Bias

Question 1

Allocation bias

Answer 1

Reduced by concealment

Investigator knows next participant will get intervention, so chooses someone who will do well

Question 2

Selection

Answer 2

Reduced by randomisation

eg response bias in a survey- participants differ from general population

Question 3

Performance

Answer 3

differences in care provided beyond just intervention

Question 4

Hawthorne effect

Answer 4

participants change behaviour as aware being studied

Question 5

Attrition

Answer 5

people withdraw from arm, eg more severe disease may be ltfu

ITT/PPA

Question 6

Detection

Answer 6

Groups differ in way information is collected

detection= differ

Question 7

Observation bias

Answer 7

researcher’s views influence data collection

Question 8

Ascertainment

Answer 8

sampling bias (some people more likely to be picked)
or observer may change the way they ask questions

Question 9

Admission rate bias

Answer 9

only select those who attend hospital

Question 10

Measurement

Answer 10

definitions of the disease are inaccurate

Question 11

Analysis

Answer 11

Incorrect statistical methods

Question 12

Publication

Answer 12

poor results not published