Please carefully read the title of this article "Developing process and product innovation through internal and external knowledge sources in manufacturing Malaysian firms: the role of absorptive capacity". Can you catch a big blunder made by the authors of this article? I am sure my readers can! Yes. The authors are “manufacturing Malaysian firms”. What else can I say? The Editors and Reviewers of "Business Process Management Journal" (Social Science Citation Index - Clarivate SSCI Impact Factor for 2022: 4.1, Scopus Q1) completely failed to identify and rectify this big blunder. Besides the grammatical, structural, theoretical, and methodological mistakes, I will highlight a few mythologies followed by the authors of this article.

Vandenberg (2006) highlighted an emerging issue in business, management, and organizational research i.e., the doctoral student training and peer review process have been plagued by myths and urban legends unbeknown to students, authors, reviewers, and editors. The actors involved in the research process blindly follow these myths because they may be taught or instructed by their supervisors to follow many statistical and methodological concepts within the research process which they presume as absolute truth but in fact they’re fake realities. These assertions about methodological myths and urban legends are, however, endorsed by contemporary organizational research methodologists like Anwar (2022), Guide and Ketokivi (2018), McIntosh, Edwards, and Antonakis (2014), and Ronkko and Evermann (2013).

We will include a similar case as an example to establish our position on creation of fake realities using false information reporting and citations. The article “Developing process and product innovation through internal and external knowledge sources in manufacturing Malaysian firms: the role of absorptive capacity” was authored by Ramayah et al. (2020). The authors justified and rationalized the use of Partial Least Square (PLS) structural equation modelling (which is in fact a path modelling technique as highlighted by, for instance, Ronkko and Evermann, 2013) for analysis of their study model. The authors mentioned under the rubric of “Data Analysis” that “The reason why is because this statistical tool enables to examine the proposed measurement and structural model, since survey research is normally not normally distributed and this have the advantage of being able to accommodate small sample sizes without data normality assumption (Chin et al., 2003) (p. 1028).” This statement is methodologically problematic and, hence, cannot be ignored.

First, authors made a heretical claim that “survey research is normally not normally distributed” and attributed this claim to Chen et al. (2003). We didn’t find this claim exactly or with the similar sense in the cited work of Chen and his colleagues and we are sure they had not mentioned this in their article. Anwar (2015) indicated that Gaussianity or normality of disturbances is an important Gaussian Linear Regression Model (GLRM) assumption theoretically rooted in the celebrated central limit theorem (CLT) of statistics. This assumption becomes more important when a researcher’s overall objective is to test hypotheses and make inferences about population using small or medium sized samples (Anwar, 2015; Gujrati, 2004). Relatedly, Ronkko and Evermann (2013), and McIntosh et al. (2014) provided decisive discussions that presenting PLS-PM as a silver bullet to analyse small sample size with non-normal data is not more than an urban legend. This position has been further endorsed by the Editors of Journal of Operation Management (Guide and Ketokivi, 2015) by stating:

"We are desk rejecting practically all PLS-based manuscripts, because we have concluded that PLS has been without exception the wrong modelling approach in the kinds of models OM researchers use. Most of the time, use of PLS is (incorrectly) justified by saying that PLS is suitable for small samples… (p.7)"

Kock and Hadaya (2018) conducted Monte Carlo simulations to disprove the PLS-PM small sample size myth. They found that sample size requirements for PLS-PM are consistent with those required in multiple regression analysis. In addition to small sample size debate, Ronkko and Evermann (2013) adopted the similar stance on suitability of PLS-PM for non-normally distributed data. They mention that the PLS-PM method obtains significance values of parameters by comparing the ratio of parameter estimate and its standard error to the student’s t-distribution. Similarly, PLS estimation is actually an OLS regression on summed scores (Ronkko and Evermann, 2013), therefore, the distributional assumptions asserted for student’s t-distribution and OLS estimation cannot be overlooked. Ringle et al. (2012) urged researchers to give heed to distributional properties of data when working with relatively small sample sizes.

Kock and Hadaya (2018) conducted Monte Carlo simulations to disprove the PLS-PM small sample size myth. They found that sample size requirements for PLS-PM are consistent with those required in multiple regression analysis. In addition to small sample size debate, Ronkko and Evermann (2013) adopted the similar stance on suitability of PLS-PM for non-normally distributed data. They mention that the PLS-PM method obtains significance values of parameters by comparing the ratio of parameter estimate and its standard error to the student’s t-distribution. Similarly, PLS estimation is actually an OLS regression on summed scores (Ronkko and Evermann, 2013), therefore, the distributional assumptions asserted for student’s t-distribution and OLS estimation cannot be overlooked. Ringle et al. (2012) urged researchers to give heed to distributional properties of data when working with relatively small sample sizes.

The above section was devoted to indicate only few of the myths followed by the authors of this article. In addition to above issues, there are many other problems can be seen in the article. For instance, authors presented the “Research Model” in Introduction section which is not an appropriate place to report it; there are no theoretical underpinnings to support the proposed hypotheses; the article needs orthographic editing etc.

We will highlight few more instances of false reporting and citations below:

Authors mentioned in sub-section "Questionnaire design": "Experience of the organization was evaluated using ten items, which were all adapted from Nieto and Quevedo's (2005) study."

If you will read Nieto and Quevedo (2005), you will come to know that they do not report 10 items to measure "Experience of the Organization". In fact, they reported a total of 8 items to measure both "Level of knowledge and experience of the organization", 4 of them measure "Level of Knowledge" and 4 measure "Level of Experience". Therefore, it is wrong to say that 10 items were adapted from Nieto and Quevedo (2005).

In the same sub-section, Authors mention:

"Absorptive capacity (knowledge-processing capabilities) is the mediating variable for this study."

This claim is strange because this study only hypothesizes direct relationships. No indirect (mediation) relationship was proposed in this study. We wonder why Editors and Reviewers didn’t catch it?

In the "Theoretical and Practical Implications" section, Authors mention:

"Further, as fourth contribution, unlike measuring absorptive capacity as a uni-dimension construct (Liu et al., 2018), we measured the elements of absorptive capacity separately. This idea will assist academician and practitioners to understand independent influence of knowledge acquisition, assimilation and utilization on product and process innovation."

By reading this statement, it seems that authors coined the idea of multi-dimensionality of absorptive capacity. However, this is a multi-level and multi-dimensional construct for instance see Bosch, Wijk, & Volberda (2006); Roberts, Galluch, Dinger, & Grover (2012). Again, this is a false claim.

Please read the full analysed article by clicking here.

References

Anwar, C.M. (2022). Emergence of false realities about the concept of "Silaturrahim": an academic social construction perspective. Tourism Critiques, (3)1: 88-97.

Anwar, M. (2015). Data health assurance in behavioural and social sciences research. European Online Journal for Natural & Social Sciences, 4(4), 725–736.

Guide, V.D.R., Jr,. & Ketokivi, M. (2015). Notes from the editors: Redefining some methodological criteria for the journal. Journal of Operations Management, 37(1): 5-8.

Gujarati, D.N. (2004). Basic Econometrics, McGraw-Hill Book Co.

Kock, N., & Hadaya, P. (2018) Minimum sample size estimation in PLS-SEM: The inverse square root and gamma-exponential methods. Information Systems Journal, 28, 227-261.

McIntosh, C.N., Edwards, J.R. and Antonakis, J. (2014). Reflections on partial least squares pathmodeling. Organizational Research Methods, 17(2): 210-251.

Ringle, C. M., Sarstedt, M., & Straub, D. W. (2012). A critical look at the use of PLS-SEM in MIS Quarterly. MIS Quarterly, 36(1): 3-14.

Rönkkö, M. & Evermann, J. (2013). A critical examination of common beliefs about partial least squares path modeling. Organizational Research Methods, 16(3): 425-448.

Vandenberg, R.J. (2006). Introduction: Statistical and methodological myths and urban legends: Where, pray tell, did they get this idea? Organizational Research Methods, 9(2): 194-201.