In: Statistics and Probability
Iconic memory is a type of memory that holds visual information for about half a second (0.5 seconds). To demonstrate this type of memory, participants were shown three rows of four letters for 50 milliseconds. They were then asked to recall as many letters as possible, with a 0-, 0.5-, or 1.0-second delay before responding. Researchers hypothesized that longer delays would result in poorer recall. The number of letters correctly recalled is given in the table.
Delay Before Recall | ||
---|---|---|
0 | 0.5 | 1 |
8 | 6 | 5 |
11 | 8 | 7 |
9 | 4 | 2 |
12 | 4 | 4 |
7 | 11 | 5 |
7 | 3 | 1 |
(a) Complete the F-table. (Round your values for MS and F to two decimal places.)
Source of Variation | SS | df | MS | F |
---|---|---|---|---|
Between groups | ||||
Within groups (error) | ||||
Total |
(b) Compute Tukey's HSD post hoc test and interpret the results.
(Assume alpha equal to 0.05. Round your answer to two decimal
places.)
The critical value is for each pairwise comparison.
Which of the comparisons had significant differences? (Select all
that apply.)
Recall following no delay was significantly different from recall following a one second delay.The null hypothesis of no difference should be retained because none of the pairwise comparisons demonstrate a significant difference.Recall following a half second delay was significantly different from recall following a one second delay.Recall following no delay was significantly different from recall following a half second delay.
Delay 0 | Delay 0.5 | Delay 1 | Total | |
Sum | 54 | 36 | 24 | 114 |
Count | 6 | 6 | 6 | 18 |
Mean, Sum/n | 9 | 6 | 4 | |
Sum of square, Ʃ(xᵢ-x̅)² | 22 | 46 | 24 |
a)
Null and Alternative Hypothesis:
Ho: µ1 = µ2 = µ3
H1: At least one mean is different.
Number of treatment, k = 3
Total sample Size, N = 18
df(between) = k-1 = 2
df(within) = N-k = 15
df(total) = N-1 = 17
SS(between) = (Sum1)²/n1 + (Sum2)²/n2 + (Sum3)²/n3 - (Grand Sum)²/ N = 76
SS(within) = SS1 + SS2 + SS3 = 92
SS(total) = SS(between) + SS(within) = 168
MS(between) = SS(between)/df(between) = 38
MS(within) = SS(within)/df(within) = 6.1333
F = MS(between)/MS(within) = 6.20
p-value = F.DIST.RT(6.1957, 2, 15) = 0.0109
Decision:
P-value < α, Reject the null hypothesis.
ANOVA | ||||
Source of Variation | SS | df | MS | F |
Between Groups | 76 | 2 | 38 | 6.20 |
Within Groups | 92 | 15 | 6.13 | |
Total | 168 | 17 |
b)
At α = 0.05, k = 3, N-K = 15, Q value = 3.67
Critical Value, CV = Q*√(MSW/n) = 3.67*√(6.1333/6) = 3.71
Comparison | Diff. = (xi - xj) | CV | Results |
x̅1-x̅2 | 3 | 3.71 | Means are not different |
x̅1-x̅3 | 5 | 3.71 | Means are different |
x̅2-x̅3 | 2 | 3.71 | Means are not different |
All that apply: