In: Statistics and Probability
Factor B: supplement
Supplement |
Placebo |
||
M =17.38 |
M =11.85 |
Trow =380 |
|
Zombie |
T =226 |
T =154 |
nrow =26 |
SS =563.08 |
SS =369.69 |
Mrow =14.62 |
|
n =13 |
n =13 |
SDrow =6.73 |
|
M =26.38 |
M =21.46 |
Trow =622 |
|
Vampire |
T =343 |
T =279 |
nrow =26 |
SS =847.08 |
SS =189.23 |
Mrow =23.92 |
|
n =13 |
n =13 |
SDrow =6.91 |
|
Tcol =569 |
Tcol =433 |
N =52 |
|
ncol =26 |
ncol =26 |
G =1002 |
|
Mcol =21.88 SDcol =8.8 |
Mcol =16.65 SDcol =6.81 |
ΣX2 =22760 |
|
N = 52
Replications, r = 13
ΣX = 1002
(ΣX)² =1004004
ΣX² = 22760
SSA = Σ((ΣXⱼ)²/nⱼ) - (ΣX)²/N = (380²/26 + 622²/26) - 1004004/52 = 1126.2308
SSB = Σ((ΣXᵢ)²/nᵢ) - (ΣX)²/N = (569²/26 + 433²/26) - 1004004/52 = 355.6923
SSBN = Σ((ΣX)²/n) - (ΣX)²/N = 1483.1538
SSAxB = SSBN - SSA - SSB = 1.2308
SSW = SST - SSA - SSB - SSAxB = 1969.0769
SST = ΣX² - (ΣX)²/N = 22760 - 1004004/52 = 3452.2308
dfA = a - 1 = 1
dfB = b-1 = 1
dfAxB = (a-1)*(b-1) = 1
dfW = ab(r-1) = 48
dfT = N-1 = 51
MSA = SSA/dfA = 1126.2308/1 = 1126.2308
MSB = SSB/dfB = 355.6923/1 = 355.6923
MSAxB = SSAxB/dfAxB = 1.2308/1 = 1.2308
MSW = SSW/dfW = 1969.0769/48 = 41.0224
F for Factor A = MSA/MSW = 27.4540
p-value for Factor A = F.DIST.RT(27.454, 1, 48) = 0.0000
Critical value for Factor A = F.INV.RT(0.05, 1, 48) = 4.0427
F for Factor B = MSB/MSW = 8.6707
p-value for Factor B = F.DIST.RT(8.6707, 1, 48) = 0.0050
Critical value for Factor B = F.INV.RT(0.05, 1, 48) = 4.0427
F for interaction = MSAxB/MSW = 0.0300
p-value for Interaction = F.DIST.RT(0.03, 1, 48) = 0.8632
Critical value for Interaction = F.INV.RT(0.05, 1, 48) = 4.0427
ANOVA | ||||||
Source of Variation | SS | df | MS | F | P-value | F crit |
Between Treatment | 1483.1538 | 3 | ||||
Factor A | 1126.2308 | 1 | 1126.2308 | 27.4540 | 0.0000 | 4.0427 |
Factor B | 355.6923 | 1 | 355.6923 | 8.6707 | 0.0050 | 4.0427 |
Interaction | 1.2308 | 1 | 1.2308 | 0.0300 | 0.8632 | 4.0427 |
Within | 1969.0769 | 48 | 41.0224 | |||
Total | 3452.2308 | 51 |
For Factor A:
As p = 0.000< 0.05, we reject the null hypothesis.
There was an effect of type of undead (zombie or vampire) on number of days
For Factor B:
As p = 0.005< 0.05, we reject the null hypothesis.
There was an effect of supplement on number of days.
For interaction:
As p = 0.863 > 0.05, we fail to reject the null hypothesis.
There is no interaction between the two factors
-------
η² for factor A:
η²A = SSA /(SST - SSB - SSAxB) = 1126.2308/(3452.2308 - 355.6923 - 1.2308) = 0.3639 = 36.39%
η² for factor B:
η²B = SSB /(SST - SSA - SSAxB) = 355.6923/(3452.2308 - 1126.2308 - 1.2308) = 0.153 = 15.3%
η² for Interaction:
η²AxB = SSAxB/(SST- SSA - SSB) = 1.2308/(3452.2308 - 1126.2308 - 355.6923) = 0.0006 = 0.06%