statistics notes

population

the entire group to be studied

sample

subset of population for whom data is collected

variable

quantity to be studied

parameter

some feature of interest of a population

statistic

corresponding feature in a sample

nominal data

data that you can categorize (Fav Color?)

ordinal data

sorting/ordering the data (Height?)

interval data

you can add or subtract the data, differences of one unit are the same scale (6 vs 5, 2000 vs 1999)

ratio

data is interval & there is a natural zero, where 0 means “no amount” (Weight?)

confounding

the actual direct reason of impact

sharpshooter effect

cherrypicking data to support a conclusion

self-selection

participation in experiment is optional

survivorship bias

only individuals who “survive” are examined when making a conclusion

discrete

gaps between data points (you can have 2 to 3 friends, but not 2.5 friends)

mode

point of highest frequency

median

middlemost data point

mean

sum of all data divided by the number of data

skewed

lots of high or low data

uniform

all data is the same across the board

bimodal

most of the data is in 2 different clusters

symmetric

mean, median, mode all same

extrapolation

predictions outside of known data

r

correlation coefficient

r squared

coefficient of determination