Health, 32, 105– 106. STACK, S. (2012). Citation classics in Suicide and Life-Threatening Behavior: A research note. Suicide and Life-Threatening Behavior, 42, 628–639. STACK, S. (2013). Citation classics in Deviant Behavior: A research note. Deviant Behavior, 34, 85–96. XIE, Y., & SHAUMAN, K. (2003). Women in science. Cambridge Harvard University Press. YANG, H., & PAN, B. (2006). Citation classics in Fertility & Sterility, 1975–2004. Fertility and Sterility, 86, 795–797. Manuscript Received: February 25, 2014 Revision Accepted: March 6, 2014
Copyright of Suicide & Life-Threatening Behavior is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Suicide and Life-Threatening Behavior 45 (1) February 2015 © 2014 The American Association of Suicidology DOI: 10.1111/sltb.12106
Predictors of Article Impact in Suicidology: The Bereavement Literature, A Research Note KARL ANDRIESSEN, MSUICIDOLOGY, KAROLINA KRYSINSKA, PHD, AND STEVEN STACK, PHD
Citation analysis has been neglected in suicidology. The present note applies a mixed-methods approach to both test and suggest hypotheses for the variation in article impact in the bereavement literature. One hundred three articles from three core suicidology journals met the criteria for inclusion in the investigation. Citations to the articles were obtained from the Web of Science. Predictor variables included structural characteristics of the author (e.g., gender) and the article itself (e.g., years since publication). A multivariate regression analysis determined that, controlling for the other variables, the most important predictor of citations was the review article (b = .461), followed by year of publication (b = .414), the multiauthored article (b = .302), publication in Suicide and Life-Threatening Behavior (SLTB) (b = .161), and male gender (b = .156). The 12 most cited articles were published between 1979 and 2004 in SLTB. The majority of these papers was written by males, were U.S. authors, and had more than one author. Four of the most cited articles were reviews. The study concludes that structural characteristics of articles and authors explained 41% of the variance in citations. The qualitative analysis determined that review papers, and papers on characteristics of suicide bereavement and psychological autopsies have been most frequently cited. Replication studies are needed for other subfields of suicidology.
Citation analysis helps to track the leading topics and the researchers who are having the greatest impact on a field. Citation analysis has blossomed in many scientific fields including anesthesiology, chemistry, fertility and sterility, deviant behavior, general medicine, pediatrics, public health, rehabilitation, KARL ANDRIESSEN, Faculty of Psychology and Educational Sciences, KU Leuven – University of Leuven, Leuven, Belgium; KAROLINA KRYSINSKA, Clinical Psychology, KU Leuven – University of Leuven, Leuven, Belgium; STEVEN STACK, Department of Psychiatry & Department of Criminology, Wayne State University, Detroit, MI, USA. Address correspondence to Karl Andriessen, Faculty of Psychology and Educational Sciences, KU Leuven – University of Leuven, Vanderkelenstraat 32, 3000 Leuven, Belgium; E-mail: [email protected]
and trauma research (Callaham, Wears, & Weber, 2002; Garfield, 1987; Hall, 1998; Holden, Rosenberg, & Barker, 2005; Key, 1988; Mehlman & Wenger, 2006; Ollerton & Sugrue, 2005; Seiber, 2010; Smith & Leggat, 2008; Stack, 2013; Yang & Pan, 2006). While there is a growing literature on such analyses, there is scant attention paid to these issues in suicidology. Only two article impact studies are available in suicide studies. Both of these are mainly descriptive works that provide lists of the most cited articles in Suicide and Life-Threatening Behavior (Cardinal, 2008; Stack, 2012). The present note contributes to the literature by applying a mixed-methods approach to citation analysis in three core suicidology journals. It goes beyond descriptive work in the previous studies
ANDRIESSEN
ET AL.
(Cardinal, 2008; Stack, 2012) by testing a series of structural hypotheses to explain the variation in citation impact in a sample of articles on bereavement and suicide, including psychological autopsy studies specifically discussing survivor issues in this type of research (Andriessen, 2014). These hypotheses involve structural characteristics of the article (e.g., year of publication, number of authors), characteristics of the author (e.g., gender, nationality), and the journal (e.g., impact factor). Second, it provides a qualitative analysis of the articles to assist in further explaining the variation in article impact on the field.
19 2005 to 2013, and for Archives of Suicide Research from 2006 to 2013. Suicidology Online was not cited in WOS. The sample contains 103 articles published during 1976 through 2013. Citations Citations to each of the bereavement articles were obtained from the online version of the WOS. The WOS has been the central source for citations to academic research since the 1970s (Garfield, 1987). Citations to articles include all those that were contained in the several thousand source journals in the WOS. Citations were gathered during January 2014.
METHODS
Predictor Variables A search for articles on suicide bereavement was undertaken by two of the authors with a background in bereavement research. Four source journals were searched: Suicide and Life-Threatening Behavior (SLTB), Crisis, Archives of Suicide Research, and Suicidology Online. The online database of each journal was searched with the search words: “bereavement,” “postvention,” and “survivor.” In addition, the tables of contents were hand searched to exclude articles from the original search that were not on suicide bereavement or postvention, and to include articles that were not identified by the original search. All published papers were included in the study, whereas announcements, errata, short comments on previously published articles, in memoria, and book reviews were excluded. Psychological autopsy studies were included when the study specifically addressed survivor’s bereavement issues in this type of research. Psychological autopsy studies that focused on and discussed merely the detection of risk factors were not included in the study. The years of the journals to be searched had to be matched with the availability of citation data in the source for citations: Thomson-Reuter’s the Web of Science (WOS). WOS citation coverage was available for SLTB from 1976 to 2013, for Crisis from
Predictor variables are categorized into three groups: characteristics of the article, the author, and the journal. A distinction is made between multiauthored articles and sole-authored articles. There is some evidence that the former receive more citations than the latter (Holden et al., 2005). Articles with two or more authors have access to larger researcher networks than articles with but one author. All else being equal, the years since publication of an article should influence citations. Articles that have been in print longer would be expected to garner more cites than recently published works. Finally, following previous work, it is assumed that review articles will receive more citations than nonreview articles (Stack, 2012). Some research has indicated a gender bias, favoring males, in publishing and receiving citations (Xie & Shauman, 2003). We include a control for the gender of the lead author of each article. Following Stack (2012), we also include a control for the nationality of the lead author. A distinction is made between U.S. authors (1) versus authors of all other nations (0). Finally, a characteristic of the journal publishing the article is controlled. Following previous work (Callaham et al., 2002;
20
PREDICTORS
Garfield, 2006), prestige is captured by the 5-year impact factor (IF) for each journal: the average number of citations an article receives during the 5 years after its publication. IFs are taken from data in the WOS. Herein, a dichotomous variable is employed where 1 = article published in SLTB and 0 = article published in Crisis or Archives of Suicide Research. SLTB had the highest IF. As there were very few articles meeting the criteria for inclusion in Archives of Suicide Research, they were combined with the articles from Crisis. In results not fully reported here, the omission of these few articles from the analysis had no significant bearing on the results.
RESULTS
The 103 articles received a total of 1,495 citations. The range in citations went from 0 to 139. The average article received 14.51 citations. There was substantial variability in article characteristics. The average number of authors per article was 2.76, with a range from 1 to 10. Most articles were multiauthored (70.9%), with 29.1% being sole authored. Thirteen papers or 12.6% were review articles. Year of publication ranged from 1976 through 2013, and 52.41 percent of the lead authors were males. A majority of lead authors (54.4%) were from the United States and the remainder were from other nations. Quantitative Analysis The results of the multivariate analysis are presented in Table 1. Controlling for the other variables, several characteristics of the article predicted citation impact. Review articles, in particular, were predictive of citations received. From the regression coefficient, review articles received 28.8 more citations than nonreview articles. The coefficient is 5.8 times its standard error. In addition, as anticipated, multiauthored articles received more citations than their sole-authored counterparts. On aver-
OF
ARTICLE IMPACT
IN
SUICIDOLOGY
age, coauthored articles received 13.8 more citations than sole-authored ones. As anticipated, newer articles received fewer citations than older ones. For example, from the regression coefficient applied to the years 2000 and 2001, all else being equal, an article published in 2001 would receive .86 fewer citations than an article published in 2000. Journal prestige also influenced article impact. Articles in SLTB received more citations (7.2 more cites) than articles in the other journals. The findings on author characteristics were split. Male gender predicted citations, with articles with male lead authors receiving, on average, 6.5 more citations than those authored by females. However, articles with U.S. lead authors did not differ from non-U.S.-authored articles in citations received. From the standardized beta coefficients, the review article was the most important predictor of citations (b = .461), followed by year of publication (b = .414) and the multiauthored article (b = .302). While significant, the remaining predictors had less of an effect on citations: male gender (b = .156) and publication in SLTB (b = .161). The model provided a reasonably good fit for explaining citations, given the significant F-ratio and size of the r-squared statistic. From the latter, the model explains 40.8% of the variance in citations. Qualitative Analysis A list of the 12 most-cited articles is presented in Table 2. The structural variables from Table 1 help to provide an explanation for why these particular articles received a large number of citations. Four of the articles are reviews (e.g., McIntosh, 1993) and the majority (8/12) have several authors. However, while soleauthored articles were at a disadvantage for being cited relative to multiauthored articles, four sole-authored articles made it into the most-cited list. These included two by Reed (1993, 1998) and one each by McIntosh (1993) and Jordan (2001). In only
ANDRIESSEN
ET AL.
21
TABLE 1
The Effects of Article Characteristics, Author Characteristics, and Citing Journal on the Scholarly Impact (Citations) of 103 Articles on Suicide Bereavement Predictor variable (A) Article characteristics Multiauthor Year of publication Review article (B) Lead author characteristics Gender: Male North American (i.e., United States) (C) Journal Suicide and Life-Threatening Behavior Constant F statistic R-Squared
Unstandardized coefficient
Standard error
t statistic
Standardized coefficient
13.80* 0.86* 28.82*
4.22 0.22 4.98
3.26 3.91 5.788
.302 .414 .461
6.49* 0.56
3.32 3.83
1.95 .88
.156 .014
7.29* 1,730.86* 11.02* 0.408
4.37 444.86 – –
1.67 3.89 – –
.161 – – –
*p < .05
two cases, the first authors are females (Kolodny, Binder, Bronstein, & Friend, 1979; Seguin, Lesage, & Kiely, 1995). The most frequently cited paper (Beskow, Runeson, & Asg ard, 1990) reviews the methodology of psychological autopsy studies with emphasis on the ethical considerations in the work with suicide survivors in this type of research. The extent to which suicide bereavement is different from nonsuicide bereavement is the subject of four papers, including one review of the literature (Bailley, Kral, & Dunham, 1999; Jordan, 2001; McIntosh, 1993; Seguin et al., 1995). The study of psycho social variables affecting the process of suicide bereavement was the focus of four other studies (Reed, 1993, 1998; Reed & Greenwald, 1991), including a longitudinal follow-up study (Brent, Moritz, Bridge, Perper, & Canobbio, 1996). One paper (Kolodny et al., 1979) deals with the grief process of therapists who lost a patient by suicide, and emphasizes the need for training and support for surviving clinicians. An overview of peer and professional support available to suicide survivors, including a review of the effectiveness of such interventions is a subject of one paper
(Jordan & McMenamy, 2004). One paper (Crosby & Sacks, 2002) presents results of a community survey regarding the prevalence of exposure to suicide in a general U.S. community sample and at the estimated impact of suicide exposure on suicidal ideation and behavior of respondents.
CONCLUSIONS
Analyses of citations to articles are useful in efforts at taking stock of a field. Such analyses can present preliminary answers to such questions as which are the most cited articles; who are the highest cited scholars; and what new developments, reviews, and debates are apparently receiving the most attention by the community of researchers in the field. While there has been much relevant work in a wide variety of scientific fields (Callaham et al., 2002; Garfield, 1987; Hall, 1998; Holden et al., 2005; Key, 1988; Mehlman & Wenger, 2006; Ollerton & Sugrue, 2005; Seiber, 2010; Smith & Leggat, 2008; Stack, 2013; Yang & Pan, 2006), there have been only two descriptive investigations in suicidology (Cardinal, 2008; Stack, 2012). The present study first contributes to the literature
22
PREDICTORS
OF
ARTICLE IMPACT
IN
SUICIDOLOGY
TABLE 2
The 12 Most Cited Articles on Bereavement and Suicide by Citations Received, Author, Year Published, Title, Journal Cites
Lead author (year of publication)
36
McIntosh, John L. (1993)
36
Crosby, Alex E. (2002)
39
Kolodny, Susan (1979)
39
Reed, Mark D. (1993)
39
Reed, Mark D. (1998)
43
Reed, Mark D. (1991)
43
Bailey, Steve (1999)
48
Jordan, John R. (2004)
53
Seguin, Monique (1995)
54
Brent, David A. (1996)
108
Jordan, John R. (2001)
139
Beskow, Jan (1990)
Article Title, Journal, Volume, Page Range Control group studies of suicide survivors: A review and critique. Suicide and Life-Threatening Behavior, 23(2), 146–161 Exposure to suicide: Incidence and association with suicidal ideation and behavior: United States, 1994. Suicide and Life-Threatening Behavior, 32(3), 321–328 The working through of patients’ suicides by four therapists. Suicide and Life-Threatening Behavior, 9(1), 33–46 Sudden death and bereavement outcomes: The impact of resources on grief symptomatology and detachment. Suicide and Life-Threatening Behavior, 23(3), 204–220 Predicting grief symptomatology among the suddenly bereaved. Suicide and Life-Threatening Behavior, 28(3), 285–301 Survivor-victim status, attachment, and sudden death bereavement. Suicide and Life-Threatening Behavior, 21(4), 385–401 Survivors of suicide do grieve differently: Empirical support for a common sense proposition. Suicide and Life-Threatening Behavior, 29(3), 256–271 Interventions for suicide survivors: A review of the literature. Suicide and Life-Threatening Behavior, 34(4), 337–349 Parental bereavement after suicide and accident: A comparative study. Suicide and Life-Threatening Behavior, 25(4), 489–498 The impact of adolescent suicide on siblings and parents: A longitudinal follow-up. Suicide and Life-Threatening Behavior, 26 (3), 253–259 Is suicide bereavement different? A reassessment of the literature. Suicide and Life-Threatening Behavior, 31(1), 91–102 Psychological autopsies: Methods and ethics. Suicide and Life-Threatening Behavior, 20(4), 307–320
through a hypothesis testing, quantitative analysis of a subfield of suicide studies, bereavement. A multiple regression analysis was able to explain 41% of the variance in citations received among 103 bereavement articles. The most important predictors were the structural characteristics of the articles themselves. Review articles received relatively more citations than other articles, a finding consistent with many of the previous studies limited to citation classics (e.g., Stack, 2012, 2013). Years since publication and the multiauthored article type were also predictive of citations irrespective of their qualitative content.
The qualitative analysis added additional considerations to explaining why some articles have more usefulness or impact than other articles to the research community. The highest number of citations reported for the review paper on methodology of psychological autopsy studies and the ethical considerations regarding involvement of survivors in this type of research (Beskow et al., 1990) seems not surprising in the context of the continuing popularity of this research method (e.g., Pouliot & De Leo, 2006) as well as the recent discussion regarding its scientific soundness (Hjelmeland, Dieserud, Dyregrov, Knizek, & Leenaars, 2012).
ANDRIESSEN
ET AL.
While coauthored articles have a much greater chance than sole-authored ones to have a high impact, nevertheless, certain sole-authored ones do have a substantial impact. This seems to be the case especially for high quality reviews of the literature, such as papers by Jordan (2001) and McIntosh (1993), which not only summarize the evidence-based knowledge, but also provide a critical review of the methodology and recommendations for future studies. The extent to which suicide bereavement is different from bereavement after other modes of death and the study of psychological and social factors affecting the process and the outcome of suicide bereavement may have appeal to a larger audience of scholars of generalized bereavement
23 processes, resulting in more citations. In addition, the question whether suicide survivors “do grieve differently” remains a challenge and inspires new studies and theoretical discussions (Andriessen & Krysinska, 2012; Jordan & McIntosh, 2011). Of interest, almost all studies looking at the process and outcome of grief after suicide were based on samples of close (i.e., firstdegree) relatives of suicide victims. These findings are in line with a recent review highlighting the lack of studies with other groups of survivors or with general population samples (Andriessen, 2014). Future research is needed to see whether the patterns uncovered in the bereavement literature will be replicated for other subfields in suicidology such as sociology and clinical research.
REFERENCES ANDRIESSEN, K. (2014). Suicide bereavement and postvention in major suicidology journals: Lessons learned for the future of postvention. Crisis, 35, 338–348. ANDRIESSEN, K., & KRYSINSKA, K. (2012). Essential questions on suicide bereavement and postvention. International Journal of Environmental Research and Public Health, 9, 24–32. BAILLEY, S. E., KRAL, M. J., & DUNHAM, K. (1999). Survivors of suicide do grieve differently: Empirical support for a common sense proposition. Suicide and Life-Threatening Behavior, 29, 256–271. BESKOW, J., RUNESON, B., & A SG ARD, U. (1990). Psychological autopsies: Methods and ethics. Suicide and Life-Threatening Behavior, 20, 307–320. BRENT, D. A., MORITZ, G., BRIDGE, J., PERPER, J., & CANOBBIO, R. (1996). The impact of adolescent suicide on siblings and parents: A longitudinal follow-up. Suicide and Life-Threatening Behavior, 26, 253–259. CALLAHAM, M., WEARS, R. L., & WEBER, E. (2002). Journal prestige, publication bias, and other characteristics associated with citation of published articles. Journal of the American Medical Association, 287, 2847–2850. CARDINAL, C. (2008). Three decades of Suicide and Life-Threatening Behavior: A bibliometric study. Suicide and Life-Threatening Behavior, 38, 260–273. CROSBY, A. E., & SACKS, J. J. (2002). Exposure to suicide: Incidence and association
with suicidal ideation and behavior: United States, 1994. Suicide and Life-Threatening Behavior, 32, 321–328. GARFIELD, E. (1987). 100 citation classics from the Journal of the American Medical Association. Journal of the American Medical Association, 257, 52–59. GARFIELD, E. (2006). The history and meaning of the journal impact factor. Journal of the American Medical Association, 295, 90–93. HALL, G. M. (1998). BJA citation classics 1945–1992. British Journal of Anesthesia, 80, 4–6. HJELMELAND, H., DIESERUD, G., DYREGROV, K., KNIZEK, B. L., & LEENAARS, A. A. (2012). Psychological autopsy studies as diagnostic tools: Are they methodologically flawed? Death Studies, 36, 605–626. HOLDEN, G., ROSENBERG, G., & BARKER, K. (2005). Bibliometrics: A potential decision making aid in hiring, reappointment, tenure, and promotion decisions. Social Work in Health Care, 41, 67–92. JORDAN, J. R. (2001). Is suicide bereavement different? A reassessment of the literature. Suicide and Life-Threatening Behavior, 31, 91–102. JORDAN, J. R., & MCINTOSH, J. L. (2011). Is suicide bereavement different? A framework for rethinking the question. In J. R. Jordan & J. L. McIntosh (Eds.), Grief after suicide (pp. 19– 42). New York: Routledge. JORDAN, J. R., & MCMENAMY, J. (2004). Interventions for suicide survivors: A review of the
24 literature. Suicide and Life-Threatening Behavior, 34, 337–349. KEY, J. D. (1988). Citation classics: Most cited articles from the archives of PM&R. Archives of Physical Medicine & Rehabilitation, 69, 1058–1059. KOLODNY, S., BINDER, R. L., BRONSTEIN, A. A., & FRIEND, R. L. (1979). The working through of patients’ suicides by four therapists. Suicide and Life-Threatening Behavior, 9, 33–46. MCINTOSH, J. L. (1993). Control group studies of suicide survivors: A review and critique. Suicide and Life-Threatening Behavior, 23, 146–161. MEHLMAN, C. T., & WENGER, D. R. (2006). The top 25 at 25: Citation classics in the Journal of Pediatric Orthopedics. Journal of Pediatric Orthopedics, 26, 691–694. OLLERTON, J. E., & SUGRUE, M. (2005). Citation classics in Trauma. The Journal of Trauma, 58, 364–369. POULIOT, L., & DE LEO, D. (2006). Critical issues in psychological autopsy studies. Suicide and Life-Threatening Behavior, 36, 491–510. REED, M. D. (1993). Sudden death and bereavement outcomes: The impact of resources on grief symptomatology and detachment. Suicide and Life-Threatening Behavior, 23, 204–220. REED, M. D. (1998). Predicting grief symptomatology among the suddenly bereaved. Suicide and Life-Threatening Behavior, 28, 285– 301.
PREDICTORS
OF
ARTICLE IMPACT
IN
SUICIDOLOGY
REED, M. D., & GREENWALD, J. Y. (1991). Survivor-victim status, attachment, and sudden death bereavement. Suicide and Life-Threatening Behavior, 21, 385–401. SEGUIN , M., LESAGE, A., & KIELY, M. (1995). Parental bereavement after suicide and accident: A comparative study. Suicide and LifeThreatening Behavior, 25, 489–498. SEIBER, J. N. (2010). Citation classics and classic citations in JAFC. Journal of Agricultural and Food Chemistry, 58, 1–3. SMITH, D. R., & LEGGAT, P. A. (2008). Ten citation classics from the Australian and New Zealand Journal of Public Health. Australian and New Zealand Journal of Public Health, 32, 105– 106. STACK, S. (2012). Citation classics in Suicide and Life-Threatening Behavior: A research note. Suicide and Life-Threatening Behavior, 42, 628–639. STACK, S. (2013). Citation classics in Deviant Behavior: A research note. Deviant Behavior, 34, 85–96. XIE, Y., & SHAUMAN, K. (2003). Women in science. Cambridge Harvard University Press. YANG, H., & PAN, B. (2006). Citation classics in Fertility & Sterility, 1975–2004. Fertility and Sterility, 86, 795–797. Manuscript Received: February 25, 2014 Revision Accepted: March 6, 2014
Copyright of Suicide & Life-Threatening Behavior is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
