Numerical Summaries Flashcards
Percentile
-A number such that of our observations fall at or below this number when ranked from smallest to largest
-Quantative
Measures of percentiles
1.Minimum
2. 25th or first quartile
3. 75th or third quartile
4. maximum
IQR
q3-q1 and gives us the range of the middle of the 50% of our data ( center)
-Measure spread
Quantitive and categorical
1.side by side box plots
2. Stacked histograms
3. Grid of histograms
Quantitive distribution
Shape
center
spread
outliers
Spread
tells us how spread out the data is and the variability
Order statistics
-numerical summaries based on the ordered ranking of a quantative variable
-What makes it useful
-No assumptions
-robust
Moment statistics
Statistics that are based on specific mathematical properties of the data.
- do make assumptions
-not robust
Median (center)
another name for the 50% percentile
Mean (center)
same thing as the average
add up all the values and divide by observations
Standard deviation (Spread)
-measures spread from the mean
-cannot be negative
-
Order statistics
median and IQR
-Robust to outliers
-More correct
Moment statistics
mean and SD
- Sensitive to skews and outliers
when do we use order stats
skewed shape or outliers
when do we use moment stats
Symmetric shape with no extreme outliers
