On the one hand, I find on occasions that even while I am in the process of providing a response, I feel uncertain about whether it is to the correct question. On the other hand, the information provided often falls short of what is necessary to inform a proper if not a comprehensive answer. I understand that sometimes the aim is to foster discussion as a means of fixing the puzzle. However, the questions often seem to be about specific current issues/difficulties. Why is it then that appropriate and necessary information is so often omitted?
I think that a lot of times the reason for that is not just people wanting to start a discussion, but a lot of people do not necessarily know the information that is going to be necessary to help someone answer their question itself. The reason I say this is from professional experience in STEM education and research: I have worked with many colleagues who understand the science and literature of their discipline, but do not know the first thing about data analysis regarding their discipline. They sometimes cannot translate from what a statistical test requires for data and the corresponding data structure itself to actually making the data look like that. I think too often we don't focus on things like data structure that can really help to solve our problems in data analysis itself. I know for a fact that most statistics courses do not harp on these things nearly as much as they should. I know doing statistical consulting that that is the most annoying thing I maybe have ever experienced. You try and try and try to walk them into what should be happening, but they still cannot answer your simple questions. That's a longer answer, but I really think it is a general lack of abstraction and comprehension of what data analysis and data science truly requires. Hopefully with collegiate emphasis on data science (I am applying for faculty positions in that interdisciplinary area), that will change.
Hello,
One of many reasons could be that some researchers want to keep brilliant ideas to themselves before they are published.
Regards
Troy,
Information communication depends on the source information, how the information is encoded, the communication channel, the decoding and the final processing at the destination. Information can be lost in any one of these stages, from not understanding the problem, to how it is formulated and transmitted to how it is received.
More specifically with respect to data analysis questions is that as Matthew mentioned there is a lack of understanding proper methodological issues in statistical analysis. and sometimes it is not present at all. Some scientist are more worried about gathering the data and do not worry on how to properly analyze the data until they face the problem. Usually at this stage it is too late and it may even be necessary to repeat the experiments because of improper setup (introduction of biases, etc.). At this point the person is so involved in just trying to solve the problem as is rather than looking at the more obvious solution of redoing the work correctly.
I do not think there is a clear and definitive answer. Research continues to evolve as do methodological approaches, and analysis. Whist some things remain constant, there is more urgency to contextualize.
I also find this frustrating, but more often than not, it seems to be a kind of "generalized" lack of knowledge. That is, the person may have trouble formulating a question adequately because they simply don't know what they are doing.
As others have said, there seems to be a tendency to know the mechanics of methods without enough understanding about how to use them in actual research applications. I'm not sure how people get into this position, but poor quality of instruction and lack of effective mentorship could have a lot to do with it.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis