Topic 2

Question 1

why are most published research findings false?

Answer 1

statistical challenges
biases
misconduct due to reward system in science

Question 2

of 49 articles, how many were shown to be wrong?

Answer 2

14 (41%)

Question 3

ppV

Answer 3

positive predictive value
probability that a research finding the has been claimed to be true, is in fact true
1-B
(1-type 2 error rate)

Question 4

Type 1 error

Answer 4

failure to reject a null hypothesis

false pos

Question 5

type 2 error

Answer 5

false negative

failure to reject a null hypothesis

Question 6

corollary 1: smaller the studies, the less likely they are to be true

Answer 6

small sample = small power

= small PPV

Question 7

Corollary 2: smaller the effect size the less likely it is to be true

Answer 7

power decreases with small effect

lots of false positives

Question 8

Corollary 3: greater the # and lesser the selection of tested relationships, less likely to be true

Answer 8

PPV depends of restudy odds (R)

research findings are more likely to be true in confirmatory designs

Question 9

Corollary 4: greater the flexibility, less likely they are to be true

Answer 9

flexibility increases potential for bias (U)

common standards =increase PPV

Question 10

Corollary 5: greater financial interests, less likely the findings are to be true

Answer 10

bias

Question 11

Corollary 6: hotter the scientific field, the less likely the research findings are to be true

Answer 11

PPV decrees with increasing number of research teams

Proteus phenomenon

Question 12

Proteus phenomenon

Answer 12

rapidly alternating extreme research claims and extremely opposite refutations
common in molecular genetics