3. Lesson objectives
At the end of the lesson, the participants
will be able to:
Enlist types of variables:
Reasons of knowing types of variables
?
Describe essentials of data editing
6. Nominal scale
In master chart in form of label, text
Example: Gender, Religion , ,
No natural order
No absolute zero
Descriptive statistics: Frequency (Count)
Proportion
7. Ordinal scale-
In master chart in form of label, text
Example: Severity of disease, Grading of
cancer,Apgar score, Grading of disability
, ,
Natural order exists
No absolute zero
Descriptive statistics: Frequency (Count)
Proportion
8. Interval scale
In master chart in form of numerical: /
Natural order exists
No absolute zero
Negative values exist
Example: Env.Temp Temp: Degrees F
Statistics: Mean, Median, Mode, SD
9. Ratio scale
In master chart in form of numerical /
Natural order exists
Absolute zeroYES
Negative values exist
Example: Height,Weight, Hb
Statistics: Mean, Median, Mode, SD
10. Summary
Characteristic Nominal Ordinal Interval Ratio
Value assigned Label Label Numerical Numerical
Meaningful
order
No Yes Yes Yes
Meaningful
interval in
values
No No Yes Yes
Meaningful
ratios in values
No No No Yes
Absolute zero No No No Yes
Possibility of
negative values
No No Yes No
12. Stats
Statistics
permissible
Choice of
statistical test
Permissible statistics
Nominal Proportion, Chi-square
,
Ordinal Proportion,, ,
rank order correlation, non-parametric
analysis of variance
Interval Mean, Median and mode, Range
percentile,
, , , , percentile
rank order correlation, non-parametric
analysis of variance
Ratio All statistics permitted for interval scales
plus the following:
geometric mean, harmonic mean,
coefficient of variation, logarithms
13. Data Editing:
Check for
Completeness
Accuracy
“Illegal entries”
out-liers
“Improbable
entries”
Completeness
Blanks
Nils
I don’t knows
14. Data Editing
Accuracy
Sample checks
Standardization
Triangulation
Illegal entries
Entry that should
not have been there
Inclusion criteria
violated
Improbable entries
Cause of death
“Rupture of uterus” in
male
4 living children to girl
aged 15
Outliers
Extreme values, Box
plot
15. Dataediting methods
Check for blanks, nils and Dont knows
Range check: Minimum & Maximum
Count check:Totals / sub-totals
Validation checks:Validation at entry stage
in MS-EXCEL
Use of Software
16. DataReduction
Tables, groups
K = 1+ (3.332 x Log 10 (n))
K Number of categories
N = number of observations
N = 200
Then K = 8.66= 9
Less than 5, more than 15 categories NOT advised
17. Data Reduction:
Descriptive Statistics
Rate, Ratio, Proportion
Rate: Numerator,
Denominator, Multiplier
Ratio: X & Y. How much
will be Y if X =(1,100,
1000)
Proportion: X /(X+Y)
Centering Constants:
Arithmetic Mean
Geometric Mean
Median
Measures of Variation
Range
Inter-quartile Range
Standard Deviation,
Variance
Percentile
95% CI