Overview
This lecture covers rounding rules, levels of measurement in data (nominal, ordinal, interval, ratio), and how to construct and interpret frequency tables in statistics.
Rounding Rules
- Round final statistical answers to one more decimal place than the original data.
- Avoid rounding during intermediate steps; round only at the final answer.
- Rely on technology to minimize rounding errors.
Levels of Measurement
- Four levels: nominal, ordinal, interval, and ratio.
- Nominal: Data are categories only, no meaningful order or arithmetic possible.
- Ordinal: Data can be ordered, but differences between values are not meaningful.
- Interval: Ordered data with meaningful differences; zero is arbitrary, so ratios don’t make sense.
- Ratio: Ordered, meaningful differences and true zero; ratios are meaningful.
Frequency Tables
- Frequency tables count occurrences of each data value in a set.
- Construct by counting how often each number appears and listing these frequencies.
- Can be visualized with dot plots for small data sets.
Relative and Cumulative Frequency
- Relative frequency = frequency of a value ÷ total number of values, often shown as a percentage.
- Cumulative frequency adds frequencies as you go down the table, reaching 100% at the end.
- Cumulative relative frequency tracks the running total of percentages.
Grouped Frequency Tables
- Used for continuous data, with class intervals (e.g., height ranges).
- Intervals should not overlap to avoid misclassification.
- Example: recording soccer players’ heights in intervals and tallying frequencies.
Example: Frequency Table Applications
- Tables can summarize event counts, like earthquake deaths by year.
- To find frequencies or percentages for specific years, sum relevant rows and divide by total.
Key Terms & Definitions
- Frequency Table — A table showing how often each value occurs in a dataset.
- Relative Frequency — The proportion of times a value occurs, calculated as frequency divided by total number of data points.
- Cumulative Frequency — Running total of frequencies up to a certain point in the data.
- Nominal Scale — Categorical data without order.
- Ordinal Scale — Categorical data with order, but no meaningful differences.
- Interval Scale — Ordered numeric data with meaningful differences, but arbitrary zero.
- Ratio Scale — Ordered numeric data with meaningful differences and an absolute zero.
Action Items / Next Steps
- Review textbook example on earthquake deaths and check solutions provided.
- Prepare to construct your own frequency tables in the next chapter.
- Practice identifying data levels of measurement in sample datasets.