Stats Unit One

statistics is the science of

variability

saying: what was

compared

saying: who’s not

here

saying: incorporate

ish-ness

data is

a representation of someone or something

tidy data is a way of

mapping the real world to a dataset

observations are the

things we are interested in

attributes

the pieces of information we are interested in

measures are the way

we collect information about observations

quantitative data

values of an attribute for an observation are numbers representing a quantity of something

categorical data

values for an attribute for an observation are selected from a set of different category labels

rating scale data

special type of categorical data, values of an attribute for an observation are selected from a predetermined scale

time series data

the values of an attribute for an observation indicate a moment in time

reliability refers to

the extent to which th data you collect from a measure truly represents and reflects the real world

data validation

the act ensuring the values collected from each observation for each attribute are reliable

variation is

universal

statistical modeling/statistical testing

building a model based on what you know, considering the predictions the model makes and comparing those predictions to what actually is happening

statistical investigative cycle

problem, plan, data, analysis, conclusion

problem -

identify a statistical question

plan -

choose a sample design, study design, and measures

data -

collect and process data

analysis -

look for patterns with summary tables, graphs, and statistical models

conclusion

interpret the results and generate new questions about the real world context

data is a representation

of someone or something

tidy data is a way of mapping the real world to a

dataset

observations are the

things we are interested in

attributes are the

pieces of information we are interested in

in tidy datasets, there is one row for

each observation

one observation

per row

one attribute

per column

rows are the

things we are interested in

types of data

quantitative, categorical, rating scale, test, time series

if someone is writing in a value then it is

text data

data cleaning

removing invalid values from a dataset

prospective data cleaning

build checks into the survey - preferred method

retrospective data cleaning

clean the data after, remove things

quantitative examples

temperature, blood pressure

categorical examples

hair color, blood type, country of origin

rating scale example

how much pain are you in between 1 and 5

text examples

customer reviews, social media posts

time series examples

date of birth, year, financial quarter

type of measure: quantitative

key giveaway: “how many”, values with decimals, scales

type of measure: categorical

key giveaway: selected from a drop down list, no order to the values

type of measure: rating scale

key giveaway: clear order to the values

type of measure: text

key giveaway: write-in data

type of measure: time series

key giveaway: “when?”