Measurement and quantitative data analysis

Numbers in research are only as good as the measurement behind them. A nominal scale merely names (male/female, pass/fail); an ordinal scale ranks (first, second); an interval scale measures equal gaps without a true zero (temperature in Celsius, many test scores); a ratio scale has a true zero (height, weight, time). Worked example: you can average test scores treated as interval data, but averaging ranks or categories misleads — the scale decides which statistics are legitimate.