Data Lab # 2 Descriptive Statistics May 28, 2008 Ivan Katchanovski, Ph.D. POL 242Y-Y
Measures of Central Tendency: The Mode The mode is the only measure of central tendency for nominal variables The mode can also be found for ordinal and interval- ratio variables Example: select a nominal variable – colony (World dataset) SPSS Command: – Analyze-Descriptive Statistics-Frequencies-Statistics-Mode – Identify the mode by a value in statistics box – Identify the mode by a value with the highest frequency – Identify the mode by a value with the highest percentage in “valid percent” column 2
Measures of Central Tendency: The Median The median is appropriate for ordinal and interval- ratio variables The median is not appropriate for nominal variables Example: Select an ordinal variable – Economic development (World dataset) SPSS Command: – Analyze-Descriptive Statistics-Frequencies-Median – Identify the median by a value in statistics box – Identify the median by a value below which 50% of cases lie in “cumulative percent” column 3
Measures of Central Tendency: The Mean The mean is only appropriate for interval-ratio variables Example: Select interval ratio variable – GDP per capita $1000 (World dataset) SPSS Command: – Analyze-Descriptive Statistics-Frequencies- Statistics-Mean – Identify the mean by a value in statistics box 4
Graphic Description of Variables: Bar Chart Bar chart – Most appropriate for nominal and ordinal variables – Shows dispersion of a variable Example: select a nominal variable – colony (World dataset) SPSS Commands: – Analyze-Descriptive Statistics-Frequencies-Charts-Bar charts-Percentages – Open Chart Editor: Select chart-Edit-SPSS Chart Object- Open – In Chart Editor: Options-Title – In Chart Editor: Elements-Show Data Labels 5
Graphic Description of Variables: Pie Chart Pie chart – Most appropriate for nominal and ordinal variables with small number of values (categories) Example: select an ordinal variable – economic development (World dataset) SPSS Commands: – Analyze-Descriptive Statistics-Frequencies-Charts-Pie charts- Percentages – Open Chart Editor: Select chart-Edit-SPSS Chart Object- Open – In Chart Editor: Options-Title – In Chart Editor: Elements-Show Data Labels 6
Graphic Description of Variables: Histogram Histogram – Most appropriate for interval-ratio variables with large number of values – Shows dispersion of a variable Example: select an interval-ratio variable – GDP per capita $1000 (World dataset) SPSS Commands: – Analyze-Descriptive Statistics-Frequencies-Charts-Histograms – Open Chart Editor: Select chart-Edit-SPSS Chart Object-Open – In Chart Editor: Options-Title – In Chart Editor: Elements-Show Data Labels 7
Skewness of Distributions Skewness of distributions – Appropriate for interval-ratio variables Example: select an interval-ratio variable – GDP per capita $1000 (World dataset) SPSS Commands: – Analyze-Descriptive Statistics-Frequencies-Statistics- Skewness/Mean/Median – Unskewed distribution if skewness statistics=0 or mean=median – Positively skewed distribution if skewness statistics>0 or mean>median – Negatively skewed distribution if skewness statistics<0 or mean<median 8
Helpful SPSS Options SPSS option for displaying variable labels instead of variable names – Open the 2000 World Values Survey Dataset – Edit-Options-General-Display labels SPSS option for selecting cases Example: selecting Canadian respondents from the World Values Survey dataset – Identify value for Canada (12) in “nation” variable (v2) – Data-Select cases-If: v2=12 9
Dependent and independent variables Describing dependent and independent variables – The 2000 World Values Survey – Questionnaires – Country specific questionnaires online at “Documentation of the Values Surveys” & “ wave” Select Canadian questionnaire and Canadian sample of the 2000 World Values Survey dataset – Identify dependent variable – Identify independent variable – Produce appropriate descriptive statistics and charts 10