In: Statistics and Probability
Begin by re-writing the problem. Minitab is required; attach or include your output. Type (use Word) your assignment; you may handwrite equations.
GPA for Randomly Selected Students in Four Business Majors |
|||
Accounting |
Finance |
Human Resources |
Marketing |
2.48 |
3.16 |
2.93 |
3.54 |
2.19 |
3.01 |
2.89 |
3.71 |
2.62 |
3.07 |
3.48 |
2.94 |
3.15 |
2.88 |
3.33 |
3.46 |
3.56 |
3.33 |
3.53 |
3.50 |
2.53 |
2.87 |
2.95 |
3.25 |
3.31 |
2.85 |
3.58 |
3.20 |
SOLUTION
Seven steps Theory:
Step 1: State the Null Hypothesis
the null would be that there will be no difference among GPAs of
different subjects (business majors). Specifically in more
statistical language the null for an ANOVA is that all the
means are the same. We state the Null hypothesis as:
H0: μ1=μ2=...=μk for k levels of an experimental treatment.
[Here, k=4]
Step 2: State the Alternative Hypothesis
HA: treatment level means not all equal
A simpler way of thinking about this is that at least one mean is
different from all others.
Step 3: Set α
α = probability of Type I Error [Here, α =
0.05]
Step 4: Collect Data
Step 5: Calculate a test statistic
For categorical treatment level means, we use an F statistic, named
after R.A. Fisher. The F value we get from the data is labeled F
(calculated). [Here, F(calculated) = 3.52]
Step 6: Construct rejection regions
As with all other test statistics, a threshold (critical) value of
F is established. This F value can be obtained from statistical
tables, and is referred to as F (critical) or F (α,df(treatment),
df(error)). [Here, F (α, 3, 24) = 3.00879]
Step 7: Based on steps 5 and 6, draw a conclusion about H0
If the F (calculated) from the data is larger than the F (α), then
you are in the Rejection region and you can reject the Null
Hypothesis with (1-α) level of confidence.
Note that modern statistical software condenses step 6 and 7 by
providing a p-value. The p-value here is the probability of getting
an F (calculated) even greater than what you observe. If by chance,
the F (calculated) = F (α), then the p-value would exactly equal to
α. With larger F (calculated) values, we move further into the
rejection region and the p-value becomes less than α. So the
decision rule is as follows:
If the p-value obtained from the ANOVA is less than α, then
Reject H0 and Accept HA.
MINITAB Output:
Initially we don't know the significance of treatments, so do not tick Tukey's option.
Once we analyze ANOVA and find that atleast one mean is different, then go for Tukey's test by clicking the checkbox.
MTB > OneWay;
SUBC> Response 'Accounting' 'Finance' 'Human Resources'
'Marketing';
SUBC> IType 0;
SUBC> Tukey 5;
SUBC> GMCI;
SUBC> TGrouping;
SUBC> GIntPlot;
SUBC> GFourpack;
SUBC> TMethod;
SUBC> TFactor;
SUBC> TANOVA;
SUBC> TSummary;
SUBC> TMeans;
SUBC> Nodefault.
One-way ANOVA: Accounting, Finance, Human Resources,
Marketing
Method
Null hypothesis All means are equal
Alternative hypothesis At least one mean is different
Significance level α = 0.05
Equal variances were assumed for the analysis.
Factor Information
Factor Levels Values
Factor 4 Accounting, Finance, Human Resources, Marketing
Analysis of Variance
Source DF Adj SS Adj MS F-Value P-Value
Factor 3 1.181 0.3937 3.52 0.030
Error 24 2.687 0.1119
Total 27 3.868
Model Summary
S R-sq R-sq(adj) R-sq(pred)
0.334584 30.54% 21.85% 5.45%
Means
Factor N Mean StDev 95% CI
Accounting 7 2.834 0.505 ( 2.573, 3.095)
Finance 7 3.0243 0.1776 (2.7633, 3.2853)
Human Resources 7 3.241 0.308 ( 2.980, 3.502)
Marketing 7 3.3714 0.2575 (3.1104, 3.6324)
Pooled StDev = 0.334584
As p-value of ANOVA table is obtained as 0.030 which is
less than alpha=0.05, we Reject the null hypothesis H0.
Yes, there is a difference in the mean GPAs by the four business
majors.
Graphs:
Should Tukey pairwise comparisons be conducted? Why or why not?
Yes, Tukey's pairwise comparisons MUST be conducted because we have Rejected the hypothesis of ANOVA carried above. It is essential to know which of the subjects (means) are different.
Tukey Pairwise Comparisons
Grouping Information Using the Tukey Method and 95% Confidence
Factor N Mean Grouping
Marketing 7 3.3714 A
Human Resources 7 3.241 A B
Finance 7 3.0243 A B
Accounting 7 2.834 B
Means that do not share a letter are significantly different.
Here, GPAs for Marketing and Accounting are significantly different.
Interpreting the results
The p-value for the multiple comparisons test is
significant only if two or more intervals do not
overlap.
The summary plot displays the multiple comparison intervals. If the
intervals for two groups do not overlap, then the standard
deviations for those groups are significantly different. The
Session window displays the Bonferroni simultaneous confidence
intervals for the standard deviations of each population. Both the
summary plot and the Session window display the p-values for the
multiple comparisons test and Levene's test.
Use Levene’s test to determine if the assumption of homogeneity of variances is valid. Give hypotheses, test statistic, p-value, decision and conclusion. Use the 0.05 level of significance.
MTB > VarTest 'Accounting' 'Finance' 'Human Resources'
'Marketing';
SUBC> Unstacked;
SUBC> Confidence 95.0;
SUBC> GInterval;
SUBC> NoDefault;
SUBC> TMethod;
SUBC> TBonferroni;
SUBC> TTest.
Test for Equal Variances: Accounting, Finance, Human Resources,
Marketing
Method
Null hypothesis All variances are equal
Alternative hypothesis At least one variance is different
Significance level α = 0.05
95% Bonferroni Confidence Intervals for Standard
Deviations
Sample N StDev CI
Accounting 7 0.505267 (0.255039, 1.55632)
Finance 7 0.177563 (0.063906, 0.76705)
Human Resources 7 0.307703 (0.201376, 0.73100)
Marketing 7 0.257451 (0.105066, 0.98082)
Individual confidence level = 98.75%
Tests
Test
Method Statistic P-Value
Multiple comparisons — 0.099
Levene 1.98 0.145
YES, the assumption of homogeneity of variances is valid as Levene's test statistic=1.98 and p-value=0.145 which is greater than alpha, so we accept H0.