are research hypotheseses misleading?

[Tnad]

1.given that it is common knowledge that one needs to formulate a hypothesis before carrying out a research ( this applies mostly to scholars writting their theses). Then, isn't true that these hypotheses acts as misleaders since they fuel the confirmation bias of the researcher?

 

2.again, if confirmation bias exists, disconfirmation bias must exist as well(seems like they are interdependent in a way!), but how can one give evidence of its existence without making proof that he is also influenced by confirmation bias?--(i mean how to argue neutrally about one of these biases)

 

3.i think a 'fair' researcher should navigate btn confirmation and disconfirmation biases. what do you think about it?

 

Thank you.

[insane_alien]

1.given that it is common knowledge that one needs to formulate a hypothesis before carrying out a research ( this applies mostly to scholars writting their theses). Then, isn't true that these hypotheses acts as misleaders since they fuel the confirmation bias of the researcher?

 

eh? the point of a hypothesis is to have a testable prediction. you then run the test in an experiment by limiting the other factors. the data will either prove the hypothesis wrong or not. nothing misleading about this, just a statement of what the experiment aims to discover.

 

3.i think a 'fair' researcher should navigate btn confirmation and disconfirmation biases. what do you think about it?

 

Thank you.

 

by your logic, even an idea is confirmation bias. it would be impossible to do anything without being heavily biased. the major flaw in your arguement is that you completely fail to take into account that the level to which a hypothesis reflects reality is not opinion but the result of experimentation and testing.

[lemur]

I think it is wise to self-reflect on your interests and biases when doing theory and research-design. Many people think their methodologies are air-tight in generating neutral answers to questions, whether the questions are biased in the first place or not; but I tend to think such methodologies reproduce the biases built into their research questions. Ideally, researchers should approach research-question formulation from different biased points-of-view. That way, you could examine the difference between research outcomes depending on the biases that the researcher had while performing the research. The first problem, however, is identifying what constitutes bias in the first place, since many scientists will be biased toward seeing their pet theories or methodologies as inherently objective and unbiased, just because it is in their professional self-interest to do so.

[Mr Skeptic]

1.given that it is common knowledge that one needs to formulate a hypothesis before carrying out a research ( this applies mostly to scholars writting their theses). Then, isn't true that these hypotheses acts as misleaders since they fuel the confirmation bias of the researcher?

 

Data can be collected without any hypothesis. Usually the researcher won't be able to see whether a specific data point in any way relates to the hypothesis, until the data is collated and analyzed. In cases where confirmation bias is a concern, the scientists often choose to do a double-blind study.

 

2.again, if confirmation bias exists, disconfirmation bias must exist as well(seems like they are interdependent in a way!), but how can one give evidence of its existence without making proof that he is also influenced by confirmation bias?--(i mean how to argue neutrally about one of these biases)

 

3.i think a 'fair' researcher should navigate btn confirmation and disconfirmation biases. what do you think about it?

 

Thank you.

 

Yes, when people know they are biased they adjust by intentionally going against their bias by the same amount they estimate their bias is.

[Tnad]

you then run the test in an experiment by limiting the other factors

I didn't mean in case of experiments but Research.

 

the point of a hypothesis is to have a testable prediction

.

exactly, you start your investigation when you have already predicted the outcome! don't u think this is a key factor that will influence yr confirmation bias.

[insane_alien]

I didn't mean in case of experiments but Research.

 

research typically involves experimentation.

 

exactly, you start your investigation when you have already predicted the outcome! don't u think this is a key factor that will influence yr confirmation bias.

 

You predict the outcome of an experiment. you don't then pretend you've done the experiment and carry on, you go do the experiment.

 

IF and ONLY IF the experiment matches the prediction do you carry on your merry way, if it fails you leave it. the experiment is unaffected by any particular biases the experimenter had (otherwise a whole lot of historical experiments were very very racist despite having nothing at all to do with race)