Interpreting Graphs Flashcards

1
Q

What does likelihood ratio nomogram show?

A

Pre-test probability
Likelihood Ratio
Post-test probability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the post-test probability?

A

Probability that a patient scoring positive on a test actually has the diseae

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

How does one get the post test probability from a likelihood nomogram?

A

Straight line drawn through pre-test probability and likelihood ratio will yield post-test probability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

How does one obtain the total sample size needed to detect group difference (effect) with specified ability (power)?

A

Altman Nomogram

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How does one draw an Altman Nomogram?

A

By joining value of power on right-hand axis with standardized difference (size of effect) on left-hand axis using a straight line

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What does the intersection of the line with the upper part of the central platform in an Altman Nomogram give us?

A

The sample size required for the study to detect the difference with a preset probability threshold for false positives, p of 0.05

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What does the intersection with the lower part of the central platform on an Altman Nomogram tell us?

A

Sample size required for study to detect difference with p of 0.01

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

How does one calculate the effect size from the Altman nomogram?

A

By the standardised difference

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Calculation for standardised difference?

A

Target difference / standard deviation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is the standardised difference referring to?

A

Target difference between 2 groups studies expressed in terms of the standard deviation of variable across the study population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

How can one estimate the standard deviation before conducting a study?

A

Searching literature

Doing small pilot study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What may one wish to calculate before starting a study?

A

Required sample size by keeping constant p-value, power and estimate an expected effect size

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

How can one use Altmans nomogram for a study already completed?

A

Can test the power of the study using the given sample size, found effect size and set p-value

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What does sample size depend on?

A

P value
Power
Expected difference
Expected variability of outcome variable in study population (SD)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What does funnel plot detect?

A

Publication bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

X and Y axis of a funnel plot?

A

X axis: measure of effect size

Y axis: measure of precision

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What can be used as measure of effect size on funnel plot?

A

Cohens D
OR
RR

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What can be used as measure of precision on funnel plot?

A

Sample size

Standard error

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Where do studies with larger sample sizes go on a funnel plot?

A

Plotted higher on y axis

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Where is true effect size obtained from?

A

Meta-analysis

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What would we expect in the relationship between larger studies and the true effect size (drawn as a vertical line)?

A

Larger studies (at the top end) to be closer to the line

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What would we expect in the relationship between smaller studies and the true effect size (drawn as a vertical line) on a funnel plot?

A

Smaller studies (at the lower end) would be spread from the line of true effect size

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Why should larger studies be closer to the line of true effect size in funnel plots?

A

Because they are more representative and have results closer to the true effect sizes

24
Q

How do we assume from a funnel plot that there is no publication bias?

A

Inverted funnel

25
Q

How do we note the presence of publication bias in a funnel plot?

A

An obvious void at either corner of the funnel plot, suggestive of missing studies

26
Q

In a ROC curve, what does the point at the start of the curve indicate?

A

Cut-off with lowest sensitivity

27
Q

In a ROC curve, what does the point at the end of the curve indicate?

A

Cut-off with lowest specifcity

28
Q

What does is the straight line made between the 2 points of a curve in a ROC curve called?

A

Chance line

29
Q

What does a chance line on a ROC curve indicate?

A

Curve of a test that performs no better than diagnosing by chance

30
Q

What is the leaf in the stem and leaf plot?

A

X axis

31
Q

What is the stem in stem and leaf plot?

A

Y axis

32
Q

Significance of leaf in stem and leaf plot

A

Each number corresponds to a single observation when the tenth difit value og the stem is added to the leaf value

33
Q

How do identify endogenous variables in path analysis

A

The variables will have arrows originating from and pointing towards them

34
Q

What is the beta regressor in path analysis?

A

Path coefficient

35
Q

What is the R square in path analysis

A

Coefficient of determination

36
Q

What is the disturbance term in path analysis

A

Error in measurement

37
Q

When might log values be used in a scatter plot?

A

To yield parametric transformation

38
Q

Calculation for precision

A

1/SE

39
Q

X and Y axis of Galbraith plot

A

X: precision
Y: Effect size

40
Q

What lines are shown on Galbraith plot?

A

Central line: line depicting pooled effect size
Line above: +2Sd line
Line below: -2SD line

41
Q

How does one identify heterogeneity in a Galbraith plot?

A

Outliers from the lines

42
Q

X and Y axis in L’Abbe plot

A

X: control group
Y: Intervention group

43
Q

What are the lines in a L’Abbe plot?

A

Dotted

Equal

44
Q

What is the dotted line in L’Abbe plot?

A

Pooled RR line - summarised RR from meta-analysis

45
Q

What is the equal line in L’Abbe Plot?

A

Indicates that the event rates are equal in two arms within trial (RR=1)

46
Q

What does it mean if a trial is below the equality line in L’Abbe plot?

A

Control groups have better event rates than intervention group

47
Q

What does each circle represent in L’Abbe plot?

A

Individual trial

48
Q

What does the size of the circle in L’Abbe trial correspond with?

A

Number of participants

49
Q

What is the upper adjacent in box whisker plot?

A

Maximum value excluding outliers

50
Q

What is the lower adjacent in box whisker plot?

A

Minimum value excluding outliers

51
Q

What is the upper quartile in box whisker plot?

A

Corresponds to 75th percentile value

52
Q

What is the lower quartile in box whisker plot?

A

Corresponds to 25th percentile value

53
Q

What is the horizontal line in box whisker plot?

A

50th percentile - median

54
Q

What does a dashed horizontal line mean in box whisker plot?

A

Mean

55
Q

What are asterisks used for in box whisker plot?

A

Extreme outliers - >2SD or 3 box lengths away from quartiles