Statistics I, VL01

Ratio scale

- assigned values are geniune numbers, "true zero" exists

- order, differences and ratios are meaningful

- data can be categorised, ranked, differences and ratios quantified

- allowed data transformations: all similarity or shape-preserving transformations

Statistical unit

member of the set of entities

Population

set of all statistical units relevant for the research question at hand

Subpopulation

subset from the population that shares certain properties

Sample

subset from the population being investigated

Variable

a unit's characteristic of interest

Value

observed outcome of a variable

Discrete variable

finite or countably infinite number of values

Continuous variable

uncountably or infinite number of values

Quasi-continuous variable

continuous variable that is measured discretely due to limitations in measurement

Quasi-continuous variable

discrete variable that is treatable as if it were continuous

Nominal scale

- values assigned as mere labels of qualitative attributes (numeric code)

- only categorisation: equal or unequal

- allowed data transformation: informations-preserving transformation, including all bijective transformations

Ordinal scale

- order matters but not the difference between values

- data can be categorised or ranked as better or worse

- allowed data transformation: all strictly monotonous or rank-preserving transformations

Interval scale

- assigned values are geniune numbers, order and differences are meaningful, scale has no "true zero"

- data can be categorised and ranked, differences can be quantified

- allowed data transformations: all linear or distance-preserving transformation

Qualitative variable

takes on a finite number of values (ordinal at most), refer to characteristics that are not quantifiable

Quantative variable

genuine numbers, refer to quantifiable characterists

Validity

to which extent a variable actually measures what it was designed to measure

Reliability

to which extent repeated measurements under the same condition yield the same results

census

gathers information from all units in a population

sample

gathers information from a sample of units that is representative for the entire population

cross-sectional data

data collected at one point in time

time-series data

repeated measurements over time

panel data/longitudinal data

follows the same set of units/panelists over time, collecting the same variable at differents points of time

primary data

data you have collected yourself from original sources

secondary data

data coming from existing surveys or administrative sources

Wird gerade gelernt (1)

Diese Begriffe lernst du gerade. Immer weiter so!