Exploring Retraction in Tourism and Hospitality Journals

Recently, retraction has received considerable attention in several fields of research. The topic, however, remains neglected in the field of tourism and hospitality. The current investigation explores the current level of retraction in tourism and hospitality research. It also draws on the perceptions of the editors-in-chief of tourism and hospitality journals to understand their attitudes toward retraction. The findings reveal that retraction is quite rare in the field of tourism and hospitality (only 5 retracted articles were identified). Empirical evidence shows that editors-in-chief have mixed perceptions concerning the current level of retraction, varying between optimism and pessimism. Optimistic editors-in-chief agreed that the rareness of retraction is reflective of research integrity and identified awareness of the best practices in academic publishing, the youthfulness of tourism as an area of study, and the role of plagiarism detection software programs as significant factors. Pessimistic editors-in-chief were reluctant to consider the low retraction rate to be an indicator of a high level of research integrity and they highlighted the difficulty of detecting malpractice, editors-in-chief’s unwillingness to engage in retraction, and the existence of unethical practices that do not lead to retraction. By identifying the retraction level and exploring the perceptions of editors-in-chief, the present study endeavours to track the progress, transparency, and integrity of tourism and hospitality research.


Introduction
Retraction is a common practice undertaken by editors-in-chief and publishers as a response to academic misconduct (e.g. plagiarism, duplicate publication) or honest errors (e.g. unintentional miscalculation of the data) with the aim of ensuring the integrity of the scientific literature. Defined as "the official declaration that a paper is so flawed that it must be withdrawn from the literature" (Van Noorden, 2011: 26), retraction is more associated with misconduct than honest errors. Over the past decade, a steady increase in the retraction rate was observed in different fields (Steen, 2011;Chen, Hu, Milbank, & Schultz, 2013;Karabag & Berggren, 2016;Moylan & Kowalczuk, 2016). While this may reflect changes in the behavior of both authors and institutions (Steen, Casadevall, & Fang, 2013); the increasing competition in the academia, the augmented need to publish, and the technological advancements that facilitated both detecting academic misconduct (e.g. plagiarism software programs) and communicating it may be among the key factors explaining the growth of retraction rate. Empirical evidence reveals that retracted authors may be affiliated to institutions in both developed and developing countries (Stretton et al., 2012;Nogueira, Gonçalves, Leles, Batista, & Costa, 2017;Fang, Steen, & Casadevall, 2012) which indicates that research misconduct is not restricted to specific geographical distribution but rather a universal issue (Resnik, Rasmussen, & Kissling, 2015).
Retraction is argued to be more common in applied sciences such as medicine and life science (Grieneisen & Zhang, 2012;Lu, Jin, Uzzi, & Jones, 2013;Hesselmann et al., 2017;Shuai, Rollins, Moulinier, Custis, Edmunds, & Schilder, 2017). Yet it has recently received considerable attention in various fields of research including economics (Cox, Craig, & Tourish, 2018), business and management (Tourish & Craig, 2018), library and information science (Ajiferuke & Adekannbi, 2018), and leadership (Atwater, Mumford, Schriesheim, & Yammarino, 2014). While journals in some fıelds of research (e.g. medicine) are active in retracting papers, many are not, specifically in the social science realm (Karabag & Berggren, 2012). To the authors' best knowledge, no study was conducted to investigate the issue of retraction in the tourism and hospitality research. This perhaps is due to the relative maturing nature of tourism research (de Esteban, Cetin, & Antonovica, 2015) and the inherently limited number of journals in the field. Therefore, the present study aims to fill this gap using a mixed-method design combining data from retracted articles with data from a survey of editors-in-chief of tourism and hospitality journals.
Accordingly, the contribution of the current study is twofold. First, it investigates the current level of retraction in the field of tourism and hospitality. This is important to track the progress and quality of research on the one hand and to obtain insights into the integrity and transparency in the field on the other hand. Second, the study gives voice to the editors-inchief of tourism and hospitality journals to understand their perceptions with respect to retraction. This is particularly important to explore as to whether the current level of retraction reflects the actual level of academic malpractice (Cox et al., 2018). By doing so, the present study adds to a growing body of literature on publishing ethics in tourism and hospitality research.

Literature Review
Academic misconduct is a major issue that erodes public trust in research. The increasing pressure for publishing and the lack of awareness of ethical practices are suggested as key reasons for involving in academic misconduct (Alrawadieh, 2018;Campos-Varela & Ruano-Raviña, 2018;McKercher, 2018;Redman & Merz, 2005). Metrification of research not only impeded the creation of new knowledge (Fennell, 2013), but it also led to the authors engaging in system manipulation to survive in a competitive academic environment (McKercher 2018b). Therefore, publishing ethics have received considerable attention over the past few years. As a response to research misconduct or honest errors, article retraction, which consists of withdrawing the article and publishing a retraction notice, emerges as a way to ensure research integrity. In this vein, retraction is one of the major governance mechanisms to identify and signal false knowledge (Furman, Jensen, & Murray, 2012).
Retraction is often perceived as a response to academic misconduct. Campos-Varela and Ruano-Raviña (2018) suggested that misconduct was the main cause of retraction of over 65% of the retracted papers in their sample. In keeping with this, Moylan and Kowalczuk (2016) found that only a small portion of retracted articles (13%) was due to honest error. Potential reasons for retraction include plagiarism, duplicate publication, questionable reviewing practices, financial disclosure violations, data falsification, data fabrication, and authorship violation (Cox et al., 2018;Steen, 2011;Karabag & Berggren, 2012;Hopp & Hoover, 2017;Bar-Ilan & Halevi, 2018;King et al., 2018). For instance, Cox et al (2018), and Moylan and Kowalczuk (2016) found that fake peer review and plagiarism were the key reasons for retraction from journals in economics. King et al (2018) found that over one-third of retracted articles were attributed to duplication. Yet, honest errors by authors can also be a potential reason for retraction (Van Noorden, 2011). For instance, in his study, Steen (2011) suggested that about one-third (31.5%) of the articles in his sample were retracted for scientific mistakes. Compared with articles retracted because of misconduct, articles with honest errors are more likely to be retracted by the author(s) of the article (Nath, Marcus, & Druss, 2006). Overall, retraction is an invitation for readers to ignore a given article (Sox & Rennie, 2006) which may lead readers to ignore the author's entire academic work. Therefore, including the reason for retracting a given paper in the retraction notice is of significant importance to distinguish offenders from honest authors. In this vein, Sox and Rennie (2006) proposed that editors-in-chief should use the word "retraction" exclusively in cases of fraud. Some authors (e.g. Wager & Williams, 2011) noted that some retractions fail to state the reason underlying the retraction decision and thus fail to distinguish honest error from intentional misconduct.
Regardless of whether it is an intentional research misconduct or honest error, retraction remains a source of harm to the reputation of different stakeholders in the scientific community including authors and their institutions, and to a lesser extent, editors-inchief and their journals. In particular, retraction constitutes a significant threat to the retracted authors' reputation which eventually harms their academic career. As noted by Steen (2011: 252), "retraction of a paper is the harshest possible punishment for a scientist". Recent research showed that authors who experience retraction are likely to suffer substantial decline in the volume of citations they receive (Azoulay, Bonatti, & Krieger, 2017;Furman et al., 2012;Fang et al., 2012;Lu et al., 2013;Shuai et al., 2017) which indicates a decreasing reliability in the retracted authors' entire academic record. Understandably, the effect of retraction on authors' reputation is greater when retraction is made because of misconduct than if it is made for honest mistake (Azoulay et al., 2017;Shuai et al., 2017). While some researchers (e.g. Foo & Tan, 2014) call for defaming authors involved in research misconduct by publishing "a shame list" as a way to curb unethical behaviours and reduce retraction, there is a wide agreement that retraction should not be a retaliatory action but rather a way to ensure the transparency and integrity of the scientific literature ( Investigating a research misconduct allegation can be complex (Resnik, Patrone, & Peddada, 2010), time-consuming (Smith, 2006) and emotionally exhausting. Typically, editors-inchief may receive allegations from researchers in the field, associate editors-in-chief, or board members about a given article. The editor may consider these allegations and contact the author(s) to address them. If the allegations are confirmed, the editor may then decide to retract the paper guided by the publisher's ethics guidelines if existed. A major problem arises from the difficulty to prove misconduct and the fact that the research community lacks agreed norms on how to respond to research misconduct (Wager & Williams, 2011). Reputable publishers such as Elsevier, Springer, and Taylor & Francis are members of the Committee on Publication Ethics (COPE).
The key function of COPE is to serve as a platform to discuss issues related to research integrity (Foo & Tan, 2014). By publishing real ethical cases anonymously, COPE provides guidelines that editors-in-chief benefit from when handling misconduct allegations. However, the scientific community is a complex institution where different actors compete and cooperate (Lacetera & Zirulia, 2009). Therefore, detecting research misconduct can be challenging. It also may be reasonable to assume that misconduct in social sciences is more likely to go unnoticed than in applied sciences which may explain the high retraction rates in applied sciences such as medicine and life science. The reproducibility of findings is a major indicator of the validity of a given publication. Authors may question the validity of the original publication if their attempts to replicate the scientific work do not reproduce the original findings (Chen et al., 2013;Resnik & Shamoo, 2017). Irreproducible results can be more harmful in fields in which published research is significantly considered to inform decisions that affect public safety and wellbeing (e.g., medicine) (Resnik & Shamoo, 2017). However, the reproducibility of findings in the social sciences may neither be possible nor of interest for many authors. For instance, if a qualitative study uses in-depth interviews to understand the perceptions or behaviours of employees in a given sector, it is unlikely that the findings will be questioned by the scientific community in the field. Specifically, the context of the study and the interpretation of the qualitative data would always serve to justify any possible suspicion. Therefore, the difficulty of replicating the scientific work and the lack of interest of reproducing findings may explain the low retraction rate in social sciences. Another challenging issue is that while editors-in-chief are largely responsible for research integrity (Marusic, Katavic, & Marusic, 2007), they may often be reluctant to make retractions to avoid potential troubles such as being sued (Cox et al., 2018). Lack of legal regulation and support from stakeholders in scientific publishing are also suggested as key reasons why editors-inchief may be unwilling to be involved in research retraction (Marusic et al., 2007).
A limited number of studies have addressed ethical issues in tourism and hospitality research (e.g., Frechtling, 2018; Moscardo, 2010). In particular, research misconduct is seldom entertained in the tourism and hospitality scholarship. This is likely to be attributed to the relative youthfulness of tourism as an area of study and the sensitivity of the topic. Most of the existing studies examined misconduct practiced by students (e.g., Goh, 2013Goh, , 2015 while much less work focused on misconduct by tourism and hospitality researchers (Deale, Schoffstall, & Lee, 2018;Alrawadieh, 2018). In a recent study, Deale et al (2018) investigated tourism and hospitality faculty members' and graduate students' perceptions of ethical research behaviours. Their findings identified some unethical behaviours including inappropriate assigning of authorship credit in research work and misuse of research funds. However, their findings concluded that the levels of some unethical behaviours occurring in hospitality and tourism are much lower than in other disciplines. Khoo-Lattimore (2018) proposed that despite the competitive research environment in tourism, early career researchers can remain ethical and successful at the same time. She also urged that more emphasis should be placed on the discussion of ethics and codes of conduct in tourism research.
It is argued that research misconduct largely goes undetected and thus observed cases are unlikely to be reflective of the overall amount of academic misconduct (Lacetera & Zirulia, 2009). In this vein, while several studies identified the retraction levels in specific research fields (e.g. Steen, 2011;Karabag & Berggren, 2012), endeavours to use these figures to understand how journals perform in terms of self-policing and research integrity remains limited. Specifically, the existing literature has largely overlooked the importance of linking and interpreting the retraction levels by considering the perceptions of the editors-inchief. Broadly speaking, research on the publication retraction is still in its fancy stage (Shuai et al., 2017), thus, the present study aims to add to a nascent stream of research by drawing on data from retracted articles and data from a survey of the editors-in-chief of tourism and hospitality journals.

Methodology
The present study used a mixed-method design to achieve two key objectives: (i) identifying the retraction level in tourism and hospitality journals and (ii) understanding the perceptions of editors-in-chief of these journals concerning retraction. To identify retracted articles in tourism and hospitality journals, keywords such as 'retraction', 'retracted', 'withdrawal', and 'redress' were researched using several research databases including ScienceDirect and Google Scholar. The popular blog RetractionWatch was also used to find out any retracted papers in the field. Finding a complete and accurate list of tourism and hospitality journals seems to be challenging. Therefore, various data sources (e.g., Web of Science, SCImago) were used to identify these journals. While working on this paper, Bob McKercher used the popular Tourism Research Information Network (TRINET) to share an updated list of specialist tourism, hospitality and events journals published around the world. Although the list was helpful in identifying additional journals, authors decided not to completely rely on the list as it contained book series and journals with unverified or not accessible websites. Finally, a total of 168 tourism and hospitality journals publishing in English were examined one by one to see whether any article was retracted. The Services Industries Journal and Journal of Business Research were also included as these two journals are popular among tourism and hospitality scholars and because they widely publish works in the field. Profit-oriented journals with questionable peerreviewing practices known as "predatory journals" (Alrawadieh, 2018) were excluded. Conference papers, working papers, and book chapters were not considered.
Following this stage, only 5 retracted articles were identified. Given the low retraction rate, it becomes interesting to investigate as to whether integrity in tourism and hospitality research is so high that editors-in-chief and publishers have no articles to retract or whether editors-in-chief are inclined to compromise malpractice. To answer these questions, a short online questionnaire using Qualtrics survey software was e-mailed to all (co)editorsin-chief (N=181) of the journals included in our study (168). The survey contained a covering letter explaining the research objectives and assuring anonymity of the editors-in-chief. The survey included a number of open-ended questions as well as questions aiming to quantify the misconduct allegations and to understand respondents' rating of the misconduct in tourism and hospitality research. For instance, the first question in the survey reads as: "our primary findings reveal that only 5 articles were retracted from over 160 tourism and hospitality journals. In your opinion, how can this be explained? (i.e., is this an indicator of high level of research integrity?). The rationale for collecting data from editors-in-chief stems from the fact that journal editors-in-chief are regarded as gatekeepers of research who are largely responsible for the integrity of the scientific record (Marusic, 2010;Marusic et al., 2007). Therefore, understanding their perceptions and getting insights into their practices are of significant importance to better understand the integrity of tourism and hospitality research.
The online survey was sent over a period of seven weeks from the 11 th of September to the 31 st of October 2018. Of the 181 (co)editors-inchief surveyed, only 21 responded achieving a low response rate of 11.6%. The qualitative data were content analyzed independently by both authors using an inductive approach (Braun & Clarke, 2006). The authors discussed their coding/abstractions and agreed on the final results presented in the study. In vivo quotes were used to enrich the discussion and to enhance the reliability of the data (Elo & Kyngas, 2008).

Findings and Discussion
As seen in Table 1, to date, only five articles were retracted in the field of tourism and hospitality which confirms the notion that retraction remains considerably rare (e.g. Furman et al., 2012;Ajiferuke & Adekannbi, 2018). These articles were retracted over the past five years from high tier SCCI-Indexed journals affiliated to Taylor & Francis. Retraction is suggested to be more frequent in the highest-impact journals (Lu et al., 2013). Both the visibility and the scrutiny are greater in high tier journals than in lower journals (Foo & Tan, 2014;Hesselmann, Graf, Schmidt, & Reinhart, 2017) which explains the prevalence of retraction in SCCI-indexed journals. Duplication was found the key reason leading to the retraction decision. The time between publication date and retraction data is significantly short for 3 of the retracted articles. According to Steen et al (2013), retraction decisions are taken faster now than in the past. By the 10 th of September 2018, these articles collectively received a total of 172 citations. Not surprisingly, the paper of Ryu and Jang which was retracted about 8 years after publication has received the highest number of citations (129 citations). There is empirical evidence that retracted articles may continue to be read and cited as valid research (Wasiak, Hamilton, Foroudi, & Faggion, 2018;Moylan & Kowalczuk, 2016;Budd, Sievert, & Schultz, 1998;Chen et al., 2013;Bar-Ilan & Halevi, 2018;Ajiferuke & Adekannbi, 2018). Citing retracted articles can potentially jeopardize the scientific literature (Chen et al., 2013). However, given the limited number of retracted articles and the recency of retraction in our sample, it remains difficult to measure as to whether retracted articles in tourism and hospitality research continue to be cited.
Retraction can be a source of embarrassment not only for authors but also for editors-in-chief (Katavić, 2014). Specifically, it seems that editors-in-chief consider retraction as having some harm on their own reputation. For instance, one of the retraction notes in our study stated: "the current Editor-in-Chief (who was not the Editor at the time of submission, acceptance and publication of this manuscript) and Taylor  Overall, the findings reveal that retraction is very uncommon in tourism and hospitality journals. This confirms the notion that retraction rate is low in the social sciences (Lu et al., 2013). For instance, Karabag and Berggren (2012) noted that business and economics journal rarely retract papers. Their analysis identified only 6 articles retracted from economics journals. The rareness of retraction in social science may be because many journals in this field do not have clear misconduct policies in place (Resnik et al., 2010).
Editors-in-chief of tourism and hospitality journals were asked to rate misconduct in the field on a 5-Likert scale ranging very serious problem to non-existent. Six respondents considered the issue of misconduct to be a very serious problem, three editors-in-chief considered it to be significant while 7 viewed misconduct to be normal. Only two respondents rated misconduct as insignificant. The rest (3) preferred not to answer.
Eight of the editors-in-chief mentioned that they receive few allegations, typically of plagiarism, about articles published in their journals. For instance, P6, an editor of an SSCI-indexed journal mentioned that only 1-2 allegations are received per year. These allegations were usually raised by peer authors.
Editors-in-chief were informed about the primary findings concerning the number of papers retracted from tourism and hospitality journals and were asked about their opinions. Specifically, editors-in-chief were enquired as to whether the current retraction level reflects the actual level of academic malpractice. Answers to these questions provided mixed results. Eight of the editors-in-chief-in-chief disagreed to consider the low retraction rate in tourism and hospitality research to be an indicator of a high level of research integrity. P1, an editor of an SSCI-indexed journal who evaluated misconduct in the field to be a very serious problem commented: "it [the fact that only 5 articles were retracted from tourism journals] does not indicate research integrity, it suggest[s] that editors and publishers do not care". Similarly, P4, an editor of an SSCIindexed journal who served as an editor-inchief for 20 years suspected that "a lot of malpractice goes unnoticed or is not pursued by the victims or publishers". P9, an editor of a growing ESCI-indexed who is particularly interested in and committed to publishing ethics suggested that: "There are academic malpractices that do not lead to retractions -e.g. friendly acceptances of papers from friends of the editor, solicited citations, forced authorship (e.g. the head of the department is written as a co-author although he/she does not fulfil the authorship criteria)" Yet, some editors-in-chief were more optimistic about the transparency and integrity of tourism research. P11 who rated misconduct in the field to be insignificant agreed that the current low level of retraction is representative of the actual level of academic misconduct and commented: "retraction usually results from significant malpractice in terms of data manipulation which is, in my view, not a common occurrence". The increasing awareness of the best practices in academic publishing, the youthfulness of tourism as an area of study, and the usage of plagiarism detection software programs prior to publishing were highlighted as significant reasons explaining the current level of retraction in tourism and hospitality. P3, an editor of an SSCI-indexed journal suggested that the level of retraction "reflects the awareness of journals and editors-in-chief to catch the problem before the paper is published (at least this is happening in the top and well established journals)". Similarly, P9 mentioned: "In the [name of the journal] we use such software [plagiarism detection software] and for the last 11 years we have rejected more than 20 submissions for plagiarism". P8, who served for 15 years as an editor-in-chief suggested that tourism "is a relatively new study field and researchers are well trained in research ethics".
Overall, editors-in-chief of tourism and hospitality journals are split into two groups; the first has some doubt about the integrity of tourism research while the second views the current low level of retraction as a reliable proxy for the level of research misconduct. Previous studies in more established fields of research suggest that significant portion of authors may engage in some academic misbehaviours. For instance, in their largescale study, Martinson, Anderson and De Vries (2005) found that over one-third of the authors in their sample admitted having engaged in at least one type of research misconduct. To better articulate the perceptions of the editorsin-chief in our study and understand as to whether they are too pessimistic or too optimistic about the integrity of tourism research, there may be a need to investigate authors' own experiences with academic misconduct.

Conclusions
The problem of scientific misconduct has become evident across many areas of study and at an increasing rate (Lu et al., 2013). As a response to misconduct, retraction has become common in some fields of research. The present study investigated the current level of retraction in the field of tourism and hospitality. Moreover, the study drew on the perception of the editors-in-chief of these journals to understand their perceptions toward retraction. By doing so, the study focused on a crucial issue influencing trust in the scholarly communication within the tourism and hospitality domain.
The study adds to the growing literature of publishing ethics and research misconduct. While retraction has received considerable attention in various fields of research particularly in life science and medicine (e.g. Nogueira et al., 2017;Furman et al., 2012;Sox & Rennie, 2006;King et al., 2018), to the authors' best knowledge, the topic was not addressed in the field of tourism and hospitality. More importantly, the present study went beyond by linking the current level of retraction to the perceptions of editors-in-chief. By doing so, the study endeavoured to explain as to whether the level of retraction is reflective of the actual research misconduct. Therefore, despite several limitations, the study makes a significant contribution by tracking the progress and quality of tourism and hospitality research. The present study shows that retraction is quite rare in tourism and hospitality journals. Qualitative data obtained from editors-in-chief offered significant yet mixed insights into the issue which makes drawing practical implications a challenging task. Based on the perceptions of editors-in-chief, it becomes difficult to assume the full integrity of tourism research or to confirm the existence of significant academic misconduct. This may be stemming from the lack of agreement among scholars on what constitutes a research misconduct (Resnik et al., 2015). The attitudes toward research misconduct fluctuate based on the research gatekeepers' levels of optimism or pessimism toward the research community. In this vein, the study does not call for more rigid self-policing on the part of the journals nor does it underestimate the volume of the research misconduct. Thus, the present study leaves several questions unanswered and encourages follow up studies.
The current study is not without limitations. First, only tourism and hospitality journals publishing in English were considered. Therefore, the findings should be regarded as indicative rather than exhaustive. Second, the low response rate achieved from the editors-inchief is a major concern. While we assume that the topic may have been of significant interest for editors-in-chief, they may have been reluctant to respond given that both authors are early career researchers. Thus, a referral sampling would have yielded a higher response rate. Closely related to this issue, editors-inchief provided short responses limiting the ability of gaining richer insights. Third, while an extensive search using different databases was undertaken to identify retracted journals, the study does not claim to have identified all retracted articles in the realm of tourism and hospitality. In particular, with a careful scrutiny of erratum statements, one can easily deduce that some research misconduct is disguised in form of erratum. Finally, as stressed early in this paper, the mixed results obtained from the editors-in-chief leave more question unanswered. This should encourage further research to address editors-in-chief's motivations for and barriers to investigate research misconduct and engagement in retraction. Future research may also address the influence of retraction on authors of retracted papers. Specifically, in the event their papers were retracted for scientific misconduct (not honest error), it may be interesting to understand their motivations, their postretraction behaviour and academic performance.