Research Critique: Determinants of Strategic Supply Chain Risk Management

In 2010 Lassar et al. did a grounded theory study (see here for more information on techniques for theory creation) on the question of what the determinants of strategic supply chain risk management might be?

Methodology

The authors used a grounded theory approach to validate their theory. GT consists of a series of interviews where each subsequent interview may build on the last one with the goal of exploring a concept from different angles. In the case of Lasser et al. the focus was on high tech companies located in Mexico. As a first step the authors build their conceptual model for the determinants of strategic supply chain risk management based on literature. They hypothesis that supply chain risk management is built on resources available for the company (resource based view), the network and performance criteria.

Resources include organizational systems and processes, information, knowledge, capabilities, partners and alliances, and other assets.

Based on this theoretical background the researchers conducted 24 interviews within eight different companies.

Results

The interviews lead to a multitude of answers in five key areas:

1. How do firms conceptualize and define supply chain management?
2. What kind of processes and systems do supply chain participants utilize in their enterprises?
3. How committed is senior management to SCM?
4. What role does performance play in supply chain management?
5. What specific risk management actions do companies employ?

Our qualitative assessment of the sample firms – multinational buyers along with suppliers – confirms that a symbiotic triad of resources, networks, and performance are key determinants of strategic risk management in supply chains in one important emerging market – Mexico.

Critique

I seldom use a critique section for an article, since usually I only present articles which I deem well written, reliable and interesting.
This study sounds very interesting at first, so I included it on my reading list, but during the evaluation process I came to the conclusion that I found it unpresentable. I really struggled with myself, but then after I continued reading it, I started to realize that this is a great chance for showing you some of the things I dislike in a scientific paper.

1. The concept proves to be very vague
   The authors build a concept for the influences on supply chain risk management from literature. The result: SCRM is based on resources, the network, and performance. The authors do discuss shortly what they subsume under these terms. But what is conspicuously missing is the reasoning for taking exactly these? Why not go with strategic choice and argue that SCRM is defined by the managers and their environment? Why is it the network and not the type of supply chain which defines the SCRM? As a result, already the first chapter leaves us with more questions than we had before.
2. No pictures
   Even though I am a fan of nicely written paragraphs where a picture of what the author meant just pops into your mind, I also know that this writing-only method is not for everyone. So it is always a good idea to support a concept with a picture containing the key elements and their relations to each other and with their “environment”. For example the authors could have used a three column layout with SCRM on top and the determinants underneath? At least if its that what they meant.
3. Methodological discussion
   Building on the definition of their framework the authors claim to use grounded theory to validate their concept. Sadly, there is no real description of the authors’ understanding of the methodology so I can only analyze the context.
   To my understanding there are two different streams within grounded theory. One (my feeling is, that it’s the more popular one) particularly states, that grounded theory is based on the idea, that the theory has to develop itself during the interviews. This is done to prevent a very important thing: Bias. Every human interaction is somehow biased. So questions during an interview, not only contain the content of the question but also prejudice and believes of the sender of the question. And of course the interviewees not only respond to the content of the question but also the prejudice and biases. In this case the authors already had their completed framework in mind, and were supposedly steering the conversations towards their framework, of course confirming their preexisting believes. Interestingly I do not think you can find this bias easily in the results, but this does not mean it is not there.
4. Results
   Even worse then a bias in the result, is the presentation of the results themselves. The goal of the interviews was to validate the preexisting believes of the authors (that’s is bad as explained above). But if I wanted to validate the concept, why not use this concept as a foundation for the outline of the discussion of the results. Why? I really have problems with seeing the correlation even between the headlines of their results (see above) and their concept? Some do correlate, but for example what does the “employment of risk management actions in a company” have to do with their concept of resources, network and performance. And in my point of view this gets worse the more you read the interpretations. So I am really asking myself: How could the authors really see their framework supported by the interviews?
5. Representativeness
   Last and just a small addition: The authors rightfully state at the end that this study cannot be viewed as representative, since it only included a small sample of companies. So the authors imply that with a bigger sample (and the same method) it would be representative…? No! Grounded theory does not want to be representative, so this argument is just mute. The goal is instead to generate a valid (grounded) theory (not to validate a preexisting theory!), which can then be tested using other methods, like a survey.

Conclusion

I have to admit that I am not an expert on grounded theory either. So I may be wrong with the forcefulness with which I presented my opinion. But even if we abstract from the methodological aspects, the work can be summarized with two points:

• Too little discussion of the methodologies and processes employed to generate these findings and
• No obvious connection between their proposed concept and their own stated findings (or again too little elaboration).

The full study can be downloaded here.