Stats Terminology

summary statistic

a single number summarizing a large amount of data. Usually percent format

Principles of Randomized Experiments

controlling, randomization, replication, blocking

observational data

data where no treatment has been explicitly applied or withheld. No causal conclusions

parameter

numerically describes a population characteristic

anecdotal evidence

data collected in a haphazard way. Faulty/erroneous

population data

all data possible in a specific source. A collection of all outcomes, responses, measurements, or counts of interest

observational study

when data is collected in a way that doesn’t directly interfere with how the data arise

nominal

a categorical variable without special ordering

numerical/quantitative variable

takes a wide range of numerical values. Think whether or not average could be taken

observational unit

formal name for a row in a table

confounding variable

correlates with both explanatory and response variables

experiment characteristics

control group, treatment group, must be replicable, blocking to hone in

experiment

conducted when one wishes to investigate the possibility of a causal connection

Statistics

science of collecting, organizing, analyzing and interpreting data to make decisions

discrete

a subtype of nunmerical variables; only can take numerical values with jumps

data matrix

convenient and common way to organize data, especially if collecting data in a spreadsheet. Each row corresponds to an observational unit; each column with a variable

sampling error

difference between parameter and statistic. Goal is to reduce the difference.

statistic

a numerical description of sample characteristics

sample

a subset or part of a population, when chosen well, eliminates bias

descriptive vs inferential statistics

collecting, organizing, analyzing data

vs

interpreting data to make decisions

data

info coming from observations, counts, measurements, or responses

ordinal

a categorical variable whose levels have a natural ordering

continuous

a subtype of numerical variables; can take numerical values without jumps

categorical/qualitative variable

when responses themselves are categories

associated variable

when two variables show some connection to each other

negative vs positive association

negative is a downward trend while positive is an upward trend

independent

if two variables aren’t associated and there’s no evident relationship between the two