Threats to Validity

In a rare moment of pristine clarity, my proffessor told us the other day, in a voice that said "of course, you all know this, but I am saying it anyway because I feel like being redundant" that there are 31 threats to the validity of an (psychological?) experiment.
THIRTY-ONE.
Seriously? For starters, I'd like to know who on earth goes around asking themselves "how many ways could my colleagues tell me that I'm wrong?" I mean, I know it's science, and the main way you know things is by figuring out how a conclusion could be wrong, but threats to validity generally don't just tell you you're wrong. They throw all your hard work out the window.
You don't even get to learn something from your wrongness (i.e., the reason books are so expensive is not because they pay their factory workers high wages, like I thought; now I should move on and make a new hypothesis). It's more like, "oh, you measured this badly, so you have to start all over again before we'll be able to know anything useful from your experiment". Wtf??
I'm not saying this threat to validity business isn't well-thought out and useful and stuff. Like my other professor said, the tools of science are there to keep us from being duped; it's a good thing to be able to say "this is true." as opposed to "this is what I think based on my experience and the anecdotes of these 5 other people". There is worth in having truth value to back up a conclusion. But sometimes I think that scientists are those people in life who are in danger of over-thinking just about everything.

For your amusement and my learning, I have therefore assembled a list of the 31* threats to validity Enjoy.

Construct validity

1. inadequate preoperational specification of constructs

2. mono-methods bias

3. mono-operation bias

4. hawthorne effect

5. experimenter expectancy (same as hawthorne effect. oops. I'm not gonna re-do the whole list tho.)

6. hypothesis guessing

7. experimenter reactivity

8. measurement reactivity

9. evaluation apprehension

Internal validity

10. ambiguity about the direction of causal influence

11. history

12. maturation

13. testing

14. mortality

15. instrumentation

16. selection

17. regression to the mean

18. selection-maturation interaction

19. interaction of selection with history

20. interaction of selection with mortality

21. diffusion or imitation of treatment

22. compensatory equalization

23. compensatory rivalry

24. resentful demoralization

statistical conclusion validity

25. experimentwise error

26. incorrect (statistical) assumptions

27. reliability of measures

28. low statistical power

29. reliability of treatment implementation

External validity

30. biased sampling

31. restricted setting

32. historical time

*On going through the prof's slides, I've realized there are actually 32 threats. Hmm... let's be science-y how many ways could I be wrong?
1. I could have called one of these things a threat when it isn't (which wouldn't be too surprising - it's REALLY hard to tell from her slides. Organization = not her best skill**).
2. Hawthorne effect and experimenter expectancies could be the same thing.***
3. She could have miscounted.

**On that note, the threats could be grouped/further classified according to:
- whether they apply more to survey research or experimental research
- applying when you reject the null
- applying when you fail to reject the null
- having to do with measurement response
- having to do with contact with the experimenter
- a few other groupings I've forgotten about because they weren't useful/were hard to understand

***On closer inspection, I think this is the answer.

Hope you enjoyed the list. I also hope it helped me learn the stupid stuff.
over 'n out.
E.O.