basic statistics Flashcards
define descriptive statistics
stats that summarise otherwise confusing, large, complicated amounts of information. normally in the form of graphs or tables.
define inferential statistics
shows relationships between population and samples, used to draw conclusion.
what is anecdata ?
data with no real context and thus limiting its usefulness.
name the four types of descriptive statistics ?
nominal, ordinal, interval, ratio.
describe nominal data
- qualitative
- no natural order
- e.g. female/male , cat/dog
- can only show if phenomenon is present or not.
- a =/ b
describe ordinal data
- values can be ranked
- e.g. bronze/silver/gold , level/gcse
- a > b
describe interval data
- distance between values has meaning
- ratio not relevant
- 0 not well defined
- e.g. change in temperature
- a-b
describe ratio data
- fractions are meaning full
- 0 is meaning full
- 0 means an absence of phenomenon
- e.g. height, weight, volume, speed, time
- a/b
describe cross sectional data
- one observation per ( person, household, firm, country )
- no time dimension
- order of observation doesn’t matter
describe time series data
- one observation per time period ( day, week, quarter, month, year )
- order of observation does matter
describe pooled cross sectional data
this is when there is a set of two cross sectionals together
describe longitudinal or panel data
-following the same subject over different times
define discrete data
a. finite
b. countably infinite
a. bounded clear countable options. (i.e. age, agree/disagree, age)
b. infinite number of options but countable ( i.e. wealth, income)
which type of data is most common in economics ?
finite data
describe continuous data
measured on the number line, uncountable infinite options ( i.e. time, distance )
describe a histogram
- similar to a bar chart but with equal even class/bar widths
- height = frequency / class width
- height = frequency density
how do you construct a frequency polygon ?
you draw a histogram and place marks in the middle of the top of the bar charts and draw them together.
how do you determine how data is skewed ?
- when the mean is on the left of the peak ( negative / leftward skewed )
- when the mean is on the right of the peak ( positively / rightward skewed )
what is rounding bias ?
when the data is less accurate because the observes like to round their observations ?
what are these averages ?
a. mode ( ungrouped )
b. mode ( grouped )
c. mean ( ungrouped )
d. mean ( grouped )
e. median ( ungrouped )
f. median ( grouped )
a. most common observation, could be multi modal and when this occurs its not a very useful descriptor of the data
b. look at frequency density not just frequency
c. sum of values/number of observations
d. sum of class midpoint x class frequency / number of obs total
e. n/2 for even number chose in-between value
f. (( L + N/2 - B ) / G )X W L = lower boundary of median class B = cumulative frequency of the groups before the med group W = group width G = frequency of the median group
what is the expected value operator ?
predicts, if we were to select a random stat : what would it say ? normally equal to the mean
how do you calculate
a. quartiles
b. quintiles
c. deciles
d. percentiles
e. inter quartile range
a. n/4 x X
b. n/5 x X
c. n/10 x X
d. n/100 x X
e. 3rd quartile - 1st quartile
how do you calculate variance for
a. population
b. sample
a. sum of ; x - mean squared
divided by n
b. same as ^ but sample mean and divided by n-1 instead
how do you calculate standard deviation ?
it is the square root of variance
how does this transportation affect data ?
a. DATA + Y
b. DATA x X
a. range of y normal
variance of y normal
expected + y
b. range x X
variance x X^2
expected x Xhow do you calculate a Z score ?
(test value - population mean) / standard deviation
