Reference no: EM132221279

Business Statistics Assignment - Analysis and Report

Report Title - The issues with using sales data, Lurking variables and sampling error

Introduction: Supermarket franchise XYZ Preliminary analysis of sales data for a new product however since there are complex issues with the data it is advisable to hire a professional statistician to analyse the data.

Possible use of sales data: Convincing suppliers to buy shelf space, Effective of offering discounts and introducing new products.

Two main issues with analysing data:-

1) Lurking variables, it can be simple to summarize a large dataset. However a lurking variable might cause you to come to the wrong conclusion googling "Simpsons paradox" will give plenty of examples to convince you that lurking variables can cause you come to an incorrect conclusion.

2) The new product was only released in selected stores (some of the stores) so there are possible issues with sampling error, an example of sampling error is that sometimes opinion polls before an election sometimes do not correctly predict who will win."

The assignment has been split into four parts -

Part 1 - Discuss a sample report from Supermarket franchise XYZ regarding the issues with sales data and include further recommendations. Explain how the statistical techniques are useful in a practical business situation.

Part 2 - Reproduce some of the sections of the sample report wing a different sample dataset provided. This will require the use of computer software such as excel.

Part 3 - Find or make a dataset that would be useful in business and give an outline of how to make a report using the dataset.

Part 4 - Discuss what are the key features of the report format, refer to the sample report in your discussion.

Part 1a -

Give an example of Simpsons Paradox. Which of the following 3 METHODS does the example of Simpsons paradox use to describe the relationship between variables.

METHOD 1 Compare means

METHOD 2 Compare proportions

METHOD 3 Use a scatterplot

Why was the particular method used?

Is understanding of Simpson's paradox useful in some Business situations, why or why not?

Part 1b -

What methods were used to describe the relationship between variables in the report and assignment? Why?

Is finding the relationship between variables useful in business why or why not" Is it useful in business.

Part 1c -

Propose some further recommendations to add to the sample report, the recommendations can either be related to the practical issues of market research or supermarkets or they can be related to statistical methodology.

Part 2a -

Discuss the dataset, in the attached example why are the rows 16 to 115 and the rows 216 to 100, 015 missing?

Part 2b,c,d,e -

Reproduction of the data summaries using your allocated sample, including details of simple calculations and comments.

Part 2f -

Calculation of confidence intervals for sales in each region using your allocated sample, note that you should use wolfram alpha to work out what you multiply the standard error by but you do not have to paste int the wolfram alpha output or the graph of the confidence intervals just change the numbers in the calculations.

Part 2g -

Just find the test statistics (in this case the test statistics is just the z score of sample proportion if p=0.5) you are not required to find the p-value".

Part 2h -

Use the p-value calculators find p-values. Also add a suitable comment to each p-value.

Part 3 -

Find or makeup a dataset that would be useful in business and give an outline of how to make a report using the dataset OR mention what variables you could add the dataset in the sample report that are related to the discussion in section Sc and discuss how you would analyse the dataset and add it to the sample report.

Part 4 -

Describe the format and structure of the sample report. NOTE: the sample report is all the blue sections of this document.

Attachment:- Assignment Data.rar

Attachment:- Assignment Files.rar