n*********n 发帖数: 95 | 1 好几年前在这里copy了这个Tips for your experiments.因为很喜欢,自己重新输入编
排了一遍,贴在实验室里,毕业后多年,收到老板的信问我这个的出处,因为他把这个
Tips放到Twitter上,立马有好几百人点赞。我也不记得了,只记得是从这个版得来的
。如果有人记得或认得,请给我留言,先谢了!
TIPS for your experiment
1. If the protocol works, don’t change it (If it isn’t broke, don’t
fix it).
2. If the protocol doesn’t work, change it.
3. If a protocol requires a laborious step, there is usually a reason for
it. Don’t try to avoid it (You are not as smart as you think).
4. Two hours of work today will save you two days of work tomorrow.
5. Every new experiment needs a positive control and a negative control.
6. There are three types of experiments:
• One which yields interesting results, no matter the outcome (Type
I),
• One which yields an interesting result, but only if it turns out
one way (Type II),
• And one which yields ambiguous, uninterpretable results
regardless of the outcome (Type III), Take your pick.
7. Label every tube and record every experiment.
8. Assume nothing. A corollary of this is that opinions don’t matter,
only data does.
9. Experiments often don’t work. Consider all the possible reason why an
experiments fails and concentrate on the likeliest causes;
10. Complicated, multi-step experiments never succeed unless they are
simplified. Break up each experiment into its components, work on one at a
time, and verity that each component works before proceeding on to the next
(Take one step at a time).
11. Don’t try to recreate the wheel. If someone else has successfully
performed the same experiment as you plan to do, talk to her.
12. Plan every experiment as if it will become a figure in a manuscript.
13. If someone asks for your help, be as helpful as you can (What comes
around, goes around).
14. It is more important to be accurate than prolific. It’s also helpful
to be prolific.
15. If an experiment is not hypothesis-driven, then it can prove nothing.
16. What matters in science is not what can possibly occur, but what
actually does occur.
17. Read widely. The most creative insights come from applying knowledge
from outside one’s field. | p********r 发帖数: 4210 | | d********m 发帖数: 3662 | | D****g 发帖数: 275 | 4 google "GREENBERG’S LAB TIPS 实验生物学圣经" | n*********n 发帖数: 95 | | s*****j 发帖数: 6435 | 6 15 是扯淡。
for
【在 n*********n 的大作中提到】 : 好几年前在这里copy了这个Tips for your experiments.因为很喜欢,自己重新输入编 : 排了一遍,贴在实验室里,毕业后多年,收到老板的信问我这个的出处,因为他把这个 : Tips放到Twitter上,立马有好几百人点赞。我也不记得了,只记得是从这个版得来的 : 。如果有人记得或认得,请给我留言,先谢了! : TIPS for your experiment : 1. If the protocol works, don’t change it (If it isn’t broke, don’t : fix it). : 2. If the protocol doesn’t work, change it. : 3. If a protocol requires a laborious step, there is usually a reason for : it. Don’t try to avoid it (You are not as smart as you think).
| N***e 发帖数: 61 | 7 1和2.说明你不敢用多的条件试。
3.说明生物就是黑砖窑。
4.为什么今天多干两小时,就会少干两天。明明就是今天多干两小时,做出来了,明天
活更多,做不出来,明天接着来。
5.reviewer总是觉得你的control不够,你还敢用a?control要加复数。
6.这就是proposal里扯的蛋。你觉得你说的是type I,其实都是type III。
7.你有没有发现实验室发Nature,Cell,Science的主儿,根本没有多少试验记录,连
数据都没留?
8和15你不觉得自己矛盾么。
9.总算说了句实话。
10.你这么做了,老板觉得你能力不够,实验进度太慢。
11.你会发现,多半情况下,对方经不住三问,为啥?参考1和2。
12.参考9,你会发现你根本没有figure。
13.参考11。
14.这是多快好省(这个是个贬义词)的委婉说法。
15.我就是好奇结果,没有hypothesis不行吗?NIH grant写多了么?
16.你把这句话强调给你的committee members试试,他们不拍死你。
17.关键是你读的文献是真的才有用。这么多文章作假,修饰数据,真是越读越糊涂。 | h**c 发帖数: 1979 | 8 8和15你不觉得自己矛盾么。
哈哈哈,生物千老自欺欺人的本领太强了 |
|