Science is..
- Thread starter Desanto
- Start date
Fountain
Sannin of the Scrolls 📜
- Joined
- Nov 16, 2012
- Messages
- 5,415
- Reaction score
- 367
They don't, In fact they will usually use a hypothesis until further evidence and facts are found, Nothing means no credibility not that is 100% false because you never know, That's why they use a hypothesis. They aren't really rejected the way i see it.
Aim64C
Anbu Operative 🎭
- Joined
- Dec 18, 2012
- Messages
- 3,681
- Reaction score
- 608
I think what you mean by the "Null Hypothesis" is what is known as "Negative Confirmation."
The Null Hypothesis is the idea that the conclusion of any experiment must be proven against an argument supporting the null conclusion.
In other words - if my experiment was to test for a relationship between sunlight exposure and plant health and my conclusion states that there is a negative correlation (more light = worse plant health); all other scientists must look at my experiment as if the conclusion is null - as if my experiment was unable to support my conclusion.
Properly utilized - the null hypothesis is part of the foundation of scientific discourse. It is, essentially, the idea that every experiment must present a persuasive argument for the validity of its conclusion. Otherwise, the hypothesis of every experiment is that there will be no conclusion - a null conclusion - to be drawn from it.
From the perspective of the one doing the experiment - it is a means of designing the experiment, of foreseeing the need for controls on the experiment.
From the perspective of peer review - it is the necessity of questioning the validity of any conclusion. So the plants with more light exposure died... but were they all in the same soil? Were they all the same plant? I see that your plants were set under a 250 kilowatt light source indefinitely - was temperature controlled... and was there a group of plants that was not set beneath a photon torpedo 24/7?
Within the world of statistics, it is incredibly important.
Do people who drink soda have better luck at casinos?
The Null Hypothesis states that there is no relationship between these two metrics. Drinking more soda should not have any statistical effect on one's wins/losses at a casino.
You have to be able to illustrate that there is a notable trend, to begin with.
How was your data collected? What is the sample size? What controls were made on the sample data?
How does this compare across similar activities - does drinking, in general, show the same trend? Alcoholic drinks? How about decaffeinated? Artificial sweeteners?
What is a 'win?' Slot machine? Poker? Roulette? Craps? Blackjack?
The Null Hypothesis is the idea that you haven't proven anything - and even when you have addressed as many concerns as practical; the validity of the experimental conclusion is fragile and can always be overturned by experiments illustrating a correlation/causation between factors that were not controlled in your experiment.
This is not the same as the assumption of impossibility - or 'proving the negative.'