The article “Tourism and hospitality firms’ response to COVID-19: the role of entrepreneurial orientation and managers’ market recovery perception” was authored by Calisto et al. (2024) and published in “Cogent Business & Management” a publication of Taylor and Francis. The journal’s Clarivate (ESCI) Impact Factor (2022) is 3. The journal is also indexed in Scopus (Q2), DOAJ, Cabell’s directories etc. The journal charges USD 1680 to publish flawed papers with nescient review.

By analysing this article, we will address an important issue of missing details of data analysis methods and incomplete results reporting found in many published articles. This article is intented for those readers who believe that merely mentioning “we applied ABC data analysis method to test the study hypotheses…” is not only unsufficient but unprofessional too. We significantly believe that authors should report complete data analysis methods and present complete reporting to support the results. In addition, journal Editors and Reviewers must ask Authors to provide sufficient details pertaining to data analysis methods and results.

Besides many other issues like language errors etc., the Authors mention that “the study follows a deductive research approach.” (p.5). This mention is problematic because this study is in fact inductive. First, the propositions are not backed by any theory; second, the applied analysis technique i.e., cluster analysis is a posteriori technique. Let’s start dissecting the analysis method this study utilized to test the propositions.

On page 6 the Authors mentioned “We conducted a cluster analysis to test our propositions…” This is the only information they provided on the analysis technique which is not fair and wrong. Therefore, we decided to bring the attention of our readers not to commit such stern mistakes while Editors and Reviewers should not allow publication of articles with incomplete information. Once conducting cluster analysis and reporting its results, one need to mention the answers to the following questions:

How Cluster Analysis was conducted and what software was used?

What clustering method was employed i.e., K-Means Cluster, Hierarchical Cluster, and Two-Step Cluster.

Is the clustering approach person-oriented or variable oriented?

How distance-or-similarity matrix was created?

What type of distance was selected and why was it chosen?

The type of distance is crucial in cluster analysis. There are many distance methods that can be selected based on the overall objective of the analysis. Some of the distances are Euclidean distance, squared Euclidean distance, Manhattan distance, the Chebyshev distance, the power distance, Phi-Square, and the percent disagreement.

The next important matter is rightly reporting the cluster analysis results.

Following are the important cluster analysis outcomes that should be reported: Initial Cluster Centers, Iteration History, Cluster Membership, Final Cluster Centers, Distance between Final Cluster Centers, ANOVA Table, and Number of Cases in Each Cluster.

We can see that almost all the essential information needed to report cluster analysis is missing in this article. We suggest Authors to appositely present the outcome of every analysis method and if they don’t know, they should learn appropriate reporting methods. We also suggest Editors and Reviewers to keep an eye on such stren mistakes committed by Authors. Publication of articles having such serious issues raises questions on the credibility of the journal and its Editors.

To read full article, please click here. Readers are welcome to share their thoughts in the comments below.