MCQ Module 2

External validity refers to:

• How 'good' or 'true' the results of a study are based on the limitation of potential bias in the

  study design and methods

• The generalizability of the findings to the 'real world' population of interest not under strict

  'experimental' control

• How credible the results of a randomized controlled trial are

Detection bias is:

• Present when there are more than 5 time points at which outcomes are measured

• Present when different assessors collect outcome data

• Present when systematic differences are identified in how the data is collected between

  groups

• All of the above

Whilst undertaking critical appraisal, to determine that there has been an attempt

on the part of the researchers to limit attrition bias you may look to see that:

• A per protocol analysis has been performed

• The study is less than 8 weeks duration to limit losses to follow up

• An available case analysis has been performed

• An intention- to-treat analysis has been performed

A quasi-randomized controlled trial is named as such because:

• Opaque sealed envelopes weren't used for allocation of subjects to groups

• A systematic but non-random method was used to allocate subjects to treatment groups

• Participants were not blinded to their treatment allocation

• All of the above

Confounding occurs:

• When the effects of two or more processes are not separated, so the apparent effect is not

  the true effect

• When investigators mix up the data collection by accident

• When there is systematic deviation of the results or inferences from the truth

• Most commonly in randomized controlled trials

Which of the following is an absolute rather than relative measure of difference?

• Relative Risk

• Odds ratio

• Risk difference

• Hazard ratio

When dealing with meta-analysis of continuous data, an effect measure you may

calculate is the:

• Odds ratio

• Relative risk

• Standardized mean difference

Meta-analysis should only be undertaken:

• On randomized controlled trials

• Normally, only in the presence of clinical and methodological homogeneity

• Where the studies to be included have similar sized populations

A sensitivity analysis may be performed during a systematic review to:

• See how sensitive the data is to the normal distribution

• To test the assumptions or decisions made by the reviewers

• Increase external validity of the findings of the review

To minimize error during the data extraction process, reviewers can:

• Use a specialized data extraction form

• Pilot test the data extraction form prior to use for the review

• Train and assess all people involved in the data extraction process

• Have blind extraction of the data from each study by two people

• All of the above