Statistics

Statistics

1. Descriptive

   • Summary of data
     • Typically found in results and methods section
       • Describes participants and variables of interest
       • Makes numbers and data easier to understand
     • UMBRELLA
       • Examines distribution of data
         • Measures of central tendency
           • Single number to describe "typical value" for given variable
             • Mean=average score
               • Arithmetic average
               • Greatly influenced by extreme values
               • Most often used form, but not always the best
             • Mode=most often value
               • Nominal data
               • Usually reported as frequency
             • Median=exact middle score
               • UNAFFECTED BY EXTREME VALUES
   • Purpose: organize data and gain general impression of data
   • Questions: "What does my data look like?" "What is the typical value?"
     • How much spread/variability?
       • Range
       • Variablity
       • Standard deviation
         • Accounts for individuality
         • Large SD=large difference between mean and scores
         • Small SD=no large difference between mean and scores
   • Frequency Distribution
     • List all values and calculate how often they occur
     • Easy way to display data
       • Histogram

2. Inferential

   • Used to examine relationship between 2+ variables and group differences between variables
   • Significance testing
     • p-value=0.05
     • Null hypothesis-no difference between groups
     • Alternative hypothesis-difference between groups
     • Ruling out Chance
       • Write out hypothesis
         • Compare results to null hypothesis
           • Determine p-value
             • For significance, p-value should be below alpha levels

3. Statistical Tests

   • T-Tests
     • Answers if mean differences exist between two groups on a continuous variable
       • One sample t-test
         • Comparing mean to known target/reference value
       • Paired sample t-test
         • Comparing means from two observations
       • Independent samples t-test
         • Compares means of 2 separate samples
         • Not dependent on each other
         • Used in intervention/experimental studies