Is it possible to prove a research hypothesis?

Firstly the big problem with this question is the use of the word “proven” this is because this is not scientific and thus would never be thought of in a research experiment. In this blog I will absolutely rip this question apart due to it being unscientific.

The biggest problem with the idea that research can’t be proven, is that some conclusion use deduction and induction to create conclusions. This is a problem because both methods of created a conclusion are far to generalized to create a coherent hypothesis which can be proven completely. For example if a conclusion has been created under deduction this conclusion would be true Alfred is fat and likes cake, thus all fat people like cake, however what if a fat person likes pies instead and is allergic to cake? This is disproved a theory. Very similar in induction e.g. if you observe 3 fat people who like eating cake, does this prove that all fat people eat cake. Of course it doesn’t, it means that 3 fat people like cake, the next fat person who you meat hates cake and is obsessed with pies. Thus this hypothesis is dis-proven.

However surely if there is no problem with you hypothesis it must be correct. For example Pavlov proved that animals salivated when seeing food. He did many tests on dogs to prove this fact and when in the present of food the dogs salivated and he used about 2 – 3 dogs in each experiment he had to create his theory. This is an example of a inductive test, due to his wide range of tests and the number of dogs he used. This means his experiment must be valid and it is a logical theory because every dog he tested salivated. Thus his experiment had been proven.

This is of course not true, there will still be dogs in the world who don’t salivate when food is put in font of them. Pavlov to prove his statement true would have looked at all the dogs the have ever been and will be. This is impossible, thus researchers use null hypothesizes to explain experiments. Null hypothesizes are hypothesizes which scientists try to disprove rather than prove. This is because it is much easier to disprove a experiment than prove an experiment. In the case of Pavlov and his dogs the reason why his hypothesis is still upheld is for the reason that there insufficient evidence that dogs do not salivate, if there was sufficient evidence his theory would be stopped being credited. This creates a paradox that in this world there are no theories which are proved there are only theories which are just waiting to be disproved.

Overall it seems that there is no way to prove if anything is true, only if it is not true. There are things which are things in this world with are almost true, however these are not proven but waiting to be dis-proven. “Nothing is true, everything is permitted”

 

 

References

Reynolds. J.L. (2006). Secret societies: inside the world’s most notorious organizations. (pp. 19). New York: Hachette Books Group USA

Le Doux, J. E. (1995). Emotion: clues from the brain. Annual Review of psychology, 46, 209 – 235

Advertisements

“Is it dishonest to remove outliers?

To understand if it is or not dishonest to remove outliners form experiments you must first understand what outliners are. Outliners are “A person or thing situated away or detached from the main body or system.” from Oxford dictionaries online. They are basically results which don’t follow the general correlation. Often they are mistakes made by the experiment, e.g. the scientist misreads the reading off of the scale when measuring the weight of something.

Firstly why it is dishonest or not to keep outliers is that, the entire scientist is removing outliner to make his results look better. It creates more correlation for the scientist so that when the scientist wants to use the data, he/she is able to create a linier graph. This is useful for presentation purposes and it look like his results are unquestionably right. However is this honest? Not really removing the outliners has created correlation however on the other side means that his results are unquestionably right, due to all follow, however let us suppose a drug was created which cured Alzheimer. Out of 100 people 98 got better however the two outliners which got significantly worse and died, is it dishonest then to remove an outline. Some people would say that out of 100 people a 98% is no problem and just to prove make sure that the drug got on the market it would be worth it. However we are not talking about if it is worth it we are asking if is honest, which it isn’t due to we are concealing the truth. You may say thought that this is an extreme case however if you made it honest to remove outliners from this experiment, the hat fits all, it stilly lying to the general public thus dishonest.

Another reason some people say that it is honest for outliners to be removed is because most of the scientific community agree that it is honest to do so is honest. A study by Orr et al. (1991) showed that 67% of the psychologists asked agreed with the statement: “Data points should be removed if they are extreme outliers and there is an identifiable reason that leads you to consider them invalid.” And 4% said “Data points should be removed from an analysis, if they lay in extreme areas … there does not need to be identifiable reason” However only 29% agreed with the statement “All data points always should be included in an analysis regardless of where they lie relative to other data points.” However what would the public think about this. What 71% of scientists are saying that they are willing to get rid of people data just to make sure that your experimental results follow a correlation, however this creates the idea to the public that they are being overlooked in the experiment. This will destroy the repetition of the scientific community for experimentation on people making it less likely to get participants due to their results might be seen as worthless.

Overall I believe that it is dishonest putting removing outliners from results. This also causes problems because it causes the data to become less reliability and gives psychology a bad name due to you manipulating for your own benefit.

 

References

Pirker. C Statistical Noise or valuable Information: The Role of extreme Cases in Marketing Research page 46

http://oxforddictionaries.com/definition/outlier?region=us

Five principles for research ethics As listed By Deborah Smith of the APA Monitor Staff (Part 1 (1 and 2 rules))

Five principles for research ethics listed by the APA

Firstly to understand the five principles you must understand what ethics is. Ethics is the “moral philosophy addressing questions such as what is good and bad, right and wrong, justice and injustice, etc” stated by Fay Shorts power point.

The first of the five ethical principles is that you have to discuss intellectual property frankly with the clients you are doing the experiment with. Out of the first two rules this rule, I have the biggest problem with. I know that these rules are put in place to keep the safety of the clients in the experiments however these rules shouldn’t hinder the validity and reliability of the experiment.

There are multiple problems in discussing intellectual properties with clients. Firstly if people know what they are being tested about they are more likely to lie. As Tangney put it “It’s almost like talking about money people don’t want to appear to be greedy or presumptuous.” For example in Milgram experiments they would have never worked. If Milgram had explained, that the clients were being tested to hurt/kill other people on command a lot less would have hurt the client to the point of death. Also some would have quit the experiment and others would have realised they were killing a person even in pretend due to being about to think about it more. The loss of this experiment would mean the loss of all the understanding of the human mind gained from that experiment. Another problem is that surely this will reduce the reliability of the experiment due to if everyone knows the experiment it means that everyone has one aspect which is all clients are the same. In psychology it is key to have lots of different people to have variation in results if everyone has something has this knowledge doesn’t it reduce the variation slightly? However thought that experiment people because unwilling to do experiments with psychologists so they lost their credibility thus keeping to ethical rules means people are still willing to be tested.

The second rule is to be conscious of multiple roles. This is where “should avoid relationships that could reasonably impair their professional performance” stated by the APA. The problem is that people who psychologists know are often the people who are most willing to do the most controversial of experiments. For example, Bangor University uses Psychology students for a lot of its third year experiments due to there is a sample base of about 300 as well as they are willing to do any experiment to get credits. However the advantage of doing this is that it reduces biases making the experiment more valid. This due to the experimenters are less likely to have favouritism towards people they already know, everyone is random.

Overall I believe that the APA has created rules which hinder the experimentation of psychologists too much and needs to reduce the ethical rules so research can continue unhindered. Out of the first two I have mentioned the first needs to be changed dramatically, or be gotten rid of completely. This is due to it ruins the validity of most experiments costing money and time of psychologists to be wasted.