Cross Sectional Studies or Surveys

Question 1

What do cross sectional studies measure

Answer 1

Cross sectional studies measure the prevalence or frequency at one point in time of a condition

Question 2

Can cross sectional studies identify causal relationships

Answer 2

no but they can suggest possible causal relationships that might be worth investigating

Question 3

What kind of study is used to identify causal relationships

Answer 3

Longitudinal studies = Observational (cohort) or experimental (controlled trial)

Question 4

What are cross sectional studies also referred to as

Answer 4

Surveys

Question 5

Why can’t cross sectional studies establish causal relationships

Answer 5

We cannot establish causal relationships as there is no time lag between exposure and event

Question 6

What 2 processes should a reliable survey go through to ensure it is reliable

Answer 6

Validation and Calibration

Question 7

What is Calibration

Answer 7

Calibration is the checking of the accuracy of a clinician’s measurements against a standard or, at the very least, the other assessors

Question 8

What is the Kappa statistic, mainly used in calibration studies

Answer 8

This is used to measure the level of agreement of assessor’s or tests or some shit uno

Question 9

What is considered a perfect, very good and good kappa score

Answer 9

1. 0 = perfect
2. 7 = very good
3. 5 = good

Question 10

What are the 2 important aspects relating to generalisability in cross-sectional studies

Answer 10

• The sample frame (i.e. the people from which the sample is drawn) should be representative and the sample should be representative
• The response rate and risk of biased response

Question 11

Why do we need to be careful of a biased response

Answer 11

The responders/non-responders could all share a particular trait

Question 12

When is the mean average used

Answer 12

When the values we are talking about are spread in a normal distribution

Question 13

When is a median average used

Answer 13

It will be the same as a mean in the normal distribution curves, so is more often used in values that are asymmetrical/skewed.

Question 14

When is the mode average used

Answer 14

When using categorial data that isn’t continuous

Question 15

What is a power calculation

Answer 15

A power calculation is a measure to tell us how many patients are required to avoid a type 1 and type 2 error

Question 16

What is a type 1 error

Answer 16

This is a false positive and is the error caused by rejecting a null hypothesis when it is true

Question 17

What is a type 2 error

Answer 17

This is a false negative and is the error caused by accepting a null hypothesis when it is not true