Statistical data analysis

Which two parameters uniquely define a normal distribution?

The mean and the variance

Why does a statistically significant result not necessarily imply practical significance?

Very large sample sizes can yield low p-values that have no practical importance

What is the null hypothesis (H0_0)

A baseline claim, often that there is no difference, association or effect

What is a location statistic in descriptive analysis?

A numerical summary representing a typical value of observed data

Why are categorical variables always classified as discrete?

They have a countable distinct set of possible values

Difference between an ordinal variable and a nominal variable?

A nominal variable has no natural ordering a ordinal variable has natural order

Cross-sectional study vs longitudinal study?

Cross-sectional: observes units at a single point in time, longitudinal: observes the same units across multiple time points

Statistical model parameter vs sample statistic?

Model parameter: fixed property of the population, sample: random variable computed from sample data

Which statistical test should be used to compare two paired, normally distributed numerical measurements from the same subjects?

The paired Student's t-test

Data Imputation

The process of replacing missing or invalid data values with substituted values

Statistical Unit

A single member or real-world entity of a population from which data is observed

Spearman vs Pearson

Sperman evaluates ranks, Pearson evaluates raw values

What range of values can Pearson's correlation coefficient r take?

-1-1

Which statistical test evaluates paired nominal binary data (e.g., pre- vs. post-treatment response)?

McNemar's test should be used

Independent vs dependent variable?

independent variable explains variation in the outcome, dependent variable is the primary variable of interest being analyzed

When comparing two unpaired nominal variables with expected cell counts under 5, which exact test replaces Pearson's chi-squared test?

Fisher's exact test

Point estimator vs interval estimator?

point estimator provides a single calculated value, interval estimator provides a range of plausible parameter values.

Which test assesses the association between two independent categorical variables summarized in a contingency table?

Pearson's chi-squared

Why use Benjamini-Hochberg procedure in multiple testing?

It controls the false discovery rate

Which non-parametric test is suitable for comparing two paired numerical samples that violate the normality assumption?

Wilcoxon signed-rank test

Which statistical test should be used to compare the means of two independent, normally distributed numerical samples?

Two-sample unpaired Student's t-test

What relationship does Pearson’s correlation r measure?

the strength and direction between two variables

Which parametric test generalises the Student's t-test to compare sample means across more than two independent groups?

One-way ANOVA

Type II Error (β)

The error committed when a false null hypothesis fails to be rejected

When comparing two independent numerical samples that are not normally distributed, which non-parametric test is appropriate?

Mann-Whitney U test

What is the null hypothesis of the Shapiro-Wilk test?

Sampled population is normally distributed

Type I Error (α)

The error committed when a true null hypothesis is incorrectly rejected

Descriptive statistics

Summarize observed data

Inferential statistics

Conclusions about populations

Categorical variable

Describes groups or labels eg blood type

Discrete variable

Takes distinct countable values

Continous variables

Takes any value in range