Inferential statistics

Inferential statistics

1. Nominal

   • Category data
   • Put in 2 or more groups
   • Doesn't give much info

2. Ordinal

   • Data for individual ppts placed in order
   • doesn't tell difference between each score

3. Interval

   • Data for participants on equal scale
     • produces lots of data
   • Most informative

4. Use of statistical tests

   • Decide which hypotheses to accept/reject

5. Significance testing

   • Differences due to chance/variables
   • P values - number statistical test gives you
     • < (or equal to) 0.05, accept alternative
     • >0.05, accept null
   • level of significance accepted: P< (or equal to) 0.05
     • If its 0.05, we are 95% sure the difference is due to variables

6. Parametric tests

   • Done when there is no anomolies
     • INTERVAL DATA - set scale
     • Groups are normal in variences

7. ERRORS

   • Type 1 - accepting alternate when null is true (false positive)
     • Type 2 - accepting null when alternate was true (false negative)

8. Distibution curves and coefficients

   • nominal
     • mean, median, mode all together at centre, 50% scores each side
   • Skewed
     • Mode is highest point on curve, median is middle, 50% scores each side, mean further along (takes into account every score)
     • Negative
       • Skew to left
       • positive
         • Skew to right