research skills 8

Question 1

what is a type 1 error ?

Answer 1

when you accept the alternative hypothesis but really the results happened due to chance
false positive

Question 2

what is a type 2 error ?

Answer 2

accept null hypothesis when you shouldn’t have
false negatives

Question 3

what do P values tell you ?

Answer 3

p= 0.001 tells you statistically very strong data
p= 0.01 tells you moderately strong data
p= 0.05 tells you statistically weak data
p > 0.05 tells you weak to very weak data

Question 4

summarize what was said in this lecture

Answer 4

Null hypothesis significance tests compare the effect size with the variation

They summarise the information in a single number- the p-value

A small p-value tells you that the data does not look a lot like random data

The p-value is not the probability that the null hypothesis is true

p=0.05 tells you there is useful evidence, but it is statistically weak

p=0.01 tells you the evidence is statistically moderately strong

To have really strong evidence, you need consistent results from repeated experiments