Theme 4:1

As we will be discussing method, and quantitative method in particular this week, I chose to read Henning Rode’s article “To share or not to share, the effects of extrinsic and intrinsic motivations on knowledge-sharing in enterprise and social media platforms” (Journal of Information Technology, volume 31., June 2016).

Rode’s article present a new phenomenon that has emerged in recent years: Enterprise Social Media Platforms (ESMP). These platforms are very similar to the social media technology we are used to for private use (e.g. Facebook) but have been developed as a means to share knowledge and information within a corporate domain. Although sharing the same sort of functions as regular social media platforms, ESMP has proved to be difficult to integrate into the working environment due to workers indifference in actively participating and sharing information on a voluntary basis. In order to determine what causes this passivity amongst workers, Rode performed a quantitative web-survey at a German company. Out of the 7946 persons who were invited, 492 participated in the survey. The questions in the survey where based on Rode’s four hypothesises with answers based on a binary scale (yes/no) and some on a 7-point scale (‘strongly disagree ’ and  ‘strongly

agree ’ and by  ‘does not apply at all ’ and  ‘applies fully).

In the discussion of the essay, Rode admits to the study’s limitation, stating that by only having one company representing the phenomenon might have affected the result. I am definitely agreeing with him as different companies have various company-cultures, which I believe affects their willingness to adopt a new system and actively participate in it. Further on, Rode is also worried about whether the questions have been understood correctly (mainly the concept of “knowledge” as in knowledge sharing) as there have been no opportunities for the participants to have the questions explained. Finally, as Rode presents the result of the study he is having a hard time figuring out a correlation between the hypothesises that were supported. This is caused by the lack of any explanatory, in-depth questions, as Rode’s only posed questions to answer “what is” (descriptive).

Although Rode is aware of the weaknesses in his study, the are a two profound limitations he does not mention: firstly, Rode is only using four hypothesises to define all of the questions. Therefore, the result of the study can only be either verifying or falsifying one or several of the hypothesises. If the actual reason why workers do not adopt the ESMP is not defined by any of the hypothesises, it will naturally be excluded from the result. This means that although Rode’s result have some support, it might not be the only (or even the most accurate) answer as the outcome might have differed if workers were to answer/discuss the issue freely.

In addition to that I also want to point out how the survey’s participants can have affected the outcome. Obviously, having 492 persons answering a survey will give some sort of valid result, but as the survey is voluntary – as is the ESMP it is trying to analyse – surely will the most passive users of the ESMP also ignore the survey. In other words: the main people that Rode is trying to analyse will not be represented in the final result.

If Rode would insist to use a quantitative method for the study he should at least make the survey mandatory for a certain amount of people in order to reach the target he is trying to analyse. Also, he should not restrict the questions to four hypothesises only but instead use a categorized framework that include more alternatives. But above all, I would suggest an additional qualitative study in order to find out why the ESMP’s have not been adopted and how it can be improved. Also, a quantitative, in-depth study would show the correlations between results (which Rode admitted to be missing in this case).

However, I do not think that this study could be performed by qualitative methods only. Although qualitative methods can give a deeper understanding of phenomena and avoid misunderstandings/misinterpretations (by interacting with the research subject directly), a qualitative result cannot be generalized the same way as quantitative. Having 10 people participate in in-depth interviews would not produce a more comprehensive result than the 492 people than answered.

Also, another issue in a qualitative study would be the lack of anonymity. In this case, it is not certain how honest people would be in interviews when answering questions about their employer.

Therefore, I would not suggest replacing a quantitative method with a qualitative, at least not in this case. I believe that only by combing the two a more useful result would emerge. 
This brings me to the article "Drumming in immersive virtual reality" by Illias Bergström et. al. It seems to me that they have tried in some ways to combine these two by performing interviews in addition to behavioural data. Still, as the study was mainly quantitative and did not pursue any in-depth interview there was no results that could explain the cause of the outcome. This shows that a method does not necessarily reach a qualitative standard only by adding interviews. I would say that this sort of quantitative study is presenting a good foundation for further research as it describes a phenomenon in a general way. The same way as Rode's article, Bergström et. al has potential to provide deeper understanding if they (or other researchers) approach the study with an additional qualitative method. 

Rode, Henning  “To share or not to share, the effects of extrinsic and intrinsic motivations on knowledge-sharing in enterprise and social media platforms”, Journal of Information Technology, volume 31, June 2016, accessed 22.09.2016