Lecture 5: Descriptive Stats Flashcards
variable (noun)
an element , feature, or factor that is liable to vary or change
two types of variables
math & computing
math variable
quantity during calculation is assumed to vary
computing variable
data item that may take on more than 1 value during the run time of the program
scales of measurement: data types
- nominal
- ordinal
- continuous
nominal measurement scale
based on the classification of an observation according to the group to which it belong – a process of categorization
EX: gender, political party, marital status
ordinal measurement scale
based on the classification of an observation according to its relationship to other observations
EX: poor-fair-good scale
types of continuous measurement scales
interval & ratio
continuous interval measurement scale
equal units of measurements, distance between two numbers are known, 0 is arbitrary
uneven interval measurement scale
number used but they do NOT have to represent ‘true numbers’ that can be added/subtracted
nominal or ordinal
continuous ratio measurement scale
equal units of measurement & a TRUE 0 at its origin
EX: mass & time
population
greek symbols
constant
samples
roman characters
variable
mode
most often
best for: nominal
median
in the middle
best for: ordinal
mean
average
best for: interval/ratio
range
=X max - X min
inter-quartile range
= X 25th percentile - X 50th percentile
variance (S^2, MS)
average of square of deviations from mean divided by degrees of freedom (N-1)
standard deviation (SD)
“positive variance”
square root of variance (MS^1/2)
coefficient of variation (CV)
percentage of spread (unit-less)
=100% x (SD/mean)
central limit theorem
the sample distribution of the mean of any independent, random variable will be normal if the sample size is large enough (30-40)
conditions for more sample points
more close samples to normal distribution
conditions for less sample points
more close original population to normal distribution
standard error mean (SE)
standard deviation of sample means from population mean
= SD/(N^1/2)
SD VS SE
SD- measures variability of individual subjects around mean
SE- assessed how accurate a sample mean reflects population mean
