Inferential Statistics

Inferential Statistics

1. Mann Whitney U

   • Independent Groups
   • At least Ordinal
   • Na: number of participants in condition A
   • Nb: number of participants in condition B
   • Ra: Sum of ranks for condition A
   • Rb: Sum of ranks for condition B
   • To determine Significance:
   • 1- Select appropriate P level (0.05)
   • 2- read across to Na and down to Nb
   • 3- significant if ≤ the critical value

2. Wilcoxin Signed Ranks

   • Repeated measures/matched pairs
   • At Least Ordinal
   • N: Number of participants excluding any with no difference between conditions
   • To Determine Significance:
   • Read across p and down N. The intersection is the critical value.
   • Significant if ≤ the critical value

3. Spearman's Rho

   • Correlational
   • At least Ordinal
   • Involves ranking each variable separately. ∑d^2 is each participants variable 1 rank - variable 2 rank squared and added up
   • To determine Significance:
     • Read across p and down N
       • If observed value is ≥ critical value, it is significant
   • The closer the observed value is to 1, the stronger the positive correlation is
   • The closer the observed value is to -1, the stronger the negative correlation

4. Binomial Signs

   • Nominal
   • Repeated measures/Matched pairs
   • Count the number of positive differences and the number of negative differences - Whichever of these is smaller is the observed value
   • To determine significance
     • 1- Select appropriate p
       • 2- Find N (no. of participants excluding ones with no difference between each conditions
         • 3- Find the critical at the intersection
           • 4- If the observed value is ≤ the critical value, results are significant

5. Chi Squared

   • Independent Groups
   • Nominal
   • O: Observed Frequency
   • E: Expected Difference (calculated by (row total x column total)/ Overall total))
   • DF: Degrees of freedom calculated by (no. of rows - 1) x (no. of columns - 1)
   • To Determine Significance:
     • 1- Find p
       • 2- Select the df
         • 3- find the critical value (intersection between p and df
           • 4- If the observed value is ≥ the critical value, results are significant so null hypothesis can be rejected