uwo psych 2801 - lesson 4

• types of graphs:

  • pie charts and bar graphs
  • histograms
  • scatterplots and line graphs
  • time-series graphs
• types of data:

  • categorical
  • numerical
• categorical data: each value represents a discrete category

  • order DOES NOT matter
• numerical data: each value represents either a real number or a place on a continuum

  • order DOES matter
• pie chart: used for categorical data

  • displaying relative frequencies (part of a whole)/proportion
• implications of pie charts:

  • not great for absolute frequencies
  • best for simple categorical data
• bar graphs: series of categories by representing each as a bar

  • for complex categorical data
  • for absolute frequencies associated with a category
  • preferred method for categorical data
• categorical data graphs: pie charts and bar charts
• numerical data charts: histograms, scatterplots and time-series graphs
• histogram: shows the distribution of numerical values

  • shows mean, range, skew and possible outliers
• you can find out if variable is normally distributed using a histogram
• binning: process of choosing interval for x axis - effects distribution
• lower bin: makes data noisier
• bin too big: lose valuable info
• types of numerical data:

  • discrete
  • continuous
• discrete numerical data: finite number of values
• continuous numerical data: infinite number of values
• continuous variables are usually assumed to be normally distributed
• if a continuous variable is not normally distributed, you can log transform it by taking the log 10 of each variable
• scatterplot: shows how TWO numerical variables are associated with each other - works well with two CONTINUOUS variables
• instead of log transforming a continuous variable, you can bin it into discrete numerical values

  • helps people understand better
  • simpler
  • helps fix normal distribution pattern
• line graph: ONE continuous and ONE discrete variable - shows how they are associated
• in a line graph, the x axis is usually the discrete variable, and the y axis is usually the continuous variable
• time series graph: type of LINE GRAPH that shows how something changes over time

  • x axis: discrete, y axis: continous
• importance of y axis: usually contains dependent measure

  • truncating axis (not starting at baseline) can exaggerate differences
  • too broad of range can minimize differences (zoomed out so lose variation)
  • flip y axis around - implies wrong message
• importance of x axis: usually associated with time

  • if you only present a portion - changes look of graph
  • exaggerate changes by zooming in
• recap of x and y axis importance:

  • truncating axis
  • expanding axis
  • ignoring conventions
  • comparing non-equivalent data (two different y axis on same graph)