
Nuremburg Code
1st recognized code of ethics

Examples of abuses of clinical trials
 Thalidomide and birth defects
 Sloan Kettering Institute
 Tuskegee sypillis experiment

1974 National Research Act
 Establised IRB's
 Establised National Commission for Protection of Human Subjects

1978 Belmont Report
 Set ethical principles and guidelines in research
 Need to consider the risks and benefits of human testing
 Need to obtain informed consent
 Subjects tested must be of sound mind

Measures of Variability
How wide the observations are spread around measure of central tendency

Range
Difference between largest and smallest value

IQR
Spread of the middle (50%) of numbers

Five number summary
 Minimum
 1st quartile
 median
 3rd quartile
 maximum

Nuremburg Code Outlines
 Voluntary informed consent
 No unnecessary research
 Research should be preceded by surveys and animal experimentation
 Research protocol should be reviewed by scientifically qualified professionals
 Research should be stopped if subject is endangered
 Physical and mental suffering should be prevented
 Subjects are free to with draw at anytime

Four Phases of Clinical Trials
 1. Evaluate safetey, determine dosage, identify side affects
 2. Evaluate safety, determine effectiveness
 3. Collect more information about safe usage, confirm effectiveness, monitor side affects, compare to alternatives
 4. Collect data on effect on specific groups, monitor long term side affects

IRB
Institutional Review Board

PHSC
Protection of Human Subjects Committe SWOSU's IRB

Components required for PHSC application
 Why the research is being done
 What the researchers want to accomplish
 What will be done during the trial and for how long
 What risks are involved in the trial
 What other treatments are available
 What benefits may be expected from the trial
 The fact that participants have the right to leave at any time

IACUC
Institutional Animal Care and Use Committee

Biosafety Committee
Ensure safety when using potentially infectious microbes in research

Descriptive Statistics
Use data to describe sample

Inferential Statistics
Use sample to make conclusions about the population

Mode
The most common observation or category

Pro's for Mode
 Easy to obtain
 Not sensitive to outliers
 appropriate for both qualititave and quantitative data
 Can use to predict the value or category of next observation
 Easy to explain
 Easy to present

Con's for Mode
 May not be a clear mode
 May not be just one mode
 Does not provide us information about the entire distribution of data
 Sensitive to how to categorize information

Mean
The average of the data

Disadvantages using mean
 Can take on a fractional value, even if original data is discrete
 Best used for symmetric data, because is sensitive to extreme measures

Median
Center of data when sorted into an array

Advantages of Median
 Relatively easy to obtain
 Can always obtain a median versus mode
 Is based on the entire distribution of data versus mode
 Not influenced by extreme values versus mean
 Appropriate for quantitative and ordinal data

Mean = median
Symmetrical

Mean > median
right skewed

Mean < median
left skewed

If you have skewed data, what measure of center is going to be affected?
mean

Trimmed mean
Drop top and bottom values

Winsorized mean
Change 5% to less extreme value


Quartiles
25,50, and 75% Q1,Q2,Q3


Find outliers by making imaginery boundaries using:
Q1(1.5*IQR)
Q3+(1,5*IQR)


Variance
 s^2
 sigma squared
 Never negative
 Accounts for each observation
 can be affected by outliers

Standard Deviation
 Square root of variance
 same units as original data

Coefficient of Variation
 s/mean*100%
 Can compare to different types of data by eliminating the units

