Visualizing Data and Distribution Characteristics
Illustrating Data and Distribution Characteristics
Why Illustrate Data?
- Visual representations (charts, graphs) are more effective than numbers alone for examining data distribution characteristics.
- They help understand central tendency and variability visually.
Principles for Chart Design
- Minimize "chart junk"; less is more.
- Plan charts before creation.
- Be clear and concise; avoid clutter.
- Label everything clearly.
- Communicate only one idea per graph.
- Keep visuals balanced (centered titles and labels).
- Maintain scale (e.g., 3 inches wide to 4 inches tall).
- Charts should be simple, conveying information straightforwardly and able to stand alone.
- Limit text to avoid distracting from the visual message.
Frequency Distributions
- A method of tallying and representing how often scores occur.
- Scores are usually grouped into class intervals (ranges of numbers).
- Rules for Class Intervals:
- Select a range of 2, 5, 10, 15, or 20 data points.
- Select 10 to 20 intervals to cover the entire data range (e.g., Range / Number of Intervals).
- Begin the lowest interval with a multiple of the interval size.
- The largest interval goes at the top of the distribution table.
Visualizing Frequency Distributions
- Histogram: A visual representation of a frequency distribution using bars.
- Frequencies are on the y-axis, class intervals (or their midpoints) on the x-axis.
- Frequency Polygon: A continuous line representing frequencies of scores within class intervals.
- Created by connecting the midpoints at the top of each bar in a histogram.
- Cumulative Frequency Distribution (Ogive): Shows the cumulative frequency of occurrences by class intervals.
- Created by adding the frequency of a class interval to all frequencies below it.
- Often plotted using class interval midpoints against cumulative frequency.
Characteristics of Distributions
- Distributions can differ in four ways: average value, variability, skewness, and kurtosis.
- Average Value: Different means shift the entire distribution along the x-axis.
- Variability: Refers to the spread or dispersion of scores (e.g., standard deviation).
- Skewness: A measure of the lack of symmetry (lopsidedness) of a distribution.
- Positively Skewed: Right tail is longer; more occurrences at lower scores. Mean > Median. Formula: Sk=s3(Xˉ−M)
- Negatively Skewed: Left tail is longer; more occurrences at higher scores. Mean < Median.
- Kurtosis: How flat or peaked a distribution appears.
- Platykurtic: Relatively flat distribution.
- Leptokurtic: Relatively peaked distribution.
- Mesokurtic: Normal or bell-shaped distribution (no kurtosis). Formula: K=n∑(sX−Xˉ)4−3
Types of Charts
- Column Chart: Compares frequencies of different categories; categories on x-axis, values on y-axis.
- Bar Chart: Identical to a column chart but categories are on the y-axis, values on the x-axis.
- Line Chart: Shows trends in data at equal intervals (e.g., over time).
- Pie Chart: Shows the proportion of an item that makes up a series of data points relative to the total.
Using Software (SPSS) for Charts
- General steps: Graphs → Chart Builder → Select chart type/design → Drag variables to axes → OK.
- Specific options for Histograms, Bar Graphs, Line Graphs, and Pie Charts are available.