
BIometryBiostatistics
What does it consist of?
 Collection, Classification
 Description, Analysis
 Interpretation, Presentation

Stats (Types & Usage)
 Descriptive & Analytical summary measures of data
 Scientific method for interpreting data obtained
 Vital Stats (PH agencies) Birth, deaths, marriage, health related conditions
 Reenforcement of other findings
 Statistical reasoning (clinical judgment, scientific judgment)

DATA:
Refer to individual values (3things)
Properties (4 things)
 Individual values: 1. presented 2.measured 3. observed
 Properties: 1. sample population 2. grouped or ungrouped
 3. quantitative or qualitative 4. discrete or continuous

Where do we get Data?
Types of Sample population?
  from a population
 Types of sample populations :(simple random, nonrandom, Stratified selected)

What is grouped? Ungrouped?
 Groups= consist of identical data, by frequency
 Ungroup= Presented individually(even if identical)

Quantitative?
Qualitative?
Discrete?
Continuous?
 Quantitative( base on Number)
 Qualitative (nonnumerical, based on categories)
 Discrete (distinct categories, limited # of possible values exist)
 Continuous (Unlimited # of possible values)

a. Accuracy?
b. Precision?
c. Reliability?
d. Validity?
 a. exactness, measure true value of what being studied
 b. consisten and reproducible
 c. stable, dependable, sound
 d. measure what is supposed to measure

a. Distribution
b. Normal
c. Skewed
a. complete summary of frequenies or proportion of a characteristic for a series of data from a population or sample.
b. AKA: Gaussian (continuous bell shape)
c. (+) skewed= to the right, () skewed= to left

Descriptive stats (Measure of central Affinity)
1. Arithmetic Mean?
2. Median?
 1. Mean (sensitive to extreme values in a series)
  Add all #/ Total #'s in series
 2. Divides series into 2 equal groups (not sensitive to extreme values), better representative of central tendency than mean.
  also it is the middle of Percentiles= Values divide series into defined percentages (level of measurment below which a specified portion of distribution falls)
  Middle number

Continued:
3. Median?
4. Mode?
5. Range?
6. Variance?
 3. if ODD # of values (lowest> highest and divide in 1/2)
  if EVEN # (2 values divide in 1/2, calcutae their mean)
4. Most occuring values in a series (Epi studies for peak of disease occurance)
5. Difference b/w highest and lowest value (measure data spread)
 6. Main use to calculate standard deviation
 SD= +ve sq. root of the variance

Analytical stats.
7. Probability?
8. Rule of probability?
9. Null hypothesis
 7. quant. expression of likelihood of occurrence
  defined in terms of relative frequency (never >1 or <0)
 8. PR (A does not occur)= 1 PR (A occurs)
 PR (A/B)= # of times A&B occur jointly or of times B occurs
 9. Samples compared are similar (any difference is due to chance)
  used to define significant difference

Screening
1.theories
2.programs
3.sensitivity
4.specificity
  testing or examining individual or large groups of people
 It separates from those who have high probability of having condition under study.
1. Theories concerned w/ diseases, diagnostic testing, used as a case finding tool, and treatment testing.
 2. Programs (must meet certian criteria)
 Segment being screened @ relatively high risk for cond.
 the disease= of enough concern to community being screened
 Undiagnosed disease= should be more responsive to TX than if diagnosed @ later symptomatic stage
 (screening test are: sensitive/specific, applicable to Lg #, easy and quick, should not cause harm, inexpensive)
 (Those with (+) results guaranteed follow up evaluation)

3. Senitivity?
4. Specificty?
 3. ability to ID correctly those who do NOT HAVE the disease (if 100% sensitivity= ())
 ability of the test to give a (+) result when person tested truly has disease.
  Is a %
 4. ability to ID correctly those who have the disease (if 100%= (+))
 ability to test to give a () result when person tested is free of disease
 Is a %

Setting Cutoff? depends on?
 Natural Hx of disease
 effectiveness of intervention (early or late)
 if disease is rare: Sensitivity=high (or cases present will be missed)
 If very lethal (early detection improves prognosis)
 (high sensitivity is necessary)(false + are tolerated, false 's are not)
Screening if very rewarding (high sensitivity)

