Dealing with bias and uncertainty

Question 1

What is research?

Answer 1

aim : develop new treatment & information to benefit people

• pre clinical and plus clinical
• bench to bedside: translate lab based research to patients

Question 2

What are two types of research?

Answer 2

1. qualitative

2. quantitative

Question 3

What are quantitative study designs?

Answer 3

• observational designs
  =>record behaviours, attitudes and symptoms - naturalistic
  -experimental design
  => researcher controls factors and records effects on intervention on outcomes

Question 4

Example of observational study

Answer 4

Background:
-scientists use pipettes to dispense precise quantities of liquids, studies show egonomic problems (work related upper limb disorders) associated with pipette use.
- Aim: estimate proportion of ergonomic problems associated with pipetting.
Methods:
cross sectional study
survey six organisation in uk
150 questionnaires

Question 5

What are some bias in this study?

Answer 5

• sample: represent population, use it to make inferences about population, so can’t be sure
  -Selection bias: systematic difference between population and sample
  -non-response bias: systematic difference between non responders and responders
  Volunteer bias: systematic difference between volunteers and population
  Generalisation: inferences about population based on sample, sample representative population.

Question 6

What is a population?

Answer 6

• restricted by geography

- restricted by time

Question 7

What is sampling error?

Answer 7

• sample estimate (estimate population value) may not equal population parameter(actual population value)
• to reduce sample error increase sample size
• confidence interval : predicting uncertainty

Question 8

What are different methods of sampling?

Answer 8

Random sampling

convenience/opportunity sampling

Question 9

Response bias

Answer 9

• in self report studies :

- systematic difference between participant response and truth

Question 10

Assessor bias

Answer 10

• systematic difference between assessor measurement of participant and truth

Question 11

Caution

Answer 11

difficult to remove bias, only minimise biases

Question 12

Hawthorne effects

Answer 12

• change in participants behaviour due to attention received in study.
• eg: participant may change the way they use pipette to look better

Question 13

Hypothesis

Answer 13

• null and alternative hypothesis

- needs to be objective