In: Statistics and Probability
You are the Finance Director of a start-up that delivers online ads to end users and you are about to approve the budget for the next fiscal year. You notice that your team has projected the annual revenue goal through equal increase of monthly ad deliveries, which assumes least number of deliveries in January and the most number of deliveries in December. Before approving the budget, you want to make sure that there is indeed such a difference between the monthly number of ads delivered. You pull out the data for the current year to perform analysis of variance on the mean number of ads delivered per month. Describe: (1) which method for analysis of variance you will choose and why; (2) what statistical results you want to see in order to approve the proposed revenue plan as it is.
Answer:
1)
Here will pick one way ANOVA (Analysis of Variances) to ensure that there is in reality such a distinction between the month to month number of promotions conveyed on the grounds that One way ANOVA thinks about the methods for at least 2 autonomous gatherings.
So as to decide if there is a factual proof that the related populace implies are altogether extraordinary.
2)
Ho : Null Hypothesis :
Consider,
μι = μ2 = μ3 .... = μ l2
i.e.,
[Here there is no contrast between the month to month number of promotions conveyed]
Ha : Alternative Hypothesis :
u1 != u2 != ... != u12
[Here there is huge distinction between the month to month number of promotions conveyed]
In the event that the determined estimation of F is less, at that point basic estimation of F, the thing that matters isn't critical.
Neglect to dismiss invalid theory i.e., Ho null hypothesis.
Decision:
The information don't bolster the case that there is huge distinction between the month to month number of promotions conveyed. We won't support the proposed income plan for what it's worth.
In the event that the determined estimation of F is more prominent, at that point basic estimation of F, the thing that matters is huge.
Reject invalid speculation i.e., Ho null hypothesis.
Decision:
The information bolster the case that there is noteworthy contrast between the month to month number of promotions conveyed. We will affirm the proposed income plan for what it's worth.