1
Q
What is nominal data
A
Values not possible to rank ie. sex, colour
2
Q
What is ordinal data
A
Ordered categories on a scale without defined gaps between numbers ie. GCS
3
Q
What is interval data
A
Continuous numbers with no reference or Zero e.g. time
4
Q
What is ratio data
A
Continuous numbers with reference/zero e.g. days since MI
5
Q
Nominal and Ordinal data is what?
A
Non-numerical, Qualitative, categorical
6
Q
Interval and ratio data is what
A
Numerical, quantitative, continuous or discrete
7
Q
Number of people in ED is continous or discrete data
A
Discrete- whole numbers
8
Q
Height is continous or discrete data
A
Continuous, can have decimals or fractions
9
Q
How do you calculate mean
A
Add up all numbers and divide by how mnay numbers there are
10
Q
How do you calculate median
A
Put all numbers in order and take away the middle value
Average middle two if evenly set
11
Q
How do you calculate mode
A
Most common value
12
Q
What is range?
A
Minimum and maximum values in data
13
Q
What does standard deviation show?
A
Spread from a mean
14
Q
In normal distribution where is the mean on the peak
A
Middle
15
Q
In normal distribution what percentage of values are within 1 SD of the mean
A
68%
16
Q
In normal distribution what percentage of values are within 2 SD of the mean
A
95%
17
Q
In normal distribution what percentage of values are within 3 SD of the mean
A
99.7%
18
Q
What is nominal data
A
Data not orderable just has a name e.g. blood type
19
Q
What is ordinal data
A
Categories are ordered in some way e.g. disease staging, pain scales
20
Q
Nominal and ordinal data are examples of what
A
Qualitative/categorical data
21
Q
What is discrete data
A
Numerical data that can only be in whole numbers e.g. visits to the GP
22
Q
What is continuous data
A
Numberical data that has no limit on values e.g. height or weight. Can add a decimal
23
Q
Discrete and continuous data are examples of what
A
Numerical/Quantitative data
24
Q
Give examples of dervied data
A
Percentages, ratios, rates, scores
25
How can we treat derived data statistically
As continuous data
26
When is a bar chart used
Categorical or discrete data- seperation between bars to show non continuous
27
For categorical or discrete data that is non continuous what chart can you use
Bar chart
28
When is a pie chart used
Represent proportional frequency
29
What chart can be used to display proportional frequency
Pie chart
30
When is a histogram used
Continuous data with proportional representation, no gaps between bars to show continuous
31
What can be used to display continuous data with proportional representation
Histogram
32
When data is unimodal what does it mean
Has one peak
33
When data is bimodal what does it mean
Has 2 peaks
34
When data is unform what does it mean
All outcomes equally as likely
35
In unimodal data when is it symmetrical
Centred around a mid point- mirror image
36
In unimodal data when is it right (positively) skewed
A long tail to the right with a few high values
37
In unimodal data when is it left (negatively) skewed
A long tail to the left with a few low values
38
If both variables are continous how can this be displayed graphically
Scatter diagram
39
How do we calculate median in a data set with an odd number total
The number in the middle
40
How do we calculate median in a data sat weith an even number total
The mean of the 2 middle numbers
41
Positive right skewed data mean to median relationship
Mean is more than median
42
Negative left skewed data mean to median relationship
Mean is less than median
43
When do we use geometric mean
Right skewed data
44
What is geometric mean
Mean created using log to balance right skewed data
45
What is weighted mean
Used when parts of data carry more weight, when equal weighting =arithmetic mean
46
Advantages of mean
Uses all data
Manageable mathematics
Known sampling distribution
47
Disadvantges of mean
Distorted by outliers
Distorted bt skewed data
48
Advantages of median
Not distorted by outliers or skews
49
Disadvantages of median
Complicated sampling distribution, ignores most data
50
What is a percentile
Measure of a percentage that has that amount of data lying below is i.e. 5th percentile
51
What is variance
The measure of the spread of data from the mean
52
What is standard deviation
The square root of the variance (average of squared differences) an average of the deviations
53
What is the t distribution
Degrees of freedom
54
What is chi squared distribution
Right skewed distribution
55
What is the F distribution
Right skewed defined by a ratio
56
Name ways to transform data
Log, square root, reciprocal, square, logistic
57
Patient sample taken completely randomly from an entired population is a
Random sample
58
Patient sample taken from one hospital and study all or some of the patients with one condition is a
Convenience sample
59
What does standard error of the mean allow us to estimate
The prescision of our estimate of a sample
60
What is a confidence interval
Either side of a standard error of a mean within standard deviation percentages, tighter estimation of data precision
61
A narrow confidence interval suggests what
Precise data estimate
62
What is an experimental study
Study that involves the investigator intervening in some way
63
What is an observational study
Study that involves observation only, no intervention
64
What is a cross sectional study
Carried out at a single point in time e.g. number with a disease at one time= prevalance
65
Can cross sectional study show disease incidence
No just prevalance- one point in time
66
What is a longitudinal study
Follows a sample over a period of time
67
What is observer bias
one observer over or under estimates a variable
68
What is Confounding bias
Spurious association due to factors related to risk and outcome
69
What is selection bias
Patients selected are not respresentative of true population
70
What is information bias
measurements incorrectly recorded
71
What is publication bias
Tendency to only publish positive results/findings
72
How can we reduce variation in data
Replication, bigger sample size, study design, blocks
73
What is a block in sampling
Grouping individuals to compare within blocks rather than over all e.g. age group
74
In what study design would you use a washout period
Crossover design
75
What is a cross over design
One individual and 2 varaibles e.g. treatment at different intervals
76
What is a factorial experiment
Looking at 2 factors in the same group e.g.
Aspirin. Beta carotene
Yes. No
Yes. Asp+Beta Asp only
No Beta only None
77
What is a phase 1 trial
Small group healthy participants, low dose looking at saftey and side effects
78
Small group healthy participants, low dose looking at saftey and side effects describes what phase trial
Phase 1
79
What is a phase 2 trial
Small healthy group with disease, does it work and is it safe with dose testing
80
Small healthy group with disease, does it work and is it safe with dose testing describes what phase trial
Phase 2
81
What is a phase 3 trial
Bigger group, more vairables, comparison to standard treatment
82
Bigger group with disease compared against current treatment
Phase 3
83
What is a phase 4 trial
Longer term follow up of patients who have been prescribed itm more patient, more longterm benefits/side effects
84
Longer term follow up of patients who have been prescribed it more patients, more longterm benefits/side effects describes what phase trial
Phase 4
85
What is a positive control in a clinical trial
Existing treatment
86
What is a negative control in a clinical trial
Placebo
87
What is a randomised control trial
Selection truly random
88
What is a systematic allocation
Selection based on factors e.g. date of attendance or day of birth
89
What is a double blind trial
Patient and clinician dont know allocation
90
Cohort study identifies smoking causes cancer which makes it a
Risk factor
91
What is healthy entrant effect
Participants healthy to start the study as need to be disease free, lag on mortality
92
What is relative risk
Risk increase or recuction due to exposure to a variable
93
What is odds ratio
OR=odds in unexposed group/odds in exposed group
94
What study gives an odd ratio
Case control studies
95
What does odds ratio show
The odds of something happening vs the odds of it not happening
96
What does the risk ratio show
Compares risk of and event in one group against the risk of an event in another
97
What is a cohort study
Group identified by exposure and followed forward to assess outcome
98
What is a case control study
Group identified by outcome and prior exposure compared
99
How do you calculate risk ratio
RR=risk in unexposed group/risk in exposed group
100
What is the null hypothesis
No effect
101
What is the alternative hypothesis
If the null hypothesis is untrue
102
At what P value can we reject the null hypothesis traditionally
0.05
103
What is a type 1 error
Rejecting a null hypothesis that is true
104
What is the error if a null hypothesis is wrongly rejected
Type 1 error
105
What is a type 2 error
Accepting a null hypothesis that is not true
106
What is the error is a null hypothesis is wrongly accepted
Type 2
107
What is a meta analysis
Systematic review of numerical results
108
What is censored data
When some data from collection is missing e.g. survival data- over surviving collection cut off
109
What is left censored data
Missing data from start point cut off
110
What is right censored data
Missing data from end point cut off
111
What is used to display survival data
Kaplan meir curve
112
What is standard deviation
Average amount of deviation in your data
113
1 SD of mean is
68%
114
2 SD of mean is
95%
115
3 SD of mean is
99.7%
116
How do you normalise left skew
square or exponantial transformation
117
How do you normalise right skew
Log or square root transformation
118
The avergae amount of variability in a data set is
Standard deviation
119
Standard error of the mean is
A measure of precision of the estimate of the mean
120
A measure of precision of the estimate of the mean is
Standard error of the mean
121
Reference range is
descriptive of where 95% (2SD ) of the sample lies
122
Confidence interval is
A range of values that is likely to contain the true population parameter, based on sample data.
123
What is the P value
Probability assuming the null hypothesis is true of obtaining test results at least as extreme as observed
124
A confidence interval is a range of values that is likely to contain the true population parameter, based on sample data is what
P value
125
Type 1 error
Wrongly rejecting null hypothesis
126
Type 2 error
Wrongly accepting null hypothesis
127
What is power
The ability to detect a difference if there really is one
128
What is a cohort trial for
Risk ratio, determines aetiologya nd natural history of a disease
129
What is case control study
Reviews cases and controls to establish risk factors gives odds ratio
130
What gives odds ratio
Case control
131
What gives risk ratio
Cohort
132
what is incidence
Number of cases in a given time/number at risk
133
What is prevalance
Number of cases in a given population
134
What is the difference between incidence and prevalance
Incidence- new diagnosis in a time frame.
Prevalance proportion of a population
135
Sensitivity
Ability of a test to detect disease
136
Specificity
Ability of a test to detect a negative
137
PPV
How likely a positive result is a postive test
138
NPV
How liklely a negative result is a negative test
139
Survival data tests
Log rank, Hazard ratio, cox proportional hazard regression
140
Hazard ratio <1 means
Less chance of dying in treatment group
141
Hazard ration >1 means
More chance of dying in treatment group
142
Relative risk is
experimental rate/control rate
143
Relative risk reduction
1-(experimental rate/control rate)
144
Absolute risk reduction
control event rate - experimental event rate
145
Absolute risk increase
experimental event rate-control event rate
146
Number needed to treat =
1/ (control event-experimental event)
147
Number needed to harm =
1/ (experimental event-control event)
FRCR
flashcards
Decks in class (5)
# Cards
