Exploring Data with Tables and Graphs

1. Frequency Distributions for Organizing and Summarizing Data

Frequency distributions are essential tools in statistics for organizing raw data into a more interpretable format. They display how often each value (or range of values) occurs in a dataset, making it easier to identify patterns and trends.

• Frequency Distribution: A table that lists data values (either individually or by intervals) alongside their corresponding frequencies (counts).

• Relative Frequency: The proportion of observations within a category, calculated as

• Purpose: To summarize large datasets, making them easier to analyze and interpret.

Example: Number of TVs in Households

Suppose we record the number of TVs in 50 randomly selected households. The data can be summarized in a frequency and relative frequency table:

2. Histograms—for Quantitative Data

A histogram is a graphical representation of the distribution of quantitative data. It uses adjacent bars to show the frequency or relative frequency of data within specified intervals (bins or classes).

• Definition: A graph consisting of bars of equal width drawn adjacent to each other (unless there are gaps in the data).

• Horizontal Axis: Represents classes of quantitative data values.

• Vertical Axis: Represents frequencies or relative frequencies.

• Bar Heights: Correspond to the frequency or relative frequency values for each class.

Important Uses of a Histogram

• Displays the shape of the data distribution.

• Shows the center of the data.

• Shows the spread (variation) of the data.

• Identifies outliers in the data.

Example: Frequency and Relative-Frequency Histograms

Using the TV data above, we can construct:

• Frequency Histogram: Bars represent the number of households for each number of TVs.

• Relative-Frequency Histogram: Bars represent the proportion of households for each number of TVs.

Example: Grouped Data (Class Width 10)

For data such as days to maturity for investments, we may use class intervals (e.g., 0-9, 10-19, etc.) with a specified class width. The frequency and relative frequency for each class are tabulated:

3. Interpreting Histograms: The CVDOT Approach

Critical thinking is required to interpret histograms effectively. The acronym CVDOT helps remember the key aspects to analyze:

• Center: Where is the middle of the data?

• Variation: How spread out is the data?

• Distribution: What is the overall shape (e.g., symmetric, skewed)?

• Outliers: Are there any data points that stand out?

• Time: If data is collected over time, are there trends or changes?

4. Common Distribution Shapes

The shape of a histogram provides insight into the underlying distribution of the data.

• Normal Distribution: A symmetric, bell-shaped curve. Most data clusters around the center, with frequencies tapering off equally on both sides.

• Skewed Right (Positively Skewed): The right tail (higher values) is longer; most data is concentrated on the left.

• Skewed Left (Negatively Skewed): The left tail (lower values) is longer; most data is concentrated on the right.

5. Assessing Normality with Normal Quantile Plots (QQ-Plots)

Normal quantile plots (also called QQ-plots) are graphical tools used to assess whether a dataset follows a normal distribution.

• Normal Distribution: The points in the QQ-plot lie reasonably close to a straight line, with no systematic deviations.

• Not Normal: The points do not lie close to a straight line, or they show a systematic pattern (e.g., curve, S-shape) that deviates from linearity.

Steps for Constructing a QQ-Plot:

1. Order the data from smallest to largest.

2. Calculate the expected z-scores for a normal distribution.

3. Plot the actual data values against the expected z-scores.

4. Assess the linearity of the plot.

Criteria for Assessing Normality:

• If the points are close to a straight line, the data is approximately normal.

• If the points deviate systematically from a straight line, the data is not normal.

Additional info: QQ-plots are especially useful for checking the normality assumption before applying statistical tests that require normality, such as t-tests or ANOVA.