SCI110: Summarising qualitative data and between-individuals - Vocabulary

Key vocabulary from lecture notes on summarising qualitative data and comparing between individuals (nominal/ordinal data, graphs, proportions, odds, and two-way tables).

distribution

A summary of qualitative data describing how observations are spread across levels of a variable.

frequency table

A table listing each level of a qualitative variable with counts or percentages for observations in that level.

qualitative data

Data expressed as categories or labels rather than numbers.

nominal data

Qualitative data with no natural order among levels (e.g., gender).

ordinal data

Qualitative data with a natural, meaningful order among levels (e.g., age groups, safety ratings).

exhaustive

Levels cover all possible values of the variable.

exclusive

Observations belong to one and only one level.

levels

The categories or values that a qualitative variable can take.

natural order

The inherent order of ordinal levels (e.g., 18–24, 25–34).

proportion

A fraction of the total; a value between 0 and 1 representing part of the whole.

percentage

A proportion multiplied by 100; a value between 0% and 100%.

population proportion (p)

The true proportion in the population (unknown in practice).

sample proportion (p̂)

The proportion estimated from a sample; a statistic.

odds

The ratio of the number of times an event happens to the number of times it does not happen.

odds ratio (OR)

The ratio of the odds of an event in one group to the odds in another group.

mode

The most frequent level(s) in a qualitative variable.

median (ordinal data)

The middle level after ordering the data; applicable to ordinal data only.

two-way table

A cross-tabulation of two qualitative variables showing counts for each combination.

confounding variable

A variable that distorts the apparent relationship between the explanatory and response variables.

Simpson’s paradox

A paradox where a trend appears within groups but reverses when groups are combined.

bar chart

A graph showing counts or percentages for each category with bars; gaps indicate discrete categories.

dot chart

A graph using dots to represent counts or percentages at each level.

pie chart

A circle divided into segments proportional to the counts of each level; best when all levels are present.

stacked bar chart

A bar chart where levels of a second variable are stacked within each level of the first variable.

side-by-side bar chart

A bar chart showing bars for different groups placed next to each other for comparison.

summary tables by rows and columns

Tables summarising data by row or column using percentages, proportions, or odds.

odds of an event

The ratio of the probability the event happens to the probability it does not.

parameter

A population value that is usually unknown and estimated by a statistic.

statistic

A numerical value computed from a sample to estimate a population parameter.