MATH1041 Study Design and Data Collection - Vocabulary Flashcards

A set of vocabulary flashcards covering key terms from Lecture 1–4 on Study Design, data collection, data sources, variable types, observational studies vs. experiments, and experimental designs.

Population

The entire group of items or individuals under consideration in a statistical study, usually defined by the research question.

Cases

The members or units from the population from which information is collected; each case corresponds to a row in a data set.

Observation

The data for a particular case; the values recorded for that case across variables.

IDs/Labels

Identification codes used to uniquely identify each case (e.g., zID).

Data Set

A collection of data arranged as rows (observations) and columns (variables), often stored as a database table.

Sample

A subset of the population actually examined to gain information about the whole.

Sample Size (n)

The number of cases in the sample.

Variable

A characteristic of the cases that can be measured, recorded or counted; each column in a data set is a variable.

Number of Variables (p)

The total number of variables recorded in the data set.

Data Set Formats

Common file formats for data: TXT, CSV, XLS, XLSX, DAT; data can be ASCII or binary.

Anecdotal Data

Information collected casually or informally from personal experiences; not systematic or representative.

Available Data

Data produced in the past for some other purpose but usable for current analysis.

Collecting Your Own Data

Data collected directly by the researcher or generated via simulation; focus on data collection design.

Census

Systematically acquiring and recording information about every item in a population.

Sample Survey

Collecting data on people by asking questions, either verbally or via questionnaires; a type of observational study.

Voluntary Response Sample

A sample where individuals choose to respond; tends to be biased toward strong opinions.

Convenience Sampling

A non-probability sampling method where units are chosen for ease of access.

Quantitative Variable

A variable that takes numerical values where arithmetic makes sense; units matter.

Categorical Variable

A variable that places an individual into one of several categories; values are non-numeric labels.

Continuous Quantitative

A quantitative variable that can take any value in an interval (e.g., height, time).

Discrete Quantitative

A quantitative variable that takes countable values (e.g., number of siblings).

Ordinal

A categorical variable with a natural order but not necessarily equal intervals between levels.

Explanatory Variable

The variable used to explain or predict the response; also called the independent variable.

Response Variable

The outcome of interest that is measured; also called the dependent variable.

Lurking Variable

An unobserved variable that may influence the interpretation of relationships between measured variables; when observed, it is called a covariable.

Covariable

A variable that is observed and may influence the interpretation of relationships between other variables.

Confounding Variable

An unobserved variable that influences the response and is related to the explanatory variable, making causal attribution difficult.

Observational Study

A study where individuals are observed and measured without deliberate intervention.

Experiment

A study where a treatment or intervention is deliberately imposed to observe responses; aimed at establishing causation.

Common Response

Two variables change due to a shared underlying cause, creating an apparent association.

Causation

A cause-and-effect relationship where changes in one variable produce changes in another; best demonstrated by well-designed experiments.

Spurious Correlation

An apparent association that is due to confounding, common response, or chance rather than a direct causal link.

Randomised Comparative Experiment

An experiment in which subjects are randomly allocated to several treatments and responses are compared across groups.

Matched Pairs Design

A design that pairs similar subjects and applies two treatments within each pair to improve precision.

Randomised Block Design

Subjects are grouped into blocks that are similar in some way; randomisation occurs within each block.

Subjects/Experimental Units

Individuals on whom the experiment is conducted and to whom treatments are applied.

Factor

An explanatory variable in an experiment that is manipulated across levels.

Levels

Different conditions or categories of a factor.

Treatment

A specific experimental condition applied to subjects, formed by combining levels of factors.

Control Group

A group that does not receive the treatment, used for comparison; may use a placebo.

Placebo

A sham treatment used as a control to account for placebo effects.

Double-Blind

A study design in which neither the subjects nor the administrators know which treatment is being given.

Compare, Randomise, Repeat, Replicate

Four core principles of good experiments: compare treatments, randomise allocation, repeat the treatment, and replicate the experiment.