Medium
ANOVA (One-Way and Two-Way)
SPSS
Free
A lecturer compares student satisfaction across three teaching formats (Recorded, Live Zoom, In-person). Group sizes are unequal. SPSS shows:
Levene’s Test: p = .003 Omnibus ANOVA (standard): F = 4.90, p = .010 Welch ANOVA: F = 6.15, p = .004
Which post-hoc approach is MOST appropriate to identify which formats differ?
A
Use Tukey HSD, because it is the default post-hoc after any significant ANOVA.
B
Use Games-Howell, because Levene’s is significant and group sizes are unequal.
C
Use Fisher’s LSD, because it has the highest power and the ANOVA is significant.
D
Do no post-hoc tests; the significant ANOVA already tells you which groups differ.