Friday, 20 August 2010

Suffering from Information Overload... BING.

The Microsoft advert for it's Bing search engine has been stuck in my head for the past few weeks now. Do you like mustaches? As someone that fell head over heels in love with Apple products about a year ago I was slightly reluctant to go back and try something that Microsoft has created. The persistent marketing tells me that, Bing promises to make search results easier to read by being 'visually organised' - I am presuming that is a fancy term for 'uses pictures', but I wouldn't know, I haven't used the system yet. I generally use Google and Pubmed to search (along with a few other sciency search engines).

This morning I have been dealing with/suffering from information overload (apparently soon to become a medical term, 'information fatigue syndrome'). I am trying to figure out a puzzle in my PhD project and I am attempting to do this by going through the literature. It's tough going. A search produces 1000s of results and the information thrown back at you can vary from being relevant to being useless. Trawling through endless papers to try and decide a) if it makes sense b)if the paper is actually talking about what you want c) if it's credible and useful. I just did a quick google search just now about 'searching for scientific papers' helpfully it threw up this result from the New Scientist stating 'most scientific papers are probably wrong', they most certainly nearly all are out of date - see my last post on e-lab books for my thoughts on making scientific information available instantly.

So, is a simple search on Bing going to change my life? Will I find all the information from the search clearly laid out and will it be obvious which site/sites I need to visit? I am going to try Bing with a search for 'number of scientific publications per year'.... here goes..... HMM, disappointing. The results page looks similar to a google search... I won't go through what results I did get, let's just say I didn't get what I was hoping for, which was a nice graph showing the number of papers published yearly per journal/country/region. Maybe this information doesn't exist, but I am sure it does. I would try searching for images, but I can't seem to find this option on Bing. Maybe I am being simple, but I can't see why I should use this over Google. I did the same search on Google, for comparison purposes didn't find a graph there either, however I did find this, from http://www.americanscientist.org/my_amsci/restricted.aspx?act=pdf&id=3263000957901

Showing that the number of publications per year is increasing (up until 1998). I presume that the same trend is going on (I will look further into this).

If the amount of information just keeps going up and up and up how on earth are we supposed to find out what we need and prevent duplicating work that someone has already done?! I really can't see what solutions there are to this, other than spending increased amount of time researching topics and trying to keep on top of what is going on in the world. For now, here are a few tips that might help:


1) Start big, if researching a new topic perform general searches with a couple of key words to get an overview of what is current/popular about the topic at that time. DO NOT STOP THERE THOUGH!

2) Learn how to use search engines properly, especially scientific ones, then you can use advance search options and narrow down the number of results you get and you get results that are more specific to your needs.

3) Learn to scan read scientific papers so you can quickly identify if the paper will be useful or not.

4) File properly, save files/links appropriately in a system you can use simply, this will save a lot of time and effort later when you suddenly think 'ah I read that somewhere, I think, now how did I find that... where will it be'

5) Make notes on things you have read and link to where you read it, this helps and can help you prevent doing dupicate information searches six months down the line

6) STAY UP TO DATE. Make sure you sign up for updates from pubmed etc and regularly (weekly) do searches on your topic so you are sure that the information you have is the most current.



By the way, I am dubious about 'information fatigue syndrome' as a medical term. All I know that trying to find information is increasingly becoming a pain in the bum due to the the fact that the amount of information we have is increasing. It makes the mind boggle. 

Friday, 13 August 2010

Science - all about the new, in an old fashioned way.

Scientists work on the unknown; they are at the forefront of knowledge. They know what is new in technology, engineering and medicine before anyone else. Ironically, the way scientists record their information is firmly stuck in the dark ages, they hand write stuff, with pen and paper. The record of the experiments they carry out is contained in handwritten lab books. Hardly anyone still works with paper and ink anymore, is this an example of where something that isn’t broken shouldn’t be fixed, or can technology help make life easier?

Filling out a lab book is annoying. Just writing one basic experiment can involve 2/3 pages of handwritten notes, 3 trips to the printer and a lot of cutting and sticking. As most experiments are repetitions of previous ones the methods are the same but the methods still need to be written out by hand in the lab book. For the majority of experiments results are revealed through a computer and the graphs created from results are also created by a computer. To record the results from the computer you have to print them out, cut and stick in your lab book. It's a laborious job. I sometimes feel a bit like a primary school student with all the cutting and sticking! It doesn’t help that the scissors I use are actually from a primary school!

Accurate recording of results, methods and thoughts is priceless. Lab books are actual legal documents, they can be called upon in court to help resolve patent disputes and plagiarism disputes. They are also invaluable in that they help you remember what you did 6 months ago! Sadly many lab books are not up to scratch, a lot of them are illegible and experimental methods are badly recorded. Making the process of recording experiments easier and simpler would save time and effort and potentially would make people record their work a little better. Some people do have extremely neat, fully completed clear lab books. I would say mine lies somewhere in between super neat and a big mess, it depends how busy I am and how much effort I can be bothered putting into it. Think about being at school and writing in exercise books, some peoples are super neat, others are a scrawling mess that only they can understand - this is what lab books are like.


These days nearly everything is done electronically, so why can’t science catch up with the times and use an electronic lab book to record lab experiments and information?  Surely other people have thought of this before me, I know many people that grumble about having to update their lab book. Data could be recorded easily and simply in an e-lab book, results and images from experiments could be directly inputted from the system/software used for the experiment, saving time and effort. Methods could be copied and pasted and altered as required - typed text is always legible, it beats handwriting anyday - this could make the court cases easier! The e-lab book could be date controlled, every time a entry is inputted the lab book could automatically record the date. Any edits to the information could be recorded on the system, so the opportunity to fudge results is reduced and the e-lab books would be as fool proof as hand writing the results/ideas. Another major advantage is one to the wider world of science, E-lab books could be made available on the Internet - no waiting 6/12 months to publish a paper the newest data would be accessible to all, this could potentially speed up science and research and reduce repetition across labs. When a research paper is produced, it could link back to the e-lab books where the experiments were carried out so the original data is available. This would be another great step forward, interactive research papers would be more useful than the standard ones we have today and also far more exciting!

So why aren't we there already?
There are of course disadvantages to using e-recording.Technology can fail, I think nearly everyone that has completed a PhD in the last few years has gone through the drama of losing some results/PhD thesis through system failures. In these cases the paper copies of their results become extremely important. Until e-recording is absolutely fail safe then there is a huge risk here. Opportunities for altering results are potentially higher if everything is recorded electronically, however I think if someone is set out to fudge results they could do that through the paper recording method too. Having a date monitor on the e-lab book possibly could help prevent results being altered. Ultimately, until industry decides that what they want and need is e-lab books, it won't happen. I very much doubt that in this economic climate academia will lead the way, but you never know. 

A change in the way people work is always difficult, people mostly dislike change, even scientists that are changing the way we look at the world through their discoveries are still creatures of habit. Some experiments are still recorded by hand, I know I have to count cell numbers by eye and hand - so I would need a way of transporting my computer to the microscope room (4 floors below me) in order to input the data. It’s not so easy to type when peering down a microscope; however, it is relatively easy to use a pen. I do own an iPhone, I could quite easily record my results where-ever I am through my e-lab book on my iPhone. Not everyone has a smart phone at the moment, which brings me on to the next barrier, cost. Technology is expensive; I imagine an e-lab book system would cost a lot more than rolling out paper lab books. I can understand why academia isn't pushing for this option. Other industries with a bit more money to play with could benefit from e-lab books, the pharmaceutical industry is regularly involved in patent disputes where lab books are brought up in court, accurate and clear e-recording of information could potentially make this process simpler and the time saved would make their scientists more productive.
Regardless of what I posted above, sometimes it’s quite nice to be reminded of what writing by hand is like, the opportunity to hand write doesn’t come very often. Writing by hand makes you think before you write, you cannot delete something in ink on paper. In this age where information and thoughts are batted around 2 a penny on the Internet as soon as they appear in someone’s brain, it can be quite nice to write something that you have thought about a lot before hand on paper, especially when it is random notes containing your thoughts, ideas and theories on a subject. I don't think you would get as many ridiculous and offensive tweets if people had to go to the trouble of handwriting them. Writing on a computer/iPhone does not give you the same feeling; it's like writing a letter, old fashioned, but nice.

Wednesday, 4 August 2010

Conflicts of Interest - What Journalists can Learn from Research

There is a report today in the Guardian covering the 'news' that various PR groups are changing images of areas (such as the country Rwanda) by giving journalists huge freebies (such as holidays to Rwanda) and then getting them to write about it. Of course they are. PR companies are experts in changing the opinions of the general public on people/places/organisations - that is what they are paid to do. I do not have a problem with this, PR is a good thing, it can change lives and educate people on certain topics.  The problem with this sponsored article writing is clear, only one view is being represented by the journalists involved as they are being 'paid' (through the freebies) to represent the people/place/organisation. This leads to an unbalanced article being reported - the journalist more than likely wouldn't have come to the same conclusions if they had been simply told by their boss to 'go and write a report on Rwanda'. Further, it leads to inaccuracies being reported, the reporters are reporting what they are told by the groups they are set up with by the PR company (the Guardian article nicely explains how inaccurate reports on corruption in Rwanda were reported).

The other side of the argument is that without the push from the PR company there wouldn't have been an article on Rwanda at all - so the PR group is raising awareness of the area, and if people are REALLY interested in Rwanda then can then go out and search about Rwanda themselves - however, people only do this if they are REALLY interested - a lot of people will take the article written by the respected journalist in the respected newspaper to be fact. The reader will not know that the report they read was sponsored by a PR company paid by the Rwandan President. The problem here is freedom of information and conflicts of interest.

All research papers have to declare any conflicts of interest, this is standard. Sometimes carrying out research with a conflict of interest is unavoidable and is not necessarily frowned upon. The most important point is the transparency of the report. Why can't this be introduced for journalists? Why isn't it in practice already? I see no barrier as why this shouldn't become standard on informative news articles. It would also help with science journalism, where 'research' is often presented to the public (e.g. a report on walnuts preventing cancer- see earlier blog) seems innocuous enough, however the research actually has a number of flaws and the research only got into the newspaper through a press release which was a result of the research being sponsored by the Californian Walnut Board. The original research paper would state that the research was sponsored by the Californian Walnut Board. No problems there, people need sponsors for research and anyone interested in the information could read the paper (with all the facts) see the flaws and see that the research was sponsored, they can then make up their own mind about the validity of the research.  If you simply read the news article about the research, none of that information is presented (neither is a link to the original research, which also should happen in any science journalism). In order for people to make a balanced judgement on an article or piece of information they need to have the full background, reasoning and sources for the article. This is impossible if information is withheld, like in many cases of science research and PR sponsored holiday guides

So PR companies, you are doing a good job - keep promoting what ever you are asked to do! Journalists - make sure your reports are clear, informative and entertaining but also make sure you declare if you have been sponsored to do the work, i.e. in the introduction state that you were asked to visit Rwanda by xxxx OR have box at the bottom of the article that declares any possible conflicts of interest. I can't see why that would be a problem and also, it would probably make journalists investigate their facts a little better - and therefore give the public better, more informative articles. It's a simple change in the way things are done, that could make a huge amount of difference - and I expect credible journalists would not have any problem with it - much like any respected researcher.

Contributors