- getting data: measure a sample to say something about a population
- example: what proportion of UVM undergrads are in CEMS?
- population: all UVM undergrads; sample: the 50 students in STAT2430A
- sample p^=5046=0.92, actual p=109251372=0.126
- way off because STAT2430A is mostly CEMS students, a convenient sample, not a representative one
- the same class gave p^=507=0.14 for out of state, just as untrustworthy
- categorical ( qualitative ) data: sorted into categories, see Categorical Variables
- nominal: no order ( colors, countries, cuisines )
- ordinal: ordered, but the gaps between levels mean nothing ( ratings, rankings )
- quantitative ( numerical ) data: measured or counted
- discrete: countable values ( students in a class, cars in a lot )
- continuous: any value in a range ( height, race time )
- the type picks the analysis: counts, proportions and chi squared tests for categorical; mean, median, standard deviation and quartiles for quantitative