Number Concepts and Reasoning

Any positive whole number

A number that cannot be written as a fraction of integers

A number expressible as a ratio of two integers with a nonzero denominator

A number only used in counting objects

Round all values to whole numbers regardless of context

Prefer exact fractions in every applied setting

Use appropriate significant figures and report uncertainty

Always show every digit provided by the calculator

Prime factorization structure

Number of digits in decimal form

Color assigned when printed

Whether digits are increasing

The data are uniformly distributed

The sample size must be one

The distribution is skewed or has outliers affecting the mean

Both mean and median are invalid

Only use visual charts to claim causality

Check for confounders and seek controlled or longitudinal evidence

Ignore sample size when correlation exists

Assume causation when correlation is strong

Estimate only and never compute exact answers

Multiply digits individually then add without carrying

Always convert to decimals before multiplying

Use a nearby round number and adjust (e.g., 100×47 − 47)

Sample size has no impact on estimates

Bigger samples always eliminate bias

Larger samples always produce smaller p-values

Big samples are never useful

Pick the model with fewer parameters without testing

Compare only training loss values

Choose the model with the most features enabled

Evaluate out-of-sample performance with a validation set