h/w for my TA

OUTLIERS?! SAY WHAT?!

Informed Consent

Is it ethical to use internet resources without the publishers consent?

The Statistics of Facebook!

what a beautiful day for blogging…sigh

Ok guys, I’m trying to try to ignore the fact there’s a bbq going on outside and write an exciting blog questioning if ‘statistically significant’ means there is an effect in the real world.

When it comes to stats we use the term ‘statistically significant’ when a result is unlikely to have occurred by chance.  Many researchers urge that tests of significance should always be accompanied by effect size statistics, which approximate the size and thus the practical importance of the difference.

If we find our significance level is lower than .05, then the results are deemed significant. This is when researchers feel they can crack open the champagne. But ‘statistically significant’ does not necessarily mean that we would see an effect in the real world. Researchers must consider  that there is a distinction between statistical significance and practical significance. A practical significance means that the treatment effect is large enough to have practical application. Whereas a statistically significant result simply shows us how likely the observed effect is to have occurred by chance (Gravetter & Forzano, 2009). 

Since Fisher (1925), it has become the norm for psychologists to use the hypothesis significance testing model to dismiss sampling error when looking at their data, despite the fact there seem a fair few flaws with the procedure.

So with this problem in mind, is there an alternative to significant testing?

The following have been suggested.

1. Effect size- The use of effect size measures has been recommended repeatedly throughout the social science literature (e.g., see Dunnette, 1966; Huberty, 1987; Rosenthal, 1992; Vaughan & Corballis, 1969). The effect measures are indicators of the strength of association between the independent variable(s) and the dependent variable(s). However it has been argued that effect size is best used along side p values in order to get the best results.

2. Confidence intervals- A confidence interval provides a medium for assessing the probability that a parameter lies within a particular range of values. Kirk (1994) showed how a mere significance test can be greatly misleading and that instead, a confidence interval can inform us of what we are really searching for in scientific investigations. However it would appear the research community could do with a little more convincing. 

 

 

Homework for my TA

http://amyray19.wordpress.com/2012/03/11/mind-over-body/#comment-59

http://statsisboring.wordpress.com/2011/10/14/is-it-dishonest-to-remove-outliers-from-our-data/#comment-53

http://bloggenrolla.wordpress.com/2012/03/11/bias-in-sona/#comment-72

http://psud56.wordpress.com/2012/03/11/should-psychology-be-written-for-the-layman-or-should-science-be-exclusively-for-scientists/#comment-32

Is it ethical to use information we find on the internet in research?

Who could have predicted that within such a short space of time, the internet would transform the way we as individuals communicate, keep in touch with one another, share opinions and ideas, express views and fears, and all behind the veil of anonymity. One example which shows how an idea can spread like wildfire is the quite a recent craze to hit Facebook- the campaign to make the African Warlord -Kony- famous. One day you’ve never heard of the name, the next its all over your news feed.

It seems you can just about find out anything you need, with the internet offering an overwhelming amount of information. Its all too tempting to involve information we have found on the internet when looking into an area of research, but to what extent can this information, ethically, be used? Everyone knows that blogs are open to the public, or in other words, anyone of the 2,267,233,742 internet users in the world. With the publisher of the blogs knowing this, does that mean that they have given us consent to use their opinions or views as data? Or must we directly ask the publisher to give full consent? 

Another way of thinking of consent is as permission. When I think of it that way, by simply publishing a blog it doesn’t mean we are giving someone permission to use our material.

A hurdle to obtaining consent is anonymity. Many can hide behind a false identity, or no identity at all. 

My initial reaction to this question was to roll my eyes at how over the top we can be when tiptoeing around the touchy subject of informed consent…but then I pondered… how would I feel if someone where to use my blogs?? So long as I was to remain anonymous I don’t think I’d have a problem with it (although if they did make contact to ask for consent I’d tell them not to waste their time!)

 

However there are plenty of blogs out there which have slightly more emotion attached than my forced statistics blog!

There are plenty of blogs out there which offer support to individuals going through a tough time, would it be ok to use their heartfelt opinions without consent?

Data from the internet can be very useful, but also full of hurdles and ethical dilemmas because of this, researchers need to approach information found on the internet carefully.

http://www.internetworldstats.com/stats.htm

http://www.simplypsychology.org/Ethics.html

Kissing deadlines!!

Homework for my TA

http://leprice91.wordpress.com/2012/02/19/reliability-how-important-is-it/#comment-49

http://danshephard.wordpress.com/2012/02/19/sona-for-the-students-or-for-the-staff/#comment-24

http://statisticalperrin.wordpress.com/2012/02/20/should-psychology-be-written-in-laymans-terms-or-kept-scientific-for-scientists/#comment-46

http://psud6d.wordpress.com/2012/02/02/is-psychological-research-biased/#comment-40

Somethings are better left untouched?

As the human race as progressed, so has our quest for knowledge. We have developed a thirst for a better understanding of why we are here, and why things happen. We have developed to a point where it is perfectly normal to want to know, not only more about our own planet, but to know about other galaxies and far beyond.

But where does this stop? Should it stop?

Are there some areas of research that could possibly have a detrimental effect upon society?

What we as psychologists study can sometimes be deemed socially sensitive topics, such as sexuality, race and gender. Some would go as far as to say that to mess around with somethings, would be ‘playing God’.

Hands up those of you who’ve seen the film Limitless? This film kinda explores this idea. A pill is created which allows you to access all of your ‘brain power’. As you can imagine the next hour and a half is followed by Bradley Cooper running around New York, trying to figure out what the heck to do with this incredibly cool power (it definitely does not involve pulling beautiful women and winning a crazy amount of money) and of course things go a bit pear shaped.

On a more serious note though…

In an enthusiastic devotion to learn more about the human thought process and behavior, it is clear from a modern day perceptive that many early psychologists were unethical in their quest for knowledge, leading to some major violations of ethics codes and standards.

One experiment which stands out in my mind is one carried out by the world renowned psychologist Dr John Money.

Money developed the theory of gender neutrality. He believed that what he called, our ‘gender identity’ (being what makes us feel, think and behave has males or females) is not determined at birth.

In 1965 he was given an opportunity to put his theory to the test. Bruce, an identical twin, at the age of 6 months old was left without a penis due to a circumcision procedure going horribly wrong. Money jumped at the opportunity, and the radical decision was made to raise Bruce as a female, and Bruce became Brenda.

Over the next three decades this case arguably became one of the most controversial in the history of psychology.

After forteen years of being Brenda, the impact of living as a female had taken its toll and Brenda reverted back to her true biological sex. Moneys theory had failed, with tragic consequences. David, at the age of 38, drove to his local supermarket and shot himself.

So was this playing god? Could the years of emotional torment have been avoided?

I believe that although research is important and our constant quest for knowledge admiral, there are some things in life better left untouched by science.

“Nothing is so fatal to the progress of the human mind as to suppose that our views of science are ultimate; that there are no mysteries in nature; that our triumphs are complete, and that there are no new worlds to conquer.”

Sir Humphry Davy

http://www.bbc.co.uk/science/horizon/2000/boyturnedgirl.shtml

 

Homework for my TA (Week 3, Semester 2)

http://standarderrorofskewness.wordpress.com/2012/02/04/do-ethics-hinder-potential-findings-of-research/#comment-28

http://psucc8.wordpress.com/2012/02/05/stanley-milgram-was-his-experiment-really-unethical/#comment-21

http://bloggenrolla.wordpress.com/2011/12/09/deception/#comment-32

http://amyray19.wordpress.com/2012/02/05/reliability-in-research/#comment-28

Would we be lying if we were to remove any outliers?

According to Grubbs the definition of an outlier is “An outlying observation, or outlier, is one that appears to deviate markedly from other members of the sample in which it occurs.” To put it simply, an outlier is a point in your data that sticks out like a sore thumb.

When it comes to the removal of outliers it all  comes down to opinion. There are no set rules or definite answers, but I believe there are ways in which you can go about outliers and certain factors that should influence your decision to determine the best course of action in regards to removing or keeping the outliers in your data set.

So what is it that causes these outliers? It is possible that an outlier will appear by chance in any given distribution. Say, for example you were measuring a participants reaction time, and each participant looks to be improving their reaction time with each trail. Except for Bob. Bobs data is showing that his reaction times differ greatly to the averages.

Now it might be that Bob is simply not very good at this task, in which case I believe it to be unfair to remove his data. Or it might be (baring in mind Bob is a student) that Bob is slightly hungover from a spontaneous trip to The Greek, and after consuming far too much alcohol ended up crawling home from Academi at 3am. It might be that in Bobs delicate state he decides that he isn’t going to cause his head any more pain by concentrating and is trying to get the experiment done as swiftly as possible in order to spend the last of his students loan on a boneless bucket from KFC and nurse his sorry, hungover self back to good health. I would like to point out and this point that Bob and the situation told above is totally fictitious and is in no way related to any individual that I know of, any relation is a complete coincidence. But lets just say this is the case, and we can see from his data that he isn’t paying very much attention, is it justified to remove his data?

One test for outliers is Grubbs’ (awesome name) test for outliers.

Grubbs’ test is based on the assumption of normality in data. That is, one should first verify that the data can be reasonably approximated by a normal distribution before applying the Grubbs’ test.

Grubbs’ test detects one outlier at a time. Multiple iterations change the probabilities of detection, and the test should not be used for sample sizes of six or less since it frequently tags most of the points as outliers.

In Grubbs’ test H₀  means that there are no outliers in the data set.

There are loooooads of different ideas and opinions, and I need to celebrate Wales’ extremely lucky victory this afternoon, so to conclude. Discussing outliers is opening a can of worms! Theres no definite answer but lots of things to consider before making that oh so important decision whether to remove them or not!

Happy Blogging!

Nadolig Llawen Simon!

http://psucd8.wordpress.com/2011/11/20/why-is-validity-important-in-research/#comment-34

http://psud5c.wordpress.com/2011/12/09/as-long-as-all-other-ethical-guidelines-are-abided-by-should-the-use-of-deception-be-allowed-to-produce-better-results/#comment-23

http://te9192.wordpress.com/2011/11/25/qualitative-vs-quantitative/#comment-43

http://psucf2psych.wordpress.com/2011/11/25/qualitative-data-is-not-as-scientific-as-qualitative-methods/#comment-39

 

 

Ethics & Stats…my favourite combination!

The declaration of Helsinki has been referred to by many as the cornerstone document of human research ethics. (WMA 2000, Bošnjak 2001, Tyebkhan 2003). The Helsinki convention has been spread by the World Medical Association as a statement of ethical principles for medical research involving human subjects. The 5 basic principals of ethics are as follows-

• Research on humans should be based on results from lab experimentations.

• Research protocol should be reviewed by committee before initiation.

• Informed consent from participants is needed.

• Research should be conducted by qualified individuals.

• The risks should not out weigh the benefits.

Both the Helsinki Convention and the BPS code require the participant are to give informed consent as an ethical guideline. However, in order for almost all experiments in Psychology to be worthwhile and the results to be valid, it would not be possible to ensure there is informed consent. Deception invalidates any informed consent gained by the experimenter, yet we find that deception is oftem unavoidable. The most important findings in Psychology have come from experiments in which deception has taken place, for example Milgram’s 1963 obedience experiment. Many would agree that what Milgram found astonishing and many would deem this experiment as one of the most important in the history of Psychology, and yet many more consider it to be unethical. But was it? After the experiment a survey showed that 84% of participants “glad” to “very glad” that they participated in the experiment. Some even wrote letters that expressed their thanks. Due to the participants being deceived, is informed consent possible?

We have all participated in a Sona experiment. I for one have been participated in an experiment in which some information had been withheld from me. Not about the task I had to do or what they were measuring, it was that they were also seeing if, as a non-smoker, I would respond differently to someone who smoked regularly. Did I feel deceived? Upset? I can honestly say…No. In fact it made the participant all the more interesting, and I actually read the debrief. If however I had known what they were measuring this might have affected the results.

I’m not saying that all of the ethics guidelines should be breeched, merely that in most cases a degree of deception is necessary.

 

Dolig Llawn pawb!