4

Question: How can I formulate a hypothesis for an engineering project that I help to develop but probably won't have time to test?

Explanation: The project is to develop a system that will train clinicians to write a certain kind of research data. This type of data will be posted into a community where clinicians can discuss and collaborate on it. I am working on this project and it will be the basis for my master thesis. The value of this project for the company is to generate more of the research data and to thus to create a community that actively discusses its products. Thus, my hypothesis could be something like: The new system will help to generate more research data and support discussion about the company's products. However, I can't test this hypothesis because the marketing and acceptance of the system will take too long. Thus, probably when the hypothesis would be testable (i.e. the first people are using it) I must have already delivered my thesis. I can't think of a hypothesis that only includes the setup of such a system.

Can anyone think of something or give me some hints on how I can come up with a hypothesis or handle this situation?

Improve this question
3
  • 1
    As this question has received some votes to close, I'd like to highlight that (a) while the question in present form is specific to engineering, it is applicable to most applied research fields, and (b) it fits the category 'Requirements and expectations of academicians' as mentioned in the Help Centre.
    – AppliedAcademic
    Commented Jun 22, 2018 at 12:49
  • I think the question is misworded. As I understand it you already have a hypothesis, but your real question is something like "How can plan the work to be able to write an acceptable thesis that must be delivered before the full results are available. There are some good answers towards that already.
    – Flyto
    Commented Jun 22, 2018 at 13:19
  • Thank you both so much for your answers. Both of them have been very helpful and I will as suggested only focus on smaller parts that I can test in the given timeframe.
    – user94400
    Commented Jun 27, 2018 at 11:44

2 Answers 2

Reset to default
3

For your specific case:

You could create a basic version of the system in a short time and conduct a study with a small number of clinicians. Have one group of them use it, and one group not use it. Define metrics that capture user variables (ease of use, tangible benefits) and system variables (how much different features are used, how much it adds to any existing discussion on products etc.). Make small changes in the system to change these variables, and map the outputs. By doing this, you already have an alpha-test, which will help you pitch the system when it is actually developed.

For the general case:

Often it is not possible to test the actual system/design/product in a limited time-frame. A good solution is to identify different parts of the system/design/product, compare these with existing systems and isolate parts that are unique and/or critical improvements. Develop just these parts, and test them either physically or through some kind of simulation. If it is a simulation, do some validation testing to increase confidence. Document everything, and generate sensible statistics during testing. Honestly report all the positive outcomes, the flaws and methods to tackle the flaws.

Improve this answer
2

So you are writing a thesis that does not contain experiments because the time frame is too narrow to test the constructed system. On the other hand, you are not writing a purely theoretical thesis because your aim is to solve a practical problem.

So you definitely need a hypothesis like the one you suggested. If you cannot do experiments yourself, you need to rely on the literature. You need to look up studies and other experiments that support the means and approaches that you suggest in your model. In this way you can have a hypothesis, make it plausible, without really "proving" it by experiment.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .