"Biggest fraud in physics in the last 50 years..."

This example of scientific fraud quite surprised me. I've never thought that it can be so big,
"successful" and last so much time...

I'm talking about German physicist Jan Hendrik Schön, who was famous in the field of semiconductors, until he was caught. He claimed to make breakthroughs in semiconductors, which could have change completely the industry. Before he was caught he received several awards as young successful researcher with a bright future. Of course, later all rewards were revoked.

In his papers he claimed that he was able to use organic materials instead of silicon and other conventional materials. Later he also declared that he built a transistor on the molecular scale, using thin layer of organic molecules. This was a huge breakthrough with grand opportunities of building organic electric circuits. Unfortunately, all of it was a lie.

How he was caught? Of course many researcher were trying to repeat his "extraordinary" experiments and obtain the same results. But what a surprise - none of them could do it. This attracted additional attention to his papers, and finally Lydia Sohn, the researcher from Princeton University, pointed out that in his graphics the noise is identical for different temperatures. He explained by mistakenly submitting the same graph twice. But it was too late.

After investigation 16 of 24 misconducts were proved. Very convenient for him he didn't use any notebooks and all electronic copies were erased due to a small space in hard drive of his computer. Poor guy, he had to publish results and then erase them from computer...

More interestingly, all coauthors were freed from the scientific misconduct. This started a discussions of responsibility of coauthors and reviewers.

He finished his career with a such shame. After all investigations and proves of his scientific frauds, his doctoral degree was revoked and he lost rights to be peer reviewer and apply for funds in 8 years.

And now I'm thinking how he was going to explain that no one could repeat his experiments? What he was thinking of? Also how he could last so long and publish 36 papers in such famous scientific journals like Nature, APL, Physical Review, Science...

Why we put the references?

To be honest, I don't have more reasons than we have said in the class... Namely:
1. To show your erudition.
2. Avoid plagiarism, divide your work and work of others.
3. Following from the previous: divide your current work and your previous work. State what you have improved, why it's important.
4. Show respect to people you like (contrary, you can do not put a reference of work of person you do not like).
5. Demonstrate your collaboration with someone, again, very close to the previous one.
6. To help the reader to understand your work, by recommending to read more about this topic.
7. Avoid repeating, make the text more laconic, without excessive sentences.
8. Following from 6-7, explain your idea by making analogies with previous works.

I think that's it. I can't think another reasons to put references. These ones I used by myself, when I was writing the paper. Maybe I forgot about some reason I used.

Bad graph, good graph...

Here are the examples I found. Nothing special, standard mistakes in graph design. Let's start with an example of bad graph.

Of course, it's only my opinion, maybe you will think contrarily. But why I think this is a bad graph? Because it is quite difficult for me to understand. In my opinion, this is the first and main criteria for an illustration. Yeah, after 2 or 3 minutes it became more clear, what is going on, but still I had a strange feeling that something left unclear. For example, this diversity of colors, mixing in the center not helping, not at all... They are just creating a chaos in my mind :). Besides, some information is impossible, well not impossible, but hard to obtain from this graph - what is the distribution of children who changed their religion.

The good graph:

Obviously it's a different type of graph and cannot be compared directly, but I like that it's clean, easy to understand. The scale is the same for whole graph, the size of letters is appropriate. The space is used well, without blank regions. 

So, these are my examples of good and bad illustration. 

Number of Journals vs Impact Factor in Biology

Here is the graphic, which illustrates the number of Journals as a function of theirs Impact factor.
Basically, it clearly demonstrates that we discussed in the class last Tuesday. Tens of Journals have IF less than 1 and only two have more than 10.
Let's see how it is in the other fields...

Reference manager

I have already written a little bit about the reference manager I use, but I will add some thing I like in Mendeley. Mendeley as you know is a free reference manager, which allows you to add your articles, articles you read and cite. As you are adding new articles, books into to your collection, you can create different folders, groups for articles, depending on their topic. I found this very useful when your library starts to grow - you have more ways to organize everything. Next time you want to find some article and you remember only the topic and name of the author you don't have to scan all papers of this author, you can just search in this folder.
Another feature I love is the citation plugin in Microsoft word. By typing the name of the author and choosing from the list of his papers you can create your reference list in just a few minutes... Also, it will be already formatted according to the standards of the scientific journal(although not all them are presented in Mendeley, but you can create your own template).
What else? Probably I will start to share my collection to my colleagues, because sometimes you need to share some paper, or tell them what interesting paper you are reading now.
So, I don't know if I will change the reference manager in future, but for this moment I'm totally happy with Mendeley.