The whole truth and nothing but truth On data falsifica8on, fabrica8on, and enhancement. Ramy K. Aziz

Transcription

1 The whole truth and nothing but truth On data falsifica8on, fabrica8on, and enhancement Ramy K. Aziz

2 Intended Learning Outcomes (ILOs) At the end of this module, each aeendee will be able to: 1. Define the acceptable rules of the a4endee s specific discipline regarding data repor9ng and manipula9on. 2. Explain the ethical problems and consequences of data fabrica9on and falsifica9on. 3. Iden9fy when data manipula9on is considered unethical. 4. Assess a figure or chart and decide whether it might have been falsified or manipulated. 5. List ways to prove data authen9city. 6. Recognize the common norms of discipline- specific data deposi9on and repor9ng.

3 Prelude: Some ethical golden rules Do to others as you would like them to do to you استفت قلبك When in doubt, follow your heart (Unless you have OCD), if you feel it s wrong, then there s something probably wrong about it!

4 Prelude: The hard ques8on Why?

5 What is truth? Before ANer

6 Can you do this to your data? h4p:// extreme- makeovers- that- are- way- too- much- unbelievable/ h4p://hotarabicmusic.com/2014/11/the- plas9c- arab- singers- before- and

7 Show me the data The currency and wealth of scholars! h4ps://figshare.com/blog/ PIs_What_happens_to_your_gradstudent_and_postdoc_s_work_when_they_leave_/111 h4ps:// dui/highest- reported- bac i- call- shenanigans/

8 Show me the data The currency and wealth of scholars! أسمع كلامك أصد قك أشوف الداتا أستعج ب! Ac8vity #1: Brainstorming/group sharing: What do different disciplines consider as the norms for transparent, honest data repor9ng

9 ACTIVITY #2: POLL!

10 Situa8ons (Ethical/ Unethical/ Gray area?) You showed your graphs to your supervisor, whose facial expressions always betray him. His face looked very much like a big tomato when he looked at the differences between your case and control averages. He controlled his voice asking you to re- plot the figure using a logarithmic scale so that the differences are clearer. Anything wrong about this?

11 Situa8ons (Ethical/ Unethical/ Gray area?) Before you prepared your histology images, your senior labmate strongly advised you to crop the figures, focusing only on the cells that changed. He asked you to make sure you crop your test and control in the same way. You felt that this may not be right. Is it?

12 Situa8ons (Ethical/ Unethical/ Gray area?) You ran a polymerase chain reac9on (PCR), and then put all your results in a gel. Aaer loading the gel, you realized that you forgot to add a nega9ve control lane, in which you run the reac9on on water, to make sure that there are no non- specific products. Your supervisor has warned you many 9mes, and you realized that she won t tolerate it this 9me without yelling at you in the lab mee9ng. You figured out a white lie that won t hurt anybody: you labeled the first lane of your gel as nega9ve control. You felt totally right about it because one of the reac9ons was nega9ve anyway, and showed no non- specific products so there is no scien9fic falsifica9on here.

13 Situa8ons (Ethical/ Unethical/ Gray area?) Everybody does it You finished your study about the link between palmar hyperhidrosis (excessive swea9ng of the hands) and successful dance career. Unfortunately, you could only find 3 current or past dancers who had the condi9on and who accepted to par9cipate in the study vs. 7 who did not have the condi9on and who also accepted to par9cipate. Your data are solid, but the sta9s9cs on 10 subjects are poor. Since your ra9os will remain the same, without hesita9on you wrote the study as if it s done on 30:70 subjects, leading to the same percentages. Everybody does it, aaer all! (Coincidentally related story: h4p:// /Dancer- suffering- profuse- swea9ng- disorder- holds- hands- boyfriend- 9me- 22- thanks- opera9on.html)

14 Image manipula8on

15 Image manipula8on

16 Image manipula8on

17 Image manipula8on

18 Ac8vity #3: Specific figures/charts à Are they genuine or manipulated? (Covered by Mohammed Yahia)

19 Take Home Messages Data are the bu4er and bread of scien9sts and scholars. Data are their currency and source of wealth. Data should be kept and presented fairly and accurately. Always abide by the three virtues: honesty, clarity, and transparency. Data visualiza9on is an art and science. It either conveys or kills your message. Some9mes, less is more. There is nothing wrong in showing a part of the picture to emphasize/focus, unless omission is inten9onal or misleading.

20 Take Home Messages If you feel it ain t right, it s probably not right, don t do it! If in doubt about it, ask! If s8ll in doubt about it, declare it (adhere to transparency)!