Descriptives

Last updated:
06 Mar 2007

Assumed knowledge

  • Francis Chapter 2 "Exploring the Distribution of a Single Variable"

Descriptive & Graphical Exercise Using SPSS

  • Open qfsall.sav (collected using the Quick Fun Survey (html)) and conduct the univariate and bivariate descriptive and graphical analyses outlined below.
  • Hint: If you have no idea where to start, you can download the syntax file (Descriptives.sps) and run it - or go straight to the output file (Descriptives.spo).  You can learn by following through each analysis in the syntax and output.  You should develop a sufficient understanding of univariate and bivariate descriptives, graphs, and SPSS to be able to produce this kind of output for each question from scratch.)

Univariate

  • Determine the level of measurement and obtain appropriate univariate descriptives and graphs to depict the nature of responses to each of the following variables:
    1. a8 (What is your favourite season?)
    2. b2 (Do you snore?)
    3. b4 (Do you believe in God?)
    4. b9 (At what age do you think you will die?)
    5. b13 (Femininity-Masculinity)
  • Statistically, you may consider using:
    • frequency / % / mode
    • median
    • mean, SD, skewness, kurtosis
  • Graphically, you may consider using:
    • bar graph
    • pie chart
    • stem and leaf plot
    • boxplot
    • histogram
    • line graph
    • error bar

Bivariate

  • Determine the level of measurement and examine bivariate distributions for the following relationships:
    (Note that unless you are an experienced statistician you probably should conduct univariate descriptives and graphs for each variable before using them in bivariate and multivariate analyses.)
    1. a1 (Gender) and a4 (Favourite colour)
      (nominal by nominal)
      • crosstabs
      • clustered bar graph
    2. b10 (Psychology student status) and b7 (Exercise)
      (nominal by ratio)
      • means table
      • bar graph (means)
    3. b14 (Physical health) and b15 (Mental health)
      (interval by interval)
      • correlation
      • scatterplot - edit by going into chart editor, double-clicking on a data point and changing to bins to represent multiple data points.
    4. b7 (Exercise) and b9 (At what age do you think you will die?)
      (ratio by ratio)
      • correlation
      • scatterplot