Professional Documents
Culture Documents
Type
a) descriptive / analytic
b) qualitative/ quantitative
DATA -
- what is your age?
- hair color?
- hair style?
Classification of variable based on relationship
1. Independent
2. Dependent
3. Intervening
4. Confounding
INDEPENDENT VARIABLE: Dependent variable
OBESITY - MI
INTERVENING
VARIABLE
• Variable through which independent variable influence dependent
variable.
FAT - ATHEROSCLEROSIS - MI
DM - DEMENTIA FAT - MI
OLD AGE HTN
Classification of variable / data based on character
Quantitative / numerical Qualitative / categorical
1.Varies in amount 1.Categories are simply used to distinguish one
2.Can be measured from another. Male / female hindu /
3.Got scale of measurement muslim
4.Obtaining by counting. 2.No magnitude
3.No scale of measurement .
4.May take numerical value which size is not
meaningful.( NID)
Type a) fixed / discrete
b)continuous
Type a) nominal
b) ordinal
Quantitative / Numerical data
Continuous Discrete / fixed
• Expressed in fraction / range • Can not be expressed in fraction.
- height / weight / BP - family member / patient number .
Qualitative / categorical data
NOMINAL ORDINAL
Can not be expressed in order. Expressed in order
• Valid
• Reliable
Type of study design
1. Observational study 2. Experimental study
DESCRIPTIVE ANALYTIC
-clinical trial
- RCT ( randomized clinical trial)
1.Case study 1.Case control. - community interventional trial
- quasi experimental trial
2.Case series . 2.Cohort study
EXAMPLE
PROSPECTIVE RETROSPECTIVE
Question
1. RCT
2. CASE CONTROL
3. COHORT
4. CASE SERIES
Concept of cross sectional and longitudinal study
Cross sectional Longitudinal study ( case control / cohort )
No follow up . Follow up
Example Example
CHARACTER
1.Prevalence (number of dengue case in 2019) 1.Incidence ( number of new case admitting in hospital)
2.Quick and inexpensive
3.No temporal relationship (cause – effect)
4.Prone to bias
5.Not good for rare diseases
Case control
1. Retrospective study
3.Appropriate for rare disease study ----- study of covid patient in china at the beginning.
Steps - case and control selection / matching / exposure measurement /analysis and interpretation
Cohort study
1. Prospective study.
2. Best – cause to effect
Example – exposed and unexposed group
smoker / non smoker – regular follow up – lung ca
• 100 student
• U need 5 male stu and 5 female stu
• Make 2 group – one male , one female
• Then randomly select
Cluster sampling
• Example – microcredit
Quota sampling
• Covid research – 2 doctor / 2 nurse / 2 imam
Snowball sampling
Example – health status about iaba abuser
Measures of location
• Central tendency – mean / median / mode
• Percentile / decile / quartile
Mean
Mode
Percentile / decile / quartile
Measures of dispersion
• Range
• Mean deviation
• Standard deviation
• Variance
• Co efficient of variance
Question
• Name the measures of dispersion
Mean deviation
Standard deviation
• More acceptable.
EXAMPLE Example
1. Chi – square / X2 test
3. Student T test 2. Fisher exact probability test
4. Proportion Z test 3. Man whitney U test
5. F test / fishers test/ ANOVA 4. Wilcoxon rank sum test
6. r test / pearson correlation test 5. Proportion Z test
6. Kruskal Wallis test/Mc Nemar test/ logistic regression
7. Friedman test / spearman rank correlation
T Test = to find out the significance of difference between two means
1. Parametric test
2. For quantitative data
3. Random sample
4. Sample size less than 30
5. Variable must be normally distributed
Z test = difference between two proportion .
• Quantitative / numerical data
• randomly selected
• Sample size should be large. >60 = definite
• Normally distributed variable
• Exm – 80% success / 20 percent failed.
Chi – Squared test
1. Non parametric test
2. For qualitative / discrete data.
3. No mean median mode
4. To measure relationship between two qualitative variable.
5. Example – smoking / cancer, vaccine/ immunity.
Deference