What role does the null hypothesis REALLY play in the scientific process?

Ever since I first started studying psychology, this has always been a question burning in the back of my mind, exactly what is the point in the null hypothesis? To me it’s always seemed terrible logic to make a hypothesis solely based on the idea that nothing significant is going to be found. What use would it be to conduct research to come to the grand conclusion, “This was pointless!” Let’s be honest though, this is a terribly unintelligent way to look at the null hypothesis. Just because it doesn’t necessarily prove something, doesn’t mean that it doesn’t hold scientific value.

Let’s talk hypothetically for a second. I’m conducting some research into child aggression, and the study involves showing children two different videos. The first being an episode of in the night garden, (for those who have never watched this, here’s a picture of iggle piggle, he’s awesome XD) the other being an episode being an episode of power rangers. We use the appropriate methodology, (independent measures, mixed sex groups, similar age range ect) and after viewing the children playing after watching the videos, we find there are no significant differences in behaviour of the children. Now potentially this could have seemed like a waste of time, however this opens a lot of debate. There may not have been any aggression change in children during these videos, but what if different programmes were used? What if the children already know one another? If we use a sample entirely of a single sex, do males show more aggression than others. For every answer there’s always another question, but it’s these questions that define the inquisitive nature we have as humans.

I’m willing to admit that the more memorable research, Zimbardo, Milgram, Skinner to name a few, have their null hypothesis rejected by significant findings. However without others before them, would they ever have been driven to conducted research in their respective fields? If it hadn’t have been for the far out and revolutionary theories of Freud, would any of these psychologists ever wanted to study behaviour? Even now, 100’s of psychologists will be conducting research inspired from any of these men, yet potentially most of them could find no significant findings. Will that stop any of them? Quite simply, no. There will never be an end to this strive for knowledge until we know everything. An infinite possibility of ideas and concepts to be either proven or discredited. Yet all of this potentially formed by the first idea. Like a seed of a tree planted, and with everyday more and more branches and leaves grow. As long as the tree has enough water (it’s thirst for knowledge) then it’s potential is unlimited. And if that isn’t both an incredible and terrifying thought, I don’t know what is…

Advertisements

2 thoughts on “What role does the null hypothesis REALLY play in the scientific process?

  1. I think null hypotheses are just as important as normal hypotheses, finding out that your experiment produced no significant results is not a waste of an experiment, but enlightens you to new thinking. For example, if you found that there was no significant relationship between eating cake and aggressive behaviour then this is a good thing, it means we can rule out cake as a trigger to kill someone. A strange example i know but im tired and just made it up off the top of my head. But the argument still stands, it means we can go out and see if cake has an affect on other behaviour, or if other confectionary causes violence. Which has actually been found with Cheetos… Apparently theres a food colouring in Cheetos that caused a girl to attack her mum, she was a blackbelt in some form of martial arts and the chemical caused hyperactivity and consequently put her mum in hospital. I tried to find this story online but it could be a myth. However, this is a pretty good paper on food colouring and hyperactivity- http://www.thelancet.com/journals/lancet/article/PIIS0140673607613063/abstract

    So we deviated off topic but i guess im trying to say that i agree, all results and hypotheses are important, at least if it comes back with no significant results it can be said that the scientific community tried and are looking busy 🙂

  2. i agree that the null hypothesis, whilst on the face of it appearing rather pointless, is actually hugely important. i seem to remember a lecturer last year saying that when we conduct experiments, we are actually testing the null hypothesis, making it an integral part of the science. However there is one major problem in science in regards to the significance of findings; Publication Bias. i agree with what you said that a non-significant result can tell you just as much about what you were testing than a significant result, and opens the door to many exciting opportunities, but as this study found (http://www.sciencedirect.com/science/article/pii/014067369190201Y) it seems that all too often rich and informative studies that don’t find significant results are not getting published. This presents a whole Smörgåsbord of problems. firstly, if anyone in the wider scientific community wanted to conduct further research, much like you were discussing, then they would not be able, as they simply wouldn’t know the studies existed. it also means things such as meta-analyses can often be very biased as well, as they can only investigate studies that have been published, as outlined by these two studies (http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020409) (http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020334). I think one of the main causes for publication bias is the fact that people view non-significant results as basically useless. but this is not the case, and as you have mentioned, you’re null hypothesis being confirmed is not the end of the world, and opens up doors for a very wide research field.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s