Blog Comments Week 8
Reply
Quit Your Jibber Jabber (HW – Week 8)
I’m sooooooo sick of reading research papers that I just don’t understand. Why must you torture me with sentences that need a second brain to comprehend!? Of course older papers are hard because they used to be written by them old folks who talk all olde worlde but is there any need for that any more. Is it time researchers got down with the kids a bit more, you get me homedizzle?
I think the answer is yes it is about time! And the main reason for this is because I believe research should become more accessible to everyone. Since the creation of the internet people have been striving to make information open to everyone and give us the freedom to get hold of it and use it as we wish, so isn’t it about time scientist #caughtup? I feel there is a need to not only make papers and research more accessible without having to be part of a university or have money to burn on buying them, but to then make that information understandable. And why is this? Well we have the potential to open ground breaking research up to the public, to inspire people and more importantly show people all the incredible things that are being discovered in the world around them.
I really can’t think of any positives as to why papers need to written with ridiculous jargon that only masters of the proffesion can understand, so i open to floor to discuss!
So to all boring and pretentious researchers out there… talk proper and you’ll be #trending before you know it, izzit.
YAWN. Caela.
#xoxoxox
Comments Week 5 Semester 2
fMRI: A Kind Of Scientifically Assisted Telepathy? Week 5 Semester 2
I recently listened to a BBC4 podcast on brain scanning (kindly left by Mike Beverley – merci beaucoup) as I very much needed inspiration to get my statistical juices going (find here.. http://www.bbc.co.uk/programmes/b00q2qyk ) Well, get my juices flowing it did! It’s a nice little podcast to listen to but what really got me inspired was one of the guys saying at the beginning that fMRI was “a kind of scientifically assisted telepathy”. Well my immediate reaction was POPPYCOCK (if we’re keeping it clean anyway). I mean come on, psychologists are scientists not the stereotypical ‘mind readers’, like when your mates mate asks you what you do and you say psychology and he thinks he’s reaaaally clever by going “well what am i thinking then?”. Not sure what your thinking but im thinking you’re an even bigger douche than these guys…
(clearly getting resentful about how often this happens to me – side note – does this happen to anyone else or do I just look like Derren Brown?! I’m really trying to keep my beard to a minimum too..) So although I think this podcast man is being preposterous I thought he deserved a chance to explain himself since it is the BBC and all.
What basically emerged from the podcast is that fMRI shows us almost a play by play of which areas of the brain have a higher blood flow during certain tasks and processes. According to these guys there’s research to show that these lovely scanners can tell when we’re lying and even predict what we’re thinking. There was mention of research that showed an area active when we look at someone we love. Is that what love has been reduced to? A bit of elevated blood flow? Waw, that’s super depressing. I’m not sure I agree that just one area accounts for love. You don’t have to be looking at a picture of someone to feel love for them. This is far too simplistic for me to buy it. I will admit though that it will take a lot more than a few activated areas on a brain scan to convince me that the force that conquers all is literally just that. It can’t be right!? There’s part of me that doesnt want to find out. Nothings sacred anymore. Our mind, our imagination, our soul, well do we really want to reduce it to a coloured dot on a screen, true or not??
There is some evidence such as this study (see vid below) which has recreated images that people are viewing from the scans. Ok to look at them they’re pretty crap and if you weren’t comparing them to each other then you couldn’t tell what it is, but, damn I have to admit, there’s something there. Everything has to start out really rubbish then get good! This could potentially get really good, especially with the ever advancing and more accurate technology becoming available. It has swayed me slightly that fMRI scanners are pretty clever, however its hardly mind reading, its simply a recreation of a stimulus, not a persons pure imagination. Recreate some crazy image i make up in my mind then you get a wizard badge.
So it looks like science is making some very successful first steps into telepathy. No it can’t quite pull dreams out of our heads yet but I never thought one day I could touch a screen and it would do stuff. The thing is touch screens are awesome and its good that people persevered with that sci fi dream, and as awesome as the notion of mind reading and telepathy is, do we really want to carry on paving the way for that kind of society in a 100 years time? In a generation where we share EVERYTHING (i mean my friend sends me pictures of his weird poo’s ffs!) shall we not start drawing the line at being able to keep it to yourself that you would do rude things to the hottie next to you? Or perhaps you can think of positive reasons for such ‘telepathy’. I’d be intrigued to know your thoughts on the matter.
PEACE OUT HOMIES
P.S. A completely irrelevant but highly amusing photo of the legend that is Paul Daniels.
Caela
xoxo
Is It Dishonest To Remove Outliers From Data? – HWK semester 2
Outliers are very common in statistics especially when dealing with large quantities of trials and I feel that in many circumstances these outliers should be removed in order to get a more accurate picture of the mean. Outliers can really skew the results and pull the data in one direction or another.
Imagine the mean is a really really hot man surrounded by really really hot women. He’s happy, he’s at home, splashing around in the jacuzzi and he’s loving it. Then comes the outlier. It’s his ex girfriend. There she is with her long rope of jealousy, screaming at him from the lobby until he removes himself from his happy place. Now he’s out of his rightful place in the jacuzzi and somewhere on the pool side. Look what she did! This one outlier had such an effect on the mean it took him right from where he should be. Surely she must be banished!?
So why is so important to remove outliers in the first place? Surely the ex girlfriend has a right to be where she is! Outliers are however sometimes caused by an invalid result. Meaning that when you turned up to that study still drunk from the night before and you couldn’t see the button you were pressing on the keyboard, that yes, it is more than likely that this result will have been picked out from the overall data. Just by looking at some overall data it is sometimes very easy to pick out these types of naughty drunken students. Other possible reasons are a poor measurement of the data you are collecting, or a bad design. It is up to the researcher to recognise where the error has come from and address it. I don’t see how it could be dishonest to remove an outlier that isn’t actually measuring what you intended in the first place.
So when is it not okay to take out outliers?? There are many circumstances such as height and IQ where in a sample you may just have a freakishly tall Stephen Hawking amongst you. Is this guy not worth a mention. I feel in these situations where you would generalise to the general population you will still find these outliers there too so you can’t take them out.
So as you can see there are times when you really should take them out but circumstances when you shouldn’t. So how do we tell!? Well it would seem there are methods to do this! How lucky! Check out Grubbs Test for Outliers. http://en.wikipedia.org/wiki/Grubbs%27_test_for_outliers. I think this sounds pretty catchy. Chauvenet’s criterion sounds pretty wild too. http://en.wikipedia.org/wiki/Chauvenet%27s_criterion. These methods are based on normal distributions, means and how far things are from the standard deviation.
So coming back to the question in hand. Is it dishonest to remove outliers? Well, no, because sometimes you need to take them out, and as long as due care and attention is paid and you can justify your reasoning then everything will be fine!!
Blog Comments – Week 11
Dearest Joe…
Three is my favourite number so I’m sticking with that amount. MERRY CHRISTMAAAAAAAAAS 😀 x
Is it unethical to use public data from the internet in research?
As Facebook slowly takes over, forums overflow with discussion and just about everyone has a blog (i hear they’re even encouraging students to do them at University!!), the internet suddenly has a wealth of information on just about anything. People divulge their fears, opinions, views, and feelings on all kinds of subjects for all to see and do so behind the anonymity of they’re computer. What we must ask ourselves therefore is this; if someone posts a viewpoint, or opinion, or interesting conversation in a public domain where anyone can access it, is it okay for a researcher to simply come and take that data and use it in their own study??
I feel on many levels that yes, it is okay to use this information. If you put something in a public place then you are giving your consent for people to use that information. If a researcher quoted a discussion in a qualitative paper is it any worse than someone quoting that discussion in a blog? If anonymity is held then the ethical principle of privacy has been upheld.
I do however see where there may be issues. If someone was posting on a support group discussion about an illness they have, this may contain some very personal and sensitive information and it could be suggested that it is morally wrong to almost expose that to people who would not normally read that information. Yes, the website is public but perhaps people go on there under the assumption that the only people who will read it are people who understand them and share similar personal experiences, not someone who is out to take their words apart and analyse them for scientific use.
This is a subject matter which is difficult to answer. Perhaps if information is taken as a whole rather than individual quotes and examples then privacy wont be invaded as much and less harm is being done because the research only talks about general themes around what they have read.
A way around this problem is to open up specific forums where the researcher could ask people if they will willingly talk about what they want to know. This way the person knows exactly what their words are being used for. The internet is an amazing tool, with a wealth of free and easy to get to information, and this can be so beneficial for research as suddenly we are reaching people we probably wouldn’t be able to, and how amazing is that! It is important to discuss issues such as this though so that we have the freedom to use the internet without fear of our privacy being abused and our information being misused. I do however feel that public internet data, if used in an appropriate manner then it could really revolutionise how efficiently we can collect data.
Comments for my TA (Week 8/9)
Dearest Joe, here are my comments for your viewing pleasure.. don’t get too excited. Maybe get drunk then you’ll think they’re better!?
Is the Qualitative method less scientific than Quantitative?
What makes a research finding important? (H/W for week 8/9)
Research comes in many different shapes and sizes from lab experiments to field studies. This makes it difficult to make a direct comparison of one finding to another because they can be as unique as you and I (represented in this rather awesome picture…)
I do however think that all research findings should have some key requirements before we can begin to ask whether they are important or not. ‘Good’ research findings should be valid, so they measure what you intend it to, and reliable, so you could repeat the study over and over and you would be able to replicate the results.
Now we’ve established our good research perhaps we should consider whether it can be generalised to the wider population. It could be argued that if you cannot relate the data to the general public or a particular proportion of society then it is of no use at all. I believe in most cases this is very true, although you could have a case study of someone with a specific brain lesion and these findings can only be used for a few others with that lesion. I think it would be ridiculous to say that because you cannot generalise that to a large group of people it is unimportant. The research may be relevant to a handful of people but if it benefits them then the research has to be defined as important.
I believe research is important if it has any form of real world implication. If a study helps us to understand a type of behaviour better, or shows us what a particular neuron in the brain does, or helps us build up a knowledge of the past then it is important. I found a study by Sandler and Boat where they have started finding some links between parental death and the development of mental disorders amongst other possible causes. This research has huge implications in the real world as it can prepare paediatricians to look out for these problems developing and it could lead to earlier diagnosis and treatment. (http://psychologydegreeguide.org/recent-psychology-research/)
I don’t think however, that research has to have such huge implications as this or necessarily have to be ‘ground breaking’ because little bits of knowledge from lots of studies can accumulate to one finding. So in the wise words of tesco… every little helps!
When people are conducting research they may often find that their results refute their hypothesis or that the results are statistically insignificant. Is a result only worth something if you predicted it or if its significant? I don’t believe this is the case at all. Insignificant results may refute really well known theories or paradigms and this is incredibly important because it may mean that others have been getting it wrong and further research is needed. Findings that also refute a hypothesis may also lead the researcher into looking at other areas where they may not have thought to look at before. An undergrad posted this in a forum and I thought it highlighted my point well…
|
“That said, even if nothing comes back significant there is still much that can be taken away from this. In fact, the lack of sig. will actually make me think more about and question some long-held beliefs I’ve had about this topic, which I think is a good thing.”
|
http://forums.studentdoctor.net/showthread.php?t=679619
I believe qualitative research also highlights the fact that a significant t-test isn’t the be all and end all of an important finding because sometimes this type of research won’t look at figures, but rather find patterns or look for unusual behaviours to open them up to more reseach. The qualitative research may not have provided an exact answer but it can provide rich and very in-depth data which can be used as a guide for further research and what to look for. It acts as a clue to what could be a huge discovery which I think is very important indeed!
I would have to say that most research is important but there may be a degree of importance. Some provide us with tiny little clues and others change our entire thinking on a subject. It could also be argued that the value of importance is subjective to a person. Is research only important to the people it directly affects? As long as research is properly conducted and reliability and validity have been ensured then it is likely that your research will play some kind of role in the further understanding of humans and their minds and their behaviours and what’s not important about that!
ItsDefinitelyCaela 