statistics

Extrapolation

extra data outside the pprediction points

interpolation

prediction inside data

Sampling strategies

• convecnience - most easily accessible

• simple random - equal/random chance of selection

• systematic - select from random starting point and fixed interval e.g. every tenth person

• stratified - form into small groups/strata with defined characteristics and then pick random from those groups

• quota - same as stratified but take sample size according to size of stratum

difference between bar chart and histogram?

bar chart — discrete data 

histogram — continuous data

describing shape of histogram:

negatively/left skewed, normal distribution, right/positively skewed

modal class

highest frequencty group in the frequency thing

how to calculate an outlier ?

Q+-  1.5 x IQR 

describing correlations

pearson product moment correlation coefficent

 finds the r and tells correlation. 0-0.25 very weak, 0.25-0,5 is weak, 0.5-0.75 is moderate, 0.75-1 is strong

residual

vertical distance between data point and regression of line