Design A Study

Question 0

Description of study

What else to include

Answer 0

Follow up
Statistics (test)
Practicality

Question 1

Description of study

PICO

Answer 1

Participants
Intervention
Comparison
Outcomes

Question 2

Hierachy of evidence

Answer 2

Systematic reviews
RCTS
Cohort studies
Case control
Ecological
Case series
Ideas, opinions

Question 3

Systems to help improve quality of published evidence

Answer 3

CONSORT - standard for reporting RCTs

STROBE - standards for reporting observational studies (cohort)

Question 4

Stat tests

If dependant variable continuous

Answer 4

Either linear regression
T test
ANOVA

Question 5

When is linear regression used

Answer 5

If ind variable continuous

Or if it is not continuous and the groups are NOT alike

Question 6

When is t test used

Answer 6

Ind variable not continuous
Groups are similar
Only 2 groups

Question 7

When is ANOVA used

Answer 7

Ind variable not continuous
Groups similar
Have more than 2 groups

Question 8

When are fgroups similar

Answer 8

Usually in RCTs

Question 9

If dependant variable is NOT continuous

Answer 9

Logistic regression

Or chi squared test

Question 10

When is logistic regression used

Answer 10

If independant variable is continuous

Or if its NOT and the groups are NOT alike

Question 11

When is chi squared test used

Answer 11

If ind variable not continuous

And groups are similar

Question 12

T test

Answer 12

To test difference between means of 2 groups

Question 13

Anova

Answer 13

Test diff between means of more than 2 groups

Question 14

Chi squared tes

Answer 14

Determine if observed freq differ from expected freq
Has PD symptom increased? Freq of yes and no
Difference between vakues expected and observed

Question 15

Linear regression

Answer 15

Assess how one or more predictor values can be used to predict continuous outcome vairiable

Do age, gender and comorbidies predict distance walked in 2 mins following physio

Question 16

Logistic regression

Answer 16

Assess predictive value pf one or more variables on a binary outcome variable

Do age, gender and comorbidities predict which pD patients will show improvement?

Question 17

Sources of bias

Answer 17

Selection of participants
Similarity or differences in groups
Obtaining info