Exploring Data with Tables and Graphs: Introductory Statistics Study Notes

Frequency Distributions

Constructing Frequency Distributions

Frequency distributions are fundamental tools for organizing quantitative data into classes or categories. They help reveal patterns and characteristics of the data set.

• Step 1: Select the number of classes (usually 5–20). Convenience and round numbers are preferred.

• Step 2: Calculate class width: $\text{Class width} \approx \frac{\text{Maximum data value} - \text{Minimum data value}}{\text{Number of classes}}$ Round up for convenience.

• Step 3: Choose the first lower class limit (minimum value or a convenient value below it).

• Step 4: List other lower class limits by adding the class width successively.

• Step 5: Enter upper class limits for each class.

• Step 6: Tally each data value in the appropriate class and sum for frequency.

Classes should not overlap, and each value must belong to exactly one class. All classes, even those with zero frequency, should be included.

Relative and Percentage Frequency Distributions

Relative frequency distributions replace class frequencies with proportions or percentages, facilitating comparison between data sets.

• Relative frequency: $\text{Relative frequency} = \frac{\text{Frequency for a class}}{\text{Sum of all frequencies}}$

• Percentage frequency: $\text{Percentage} = \frac{\text{Frequency for a class}}{\text{Sum of all frequencies}} \times 100\%$

• The sum of percentages should be close to 100% (allowing for rounding errors).

Normal and Skewed Distributions

Frequency distributions can reveal the shape of the data, such as normal (bell-shaped), uniform, or skewed distributions.

• Normal Distribution: Frequencies start low, increase to a peak, then decrease, forming a symmetric shape.

• Skewed Distribution: Not symmetric; extends more to one side. Right-skewed (positively skewed) has a longer right tail; left-skewed (negatively skewed) has a longer left tail.

Gaps in distributions may indicate data from multiple populations.

Normal Quantile Plots

Normal quantile plots are used to assess whether a data set follows a normal distribution.

• If points lie close to a straight line, the distribution is normal.

• Systematic deviations from a straight line indicate non-normality.

Graphical Methods for Data Exploration

Histograms

A histogram is a bar graph for quantitative data, where bars are adjacent and represent frequency or relative frequency.

• Horizontal axis: classes of data values

• Vertical axis: frequency or relative frequency

• Objective: Understand data characteristics

Dotplots

Dotplots display each data value as a dot above a horizontal scale. Dots are stacked for repeated values.

• Shows the shape of the distribution

• Allows reconstruction of original data values

Stemplots (Stem-and-Leaf Plots)

Stemplots separate each value into a stem and a leaf, often after rounding data values.

• Shows distribution shape

• Retains original data values

• Data are sorted

Time-Series Graphs

Time-series graphs plot quantitative data collected at different points in time, revealing trends and changes.

• Useful for identifying patterns over time

Bar Graphs

Bar graphs use bars of equal width to show frequencies of categories for categorical (qualitative) data.

• Shows relative distribution of categories

• Bars may be separated by gaps

Pareto Charts

Pareto charts are bar graphs for categorical data, with bars arranged in descending order of frequency.

• Highlights most important categories

• Facilitates comparison of categories

Pie Charts

Pie charts depict categorical data as slices of a circle, with slice size proportional to frequency. They are common but less effective than Pareto charts.

• Shows distribution of categorical data

• Not recommended for statistical analysis due to lack of scale and inefficiency

Frequency Polygons

Frequency polygons use line segments connected to points above class midpoints, similar to histograms but with lines instead of bars.

• Can use relative frequencies for comparison

• Multiple frequency polygons can be plotted for easy comparison

Deceptive Graphs and Pictographs

Graphs can be misleading if the vertical scale does not start at zero or if pictographs exaggerate differences using area or volume.

• Doubling the sides of a square increases area by four times

• Doubling the sides of a cube increases volume by eight times

Correlation and Regression

Scatterplots

Scatterplots display paired (x, y) quantitative data, with the x-axis for the first variable and y-axis for the second.

• Used to assess correlation between variables

• Linear correlation: points form a pattern approximated by a straight line

• Correlation does not imply causation

Regression Line

The regression line (line of best fit) is the straight line that best fits the scatterplot of paired data, often determined by the least-squares method.

• Used for prediction and analysis of relationships

Key Vocabulary

• Center: A value representing the middle of the data set.

• Variation: The degree to which data values differ.

• Distribution: The shape or spread of data values.

• Outliers: Values far from the majority of the data.

• Time: Changes in data characteristics over time.

• Frequency Distribution: Table showing categories and frequencies.

• Lower/Upper Class Limits: Smallest/largest values in each class.

• Class Boundaries: Numbers separating classes without gaps.

• Class Midpoints: Middle value of each class: $\text{Class midpoint} = \frac{\text{Lower class limit} + \text{Upper class limit}}{2}$

• Class Width: Difference between consecutive lower class limits.

• Cumulative Frequency Distribution: Frequency for each class is the sum of that class and all previous classes.

Example: Constructing a cumulative frequency distribution: If the first frequency is 1, the cumulative frequency is 1; if the next is 2, the cumulative frequency is 3, and so on.