BIOSTATISTICS
1. The more prevalent a disease, the more will be it's P.P.V..i.e., less will be it's False Positives. Less prevalent
more false positives.
---- As prevalence Increases --- PPV Increases ---- False +ve Decreases.
---- As prevalence Decreases -- PPV decreases ---- False +ve Increases.
2. Standard Deviation
If values are same as mean, then SD = 0.
3. The NORMAL Distribution
The probability of a reading to fall outside 2 SD is 5% ie., 5/100 = 1/20 = 0.05.
Normal Curve
-----------------------
Mean = Median = Mode = 0.
S.D = 1.
Area = 1.
No: of Observations - Infinite.
Relative Deviate or Standard Normal Deviate (Z)... Z = (x - Mean) / SD.
Z------> The distance of a value 'x' from Mean, in Normal curve, in terms of SD is called as RD/SND/Z.
4. Determinants of causality - Criteria is Bradford- Hill Criteria.
It does not include Concurrency.... ie., the cause and effect need not be present simultaneously.
Criteria includes.... 1. Temporal Sequence 2. Biological Gradient 3. Strength of Association etc.
5. Two variables can be simultaneously plotted in scatter plot diagram / Correlation Diagram.
6. Degree of Freedom = (c-1) x (r-1), where 'c' is the number of columns and 'r' is the number of rows.
Note : Although this is the usual formula used, there is exception. Please check it.
7. Random / Probability Sampling is used in Inferential Statistics
Are of following types--
1. Simple Random Sampling -
2. Stratified Random Sampling - Population is stratified into internally homogenous
groups...Population ---> Village and Town.
Into Hindus and Christians.
3. Cluster Sampling - When Sample size is Very large.
4. Systematic Sampling - for eg:- Every 3rd person admitted in a Hospital.
8. Standard Error of Mean = S.E. = Standard Deviation / Sq.rt of 'n'.
So in Random sampling, if sample size is increased, the Standard Error would decrease.
No comments:
Post a Comment
Please add your comments or suggestions here..
Note: Only a member of this blog may post a comment.