5 ms·

This is a critical problem in the Scientific realm right now. Too much emphasis is placed on research that breaks ground - and not nearly enough is put on resea

by paged90 14y ago

This is a critical problem in the Scientific realm right now. Too much emphasis is placed on research that breaks ground - and not nearly enough is put on research that simply didn't work. For example, suppose I have a hypothesis, and I test it out. To fully reject that hypothesis I might expend a year of lab time. This results in a year 'wasted' if I found no result, and expected one (an alternative would be you got no result and it was surprising - thus you found something exciting, but often this isn't the case). This is a year wasted for me, but with so many labs focusing on the same areas, it means that many can double up in their 'time-wasting' endeavours. This effectively means labs are trying the same hypothesis and failing, wasting cumulative time on the same dead-end questions. Every piece of research is a vital piece of an ever-growing puzzle that helps us deduce exactly what is going on. Unfortunately it seems focus has centered on those discoveries that 'cure cancer' or 'find the answer to immortality'. There needs to be a massive re-haul on the way we treat research. Quality research is quality research - regardless of the results it garners.

greeneggs 14y ago
"Quality research is quality research - regardless of the results it garners." Not true. Research with a negative outcome is less valuable. It creates less knowledge, because there are many more negative statements you can say about how things work than positive statements. Negative statements are more trivial than "vital." Failed research projects are a combination of bad luck and bad planning, and you can't separate out these components. Even if more failed projects are published (further stressing the peer-review system), nobody's going to get a job based on a failed experiment. And what evidence do you have that this is a problem? Or that this is more of a problem than before? Even if so, "a massive re-haul on the way we treat research" is not the solution; better networking would be enough to avoid duplicated efforts.
toufka 14y ago
This is true, to an extent. However, as a good scientist I try to frame my questions so that the answer cannot be 'no'. Instead of asking, "does this object do this?" I ask, "what does this object do?" or, "I've logically deduced that this object must do A or B; now I will do an experiment to falsify A, and another to falsify B - whichever stands must be true." Instead of asking questions with a binary result, I try to frame them with an open ended result. It's not always easy to do. And sometimes some answers are more interesting/cooler than others. Further, there are some disciplines/questions which can only be answered with a yes/no answer. But because we know quite well the risks of failure we oftentimes try to mitigate it up front.

by paged90 14y ago

greeneggs 14y ago

"Quality research is quality research - regardless of the results it garners." Not true. Research with a negative outcome is less valuable. It creates less knowledge, because there are many more negative statements you can say about how things work than positive statements. Negative statements are more trivial than "vital." Failed research projects are a combination of bad luck and bad planning, and you can't separate out these components. Even if more failed projects are published (further stressing the peer-review system), nobody's going to get a job based on a failed experiment. And what evidence do you have that this is a problem? Or that this is more of a problem than before? Even if so, "a massive re-haul on the way we treat research" is not the solution; better networking would be enough to avoid duplicated efforts.

toufka 14y ago

This is true, to an extent. However, as a good scientist I try to frame my questions so that the answer cannot be 'no'. Instead of asking, "does this object do this?" I ask, "what does this object do?" or, "I've logically deduced that this object must do A or B; now I will do an experiment to falsify A, and another to falsify B - whichever stands must be true." Instead of asking questions with a binary result, I try to frame them with an open ended result. It's not always easy to do. And sometimes some answers are more interesting/cooler than others. Further, there are some disciplines/questions which can only be answered with a yes/no answer. But because we know quite well the risks of failure we oftentimes try to mitigate it up front.