1
00:00:00,000 --> 00:00:02,411
Can week-old pizza cause 
psychedelic hallucinations?

2
00:00:02,411 --> 00:00:03,596
Does coffee makes you smarter

3
00:00:03,596 --> 00:00:05,658
or does it just make you 
do dumb stuff faster?

4
00:00:05,658 --> 00:00:06,690
Like much much of psychology itself,

5
00:00:06,690 --> 00:00:08,810
questions like this can seem pretty intuitive.

6
00:00:08,810 --> 00:00:11,571
I mean, people may not be the easiest 
organisms to understand, but...

7
00:00:11,571 --> 00:00:13,111
You're a person, right?

8
00:00:13,111 --> 00:00:16,891
So you must be qualified to draw, like,
some conclusions about other people,

9
00:00:16,891 --> 00:00:18,105
and what makes them tick.

10
00:00:18,105 --> 00:00:21,412
But it's important to realize that 
your intuition isn't always right.

11
00:00:21,412 --> 00:00:24,701
In fact, sometimes it is exactly wrong,

12
00:00:24,701 --> 00:00:28,195
and we tend to grossly underestimate 
the dangers of false intuition.

13
00:00:28,195 --> 00:00:31,277
If you have some idea about a person and
their behavior that turns out to be right,

14
00:00:31,277 --> 00:00:34,212
that reinforces your trust in your intuition.

15
00:00:34,212 --> 00:00:35,329
Like if I warned my buddy Bob

16
00:00:35,329 --> 00:00:38,332
against eating the deep-dish pizza 
that's been in the fridge for the past week,

17
00:00:38,332 --> 00:00:40,649
but he eats it anyway and
soon starts to wig out,

18
00:00:40,649 --> 00:00:43,219
I'm gonna say: "Dude, I told you so!"

19
00:00:43,219 --> 00:00:45,196
But if I'm wrong, and he's totally fine,

20
00:00:45,196 --> 00:00:48,473
I'll probably won't even
think about it ever again.

21
00:00:48,473 --> 00:00:52,445
This is known as hindsight bias, 
or the "I-knew-it-all-along" phenomenon.

22
00:00:52,445 --> 00:00:54,634
This doesn't mean that common sense is wrong,

23
00:00:54,634 --> 00:00:56,331
it just mean that our intuitive sense

24
00:00:56,331 --> 00:01:00,143
more easily describes what JUST happened 
than what WILL happen in the future.

25
00:01:00,143 --> 00:01:02,666
Another reason you can't 
blindly trust your intuition

26
00:01:02,666 --> 00:01:05,120
is your natural tendency toward overconfidence.

27
00:01:05,120 --> 00:01:09,461
Sometimes, you just really, really 
feel like you're right about people

28
00:01:09,461 --> 00:01:11,945
when actually, you're really, really wrong!

29
00:01:11,945 --> 00:01:13,112
We've all been there...

30
00:01:13,112 --> 00:01:15,654
We also tend to perceive order in random events,

31
00:01:15,654 --> 00:01:17,437
which can lead to false assumptions.

32
00:01:17,437 --> 00:01:19,526
For example, if you flip a coin five times,

33
00:01:19,526 --> 00:01:25,018
you have equal chances of getting all tails 
as you do getting alternating heads and tails,

34
00:01:25,018 --> 00:01:28,821
but we see the series of five tails 
as something unusual, as a streak,

35
00:01:28,821 --> 00:01:33,138
and thus, giving that result some kind of 
meaning that it very definitely does not have.

36
00:01:33,138 --> 00:01:35,432
That is why we have the methods and safeguards

37
00:01:35,432 --> 00:01:38,757
of psychological research and experimentation,

38
00:01:38,757 --> 00:01:41,533
and the glorious process of scientific inquiry.

39
00:01:41,533 --> 00:01:43,623
They help us to get around these problems

40
00:01:43,623 --> 00:01:48,184
and basically, save the study of our 
minds from the stupidity of our minds.

41
00:01:48,184 --> 00:01:52,523
So I hope that it won't be a spoiler if I 
tell you now that pizza won't make you trip,

42
00:01:52,523 --> 00:01:55,999
and coffee doesn't make you smart. Sorry.

43
00:01:55,999 --> 00:01:59,639
[on-screen animations and 
ribbons of science sentences]

44
00:02:03,679 --> 00:02:05,946
Title screen says "Episode 2: 
Research & Experimentation."

45
00:02:05,946 --> 00:02:07,683
In most ways, psychological research

46
00:02:07,683 --> 00:02:10,440
is no different than in the 
other scientific discipline.

47
00:02:10,440 --> 00:02:14,830
Like, step one is always figuring out how to 
ask general questions about your subject,

48
00:02:14,830 --> 00:02:18,420
and turn them into measurable, 
testable propositions.

49
00:02:18,420 --> 00:02:21,354
This is called "operationalizing" your questions.

50
00:02:21,354 --> 00:02:22,708
So you know how the scientific method works.

51
00:02:22,708 --> 00:02:24,965
It starts with a question and a theory.

52
00:02:24,965 --> 00:02:28,153
And I don't mean theory in the 
sense of like, a hunch that says

53
00:02:28,153 --> 00:02:30,489
"a quad-shot of espresso 
makes you think better."

54
00:02:30,489 --> 00:02:36,107
Instead, in science, a theory is what explains 
and organizes lots of different observations

55
00:02:36,107 --> 00:02:37,308
and predicts outcomes.

56
00:02:37,308 --> 00:02:39,383
And when you come up 
with a testable prediction,

57
00:02:39,383 --> 00:02:40,788
that's your hypothesis.

58
00:02:40,788 --> 00:02:42,771
Once your theory and hypothesis are in place

59
00:02:42,771 --> 00:02:45,675
you need a clear and common 
language to report them with.

60
00:02:45,675 --> 00:02:47,702
So, for example,
defining exactly what you mean

61
00:02:47,702 --> 00:02:50,204
by "thinking better" with 
your espresso hypothesis

62
00:02:50,204 --> 00:02:53,107
would allow other researchers 
to replicate the experiment.

63
00:02:53,107 --> 00:02:55,152
And replication is key.

64
00:02:55,152 --> 00:02:57,722
You can watch a person exhibit 
a certain behavior once,

65
00:02:57,722 --> 00:02:59,514
and it won't prove very much.

66
00:02:59,514 --> 00:03:03,641
But if you keep getting consistent results 
even as you change subjects or situations,

67
00:03:03,641 --> 00:03:05,117
you're probably onto something.

68
00:03:05,117 --> 00:03:08,559
This is a problem with one popular
type of psychological research:

69
00:03:08,559 --> 00:03:11,972
case studies, which take an 
in-depth look at one individual.

70
00:03:11,972 --> 00:03:13,976
Case studies can sometimes be misleading,

71
00:03:13,976 --> 00:03:16,432
because by their nature, 
they can't be replicated;

72
00:03:16,432 --> 00:03:18,428
so, they run the risk of over-generalizing.

73
00:03:18,428 --> 00:03:20,974
Still, they're good at showing 
us what CAN happen,

74
00:03:20,974 --> 00:03:24,822
and end up framing questions for more 
extensive and generalizable studies.

75
00:03:24,822 --> 00:03:26,301
They're also often memorable

76
00:03:26,301 --> 00:03:30,384
and a great story-telling device psychologists
use to observe and describe behavior.

77
00:03:30,384 --> 00:03:34,494
Like, say, the smell of coffee makes 
Carl suddenly anxious and irritable.

78
00:03:34,494 --> 00:03:37,358
That obviously doesn't mean that it 
has the same effect on everyone.

79
00:03:37,358 --> 00:03:41,179
In fact, Carl has terrible memories
associated with that smell,

80
00:03:41,179 --> 00:03:43,277
and so his case is actually quite rare.

81
00:03:43,277 --> 00:03:44,032
Poor Carl... :(

82
00:03:44,032 --> 00:03:45,808
But, you will still have to look at lots of

83
00:03:45,808 --> 00:03:47,895
other cases to determine that conclusively.

84
00:03:47,895 --> 00:03:51,685
Another popular method of psychological 
research is naturalistic observation,

85
00:03:51,685 --> 00:03:55,493
where researchers simply watch 
behavior in a natural environment,

86
00:03:55,493 --> 00:03:57,912
whether that's chimps poking 
anthills in the jungle,

87
00:03:57,912 --> 00:04:01,160
kids clowning in a classroom, 
or drunk dudes yelling at soccer games.

88
00:04:01,160 --> 00:04:03,852
The idea is to let the subjects 
just "do their thing"

89
00:04:03,852 --> 00:04:06,451
without trying to manipulate 
or control the situation.

90
00:04:06,451 --> 00:04:08,732
So yeah, basically just spying on people.

91
00:04:08,732 --> 00:04:12,456
Like case studies, naturalistic observations 
are great at describing behavior,

92
00:04:12,456 --> 00:04:14,819
but they're very limited in explaining it.

93
00:04:14,819 --> 00:04:18,435
Psychologists can also collect behavioral
data using surveys or interviews,

94
00:04:18,435 --> 00:04:21,379
asking people to report 
their opinions and behaviors.

95
00:04:21,379 --> 00:04:24,426
Sexuality researcher Alfred Kinsey
famously used this technique

96
00:04:24,426 --> 00:04:27,580
when he surveyed thousands of men 
and women on their sexual history

97
00:04:27,580 --> 00:04:28,914
and published his findings.

98
00:04:28,914 --> 00:04:32,668
in a pair of revolutionary texts: 
"Sexual Behavior in the Human Male"

99
00:04:32,668 --> 00:04:34,834
and "Sexual Behavior in the Human Female."

100
00:04:34,834 --> 00:04:36,852
Surveys are a great way to access people's

101
00:04:36,852 --> 00:04:38,760
consciously held attitudes and beliefs,

102
00:04:38,760 --> 00:04:43,498
but how to ask the questions can be tricky;
subtle word choices can influence results.

103
00:04:43,498 --> 00:04:46,552
For example, more forceful 
words like "ban" or "censor"

104
00:04:46,552 --> 00:04:49,606
may elicit different reactions
than "limit" or "not allow."

105
00:04:49,606 --> 00:04:53,261
Asking: "Do you believe in space aliens?" 
is a much different question than

106
00:04:53,261 --> 00:04:56,080
"Do you think that there is intelligent 
life somewhere else in the universe?"

107
00:04:56,080 --> 00:04:59,341
It's the same question, but in the first, 
the subject may assume that you mean

108
00:04:59,341 --> 00:05:01,794
"aliens visiting the Earth 
and making crop circles

109
00:05:01,794 --> 00:05:03,697
and abducting people and poking them."

110
00:05:03,697 --> 00:05:05,677
And if how you phrase surveys is important,

111
00:05:05,677 --> 00:05:07,332
so is who you ask.

112
00:05:07,332 --> 00:05:10,543
I could ask a room full of students at a pacifist 
club what they think about arms control,

113
00:05:10,543 --> 00:05:13,896
but the results wouldn't be a representative 
measure of where the students stand,

114
00:05:13,896 --> 00:05:16,640
because there's a pretty clear
sampling bias at work here.

115
00:05:16,640 --> 00:05:20,125
To fairly represent a population,
I'd need to give a random sample

116
00:05:20,125 --> 00:05:21,973
where all members of the target group

117
00:05:21,973 --> 00:05:25,995
(in this case, students) had an equal chance
of being selected to answer the question.

118
00:05:25,995 --> 00:05:27,905
So, once you've described behavior

119
00:05:27,905 --> 00:05:30,496
with surveys, case studies, 
or naturalistic observation,

120
00:05:30,496 --> 00:05:34,357
you can start making sense out of it 
and even predict future behavior.

121
00:05:34,357 --> 00:05:35,568
One way to do that is to look at

122
00:05:35,568 --> 00:05:39,462
how one trait or behavior is related 
to another or how they correlate.

123
00:05:39,462 --> 00:05:41,120
So let's get back to my buddy Bob,

124
00:05:41,120 --> 00:05:44,419
who seems to think that his refrigerator
is actually some kind of time machine

125
00:05:44,419 --> 00:05:46,241
that can preserve food indefinitely.

126
00:05:46,241 --> 00:05:48,962
Let's say that Bob is just tucked into 
a lunch of questionable leftovers...

127
00:05:48,962 --> 00:05:51,687
Pizza that may very well have 
had a little bit of fungus on it...

128
00:05:51,687 --> 00:05:55,402
But he was hungry. And lazy.
And so he doused it in sriracha.

129
00:05:55,402 --> 00:05:57,342
Suddenly, he starts seeing things.

130
00:05:57,342 --> 00:05:59,520
Green armadillos with laser-beam-eyes.

131
00:05:59,520 --> 00:06:02,910
From here we can deduce that eating 
unknown fungus predicts hallucination.

132
00:06:02,910 --> 00:06:06,617
That's a correlation; 
but correlation is not causation.

133
00:06:06,617 --> 00:06:10,758
Yes, it makes sense that eating questionable 
fungus would cause hallucinations,

134
00:06:10,758 --> 00:06:14,223
but it's possible that Bob was already 
on the verge of a psychotic episode

135
00:06:14,223 --> 00:06:16,691
and those fuzzy left-overs were actually benign!

136
00:06:16,691 --> 00:06:19,689
Or, they could be an entirely 
different factor involved,

137
00:06:19,689 --> 00:06:21,658
like maybe he hadn't slept in 72 hours

138
00:06:21,658 --> 00:06:23,799
or had an intense migraine coming on,

139
00:06:23,799 --> 00:06:26,279
and one of those factors 
caused his hallucinations.

140
00:06:26,279 --> 00:06:28,369
It's tempting to draw 
conclusions from correlations,

141
00:06:28,369 --> 00:06:30,114
but it's super important to remember

142
00:06:30,114 --> 00:06:34,159
that correlations predict the POSSIBILITY 
of a cause-and-effect relationships;

143
00:06:34,159 --> 00:06:35,821
they can not prove them.

144
00:06:35,821 --> 00:06:38,827
So we've talked about how to describe 
behavior without manipulating it,

145
00:06:38,827 --> 00:06:41,959
and how to make connections and 
predictions from those findings,

146
00:06:41,959 --> 00:06:43,595
but that can only take you so far.

147
00:06:43,595 --> 00:06:46,445
To really get to the bottom of 
cause-and-effect behaviors,

148
00:06:46,445 --> 00:06:48,323
you're gonna have to start experimenting.

149
00:06:48,323 --> 00:06:51,184
Experiments allow investigators
to isolate different effects

150
00:06:51,184 --> 00:06:53,494
by manipulating an independent variable

151
00:06:53,494 --> 00:06:57,214
and keeping all other variables constant
(or as constant as you can).

152
00:06:57,214 --> 00:06:59,643
This means that they need at least two groups:

153
00:06:59,643 --> 00:07:02,354
the experimental group, which is 
gonna get "messed with";

154
00:07:02,354 --> 00:07:04,816
and the control group, which is 
not going to get "messed with".

155
00:07:04,816 --> 00:07:06,547
Just as surveys use random samples,

156
00:07:06,547 --> 00:07:09,813
experimental researchers need to 
randomly assign participants to each group

157
00:07:09,813 --> 00:07:14,482
to minimize potential confounding variables 
or outside factors that may skew the results.

158
00:07:14,482 --> 00:07:16,794
You don't want all grumpy teenagers in one group

159
00:07:16,794 --> 00:07:19,047
and wealthy Japanese servers
in the other; they gotta mingle.

160
00:07:19,047 --> 00:07:24,185
Sometimes one or both groups are not
informed about what's actually being tested.

161
00:07:24,185 --> 00:07:27,285
For example, researchers can test
how substances affect people

162
00:07:27,285 --> 00:07:30,027
by comparing their effects to
placebos, or inert substances.

163
00:07:30,027 --> 00:07:32,453
And often, the researchers themselves

164
00:07:32,453 --> 00:07:35,164
don't know which group is experimental
and which is control,

165
00:07:35,164 --> 00:07:39,031
so they don't unintentionally influence
the results through their own behavior.

166
00:07:39,031 --> 00:07:41,044
In which case, it's called...

167
00:07:41,044 --> 00:07:43,335
You guessed it! A double-blind procedure.

168
00:07:43,335 --> 00:07:46,178
So let's put these ideas into practice
in our own little experiment.

169
00:07:46,178 --> 00:07:48,698
Like all good work, it starts with a question.

170
00:07:48,698 --> 00:07:51,003
So the other day, my friend 
Bernice and I were debating.

171
00:07:51,003 --> 00:07:53,531
We were debating coffee 
and its effect on the brain.

172
00:07:53,531 --> 00:07:56,882
Personally, she's convinced that coffee 
helps her focus and think better,

173
00:07:56,882 --> 00:08:00,294
but I get all jittery, like a caged meerkat 
and can't focus on anything.

174
00:08:00,294 --> 00:08:03,806
And because we know that over-confidence 
can lead to belief that are not true,

175
00:08:03,806 --> 00:08:05,783
we decided to do some critical thinking.

176
00:08:05,783 --> 00:08:07,422
So let's figure out our question:

177
00:08:07,422 --> 00:08:10,071
"Do humans solve problems 
faster when given caffeine?"

178
00:08:10,071 --> 00:08:13,086
Now we've got to boil that down 
into a testable prediction.

179
00:08:13,086 --> 00:08:16,941
Remember: Keep it clear, simple, and 
eloquent so that it can be replicated.

180
00:08:16,941 --> 00:08:20,002
"Caffeine makes me smarter" 
is not a great hypothesis.

181
00:08:20,002 --> 00:08:21,774
A better one would be, say...

182
00:08:21,774 --> 00:08:26,941
"Adults humans given caffeine will navigate 
a maze faster than humans not given caffeine."

183
00:08:26,941 --> 00:08:30,852
The caffeine dosage is your independent
variable (the thing that you can change).

184
00:08:30,852 --> 00:08:32,261
So, you'll need some coffee.

185
00:08:32,261 --> 00:08:34,118
Your result or dependent variable--

186
00:08:34,118 --> 00:08:37,125
(the thing that depends on 
the thing that you can change),

187
00:08:37,125 --> 00:08:40,279
is going to be the speed at which the 
subject navigates this giant corn maze.

188
00:08:40,279 --> 00:08:42,794
Go out on the street, wrangle up a 
bunch of different kinds of people,

189
00:08:42,794 --> 00:08:44,986
and randomly assign them 
into three different groups.

190
00:08:44,986 --> 00:08:48,132
Also at this point, the American 
Psychological Association suggests

191
00:08:48,132 --> 00:08:50,972
that you acquire everyone's 
informed consent to participate.

192
00:08:50,972 --> 00:08:53,708
You don't want to force anyone
to be in your experiment,

193
00:08:53,708 --> 00:08:55,285
no matter how cool you think it is.

194
00:08:55,285 --> 00:08:58,184
So the control group gets 
a placebo (in this case, decaf).

195
00:08:58,184 --> 00:09:00,603
Experimental group 1 gets 
a low dose of caffeine,

196
00:09:00,603 --> 00:09:02,253
which we'll define at 100mg

197
00:09:02,253 --> 00:09:04,595
(just an eye opener, 
like a cup of coffee's worth).

198
00:09:04,595 --> 00:09:07,411
Experimental group 2 gets 500 mg

199
00:09:07,411 --> 00:09:10,334
(more than a quad-shot of espresso
dumped in a Red Bull).

200
00:09:10,334 --> 00:09:12,619
Once you dose everyone, 
turn them loose in the maze

201
00:09:12,619 --> 00:09:14,803
and wait at the other end with a stopwatch.

202
00:09:14,803 --> 00:09:18,063
All that's left is to measure your 
results from the three different groups

203
00:09:18,063 --> 00:09:19,910
and compare them, just to see if 
there were any conclusive results.

204
00:09:19,910 --> 00:09:21,940
If the highly-dosed folks got through it

205
00:09:21,940 --> 00:09:23,970
twice as fast as the low dose, placebo groups

206
00:09:23,970 --> 00:09:25,962
then Bernice's hypothesis was correct

207
00:09:25,962 --> 00:09:29,043
and she can rub my face in it, 
saying she was right all along,

208
00:09:29,043 --> 00:09:32,521
but really, that would just be 
the warm flush of hindsight bias

209
00:09:32,521 --> 00:09:35,493
telling her something she didn't 
really know until we tested it.

210
00:09:35,493 --> 00:09:38,336
Then, because we've used clear 
language in defining our parameters,

211
00:09:38,336 --> 00:09:41,443
other curious minds can easily 
replicate this experiment and

212
00:09:41,443 --> 00:09:43,588
we can eventually pool all the data together,

213
00:09:43,588 --> 00:09:47,809
and have something solid to say about what
that macchiato was doing to your cognition.

214
00:09:47,809 --> 00:09:50,732
Or at least the speed at which 
you can run through a maze.

215
00:09:50,732 --> 00:09:51,702
Science!

216
00:09:51,702 --> 00:09:54,107
Probably the best tool that you have
for understanding other people.

217
00:09:54,107 --> 00:09:56,513
Thanks for watching this episode
of Crash Course Psychology!

218
00:09:56,513 --> 00:09:57,745
If you've paid attention,

219
00:09:57,745 --> 00:10:00,517
you've learned how to apply the scientific
method to psychological research

220
00:10:00,517 --> 00:10:05,387
through case studies, naturalistic observation, 
surveys and interviews, and experimentation.

221
00:10:05,387 --> 00:10:08,850
You've also learned about different 
kinds of bias in experimentation,

222
00:10:08,850 --> 00:10:11,443
and how research practices help us avoid them.

223
00:10:11,443 --> 00:10:13,352
Thanks specially to our Subbable subscribers,

224
00:10:13,352 --> 00:10:15,815
who make this and all of Crash Course possible.

225
00:10:15,815 --> 00:10:20,255
If you'd like to contribute to help us keep 
Crash Course going and also get awesome perks,

226
00:10:20,255 --> 00:10:24,559
like an autographed science poster or 
even be animated into an upcoming episode

227
00:10:24,559 --> 00:10:27,574
go to subbable.com/crashcourse to find out how.

228
00:10:27,574 --> 00:10:40,885
[Host reads the credits]