Quantitative data analysis
- 1. POL / SOC 360-01 Spring 2015
- 4. Once we have… Chosen research question Completed literature review Selected appropriate research design ▪ Survey; Experiment; Observation; Content Analysis Gathered data We can use statistics to quantify our results for presentation and publication
- 5. Understand your own and others’ results What do resulting numbers mean? Awareness of statistical results Remember “innumeracy?”
- 7. Array of rows and columns that stores observed values and variables Similar to Excel Spreadsheet What goes into these cells? Review different types of variables
- 8. Nominal: Classification of observations into categories Examples: Religious Faith, Race Ordinal: Observations can be compared by having more or less of a particular attribute; uncertainty of equality between values Example: Olympic Performance (Gold, Silver, Bronze medals) Interval: Intervals between values assigned to observations have meaning and no meaningful zero point Examples: Temperature; Dates Ratio: Interval variable properties with true zero point Examples: Income;Years of Education
- 10. Table showing number of observations and each value of a variable “Lists” each variable’s possible values and how often each occurs
- 11. Raw Frequency Number of observations of a given variable Relative Frequency Number that transforms raw frequency into proportion or percentage ▪ Proportion ▪ Percentage Cumulative Frequency Portion of total that is above or below a certain point
- 12. Too Much Influence Frequency Proportion Relative Frequency Cumulative Frequency StronglyAgree 333 .34 33.5 33.5 Agree 533 .54 53.6 87.1 Uncertain 38 .04 3.8 90.9 Disagree 75 .08 7.5 98.4 Disagree Strongly 16 .02 1.6 100 Totals 995 1.01 100
- 14. Describe characteristics or properties of a set of numbers Two MainTypes: Measures of CentralTendency Measures of Dispersion
- 16. MEAN Locates the middle or center of a distribution Most familiar measure of central tendency; “average” Add values of variable and dividing total by total number of values MEDIAN Divides distribution in half Odd-Numbered Set Even-Numbered Set Most appropriate with ordinal-level data
- 17. Commonly used when dealing with nominal or categorical data Category with the greatest frequency of observations If distribution has one mode = unimodal If distribution has two modes = bimodal If distribution has many modes = multimodal
- 19. No variability (all scores have same value), then variability = 0 Measure will always be positive number (cannot be “less than zero” variation) Greater variability of data, larger the measure
- 20. Largest (maximum) value of variable minus smallest (minimum) value INTERQUARTILE RANGE Divide observations into four equal-sized portions First batch contains 25% of cases, 2nd would have 25%, and so would the 3rd and 4th grouping; division points are called quartiles Finding range, but using 3rd quartile (Q3) as maximum and 1st quartile (Q1) as minimum values RANGE
- 21. INSTRUCTIONS In groups, calculate the various descriptive statistics for each set of numerical data.
- 23. Variance Standard Deviation
- 26. Displays distribution of one variable for each category of another variable Steps to Creating Cross-Tabs: #1: Record respondents’ answers to question #2: Create categories for table #3: Count number of respondents who fall into each category #4: Convert tallies to frequencies; add up row and column tables
- 29. Statistical technique centered on expressing relationship between two quantitative variables with a linear equation Idea of “Best Fit Line” Correlation Coefficient (r) Coefficient of Determination (R2)