Project #90671 - SPSS STATS


Marketing Research Data analysis homework


 (Due 11/02/2015)




The purpose of the assignment is to demonstrate your analysis skills. You will be provided with a dataset containing store level Nielsen data for a food item.  In class you have been exposed to the basic terminology and variables that typify scanner datasets, as well as analysis tools such as frequency tables, cross tabulations, and hypothesis testing; so this is your opportunity to put your knowledge to work.  You have the option to work on the homework alone or in pairs. SPSS or Excel can be used for your analysis.




DATA OVERVIEW                                                                                                                       


The data file contains outcome and predictor measures for four brands in a single food product category.  The brand names are disguised and labeled {A,C,D,X}.  Brand X denotes the private label product offering for the category.  The data cover weekly observations of the included variables measured at the store level.  The time period of the data spans 51 weeks, with week one corresponding to the first week of 2007.  There are 18 stores in the dataset, each with 51 weekly observations.  These stores are classified into 2 separate markets.  With 18 stores and 51 observations per store, the length of the data file is 918 records.




Week [1 : 51]      (for)     Stores [n=18]     (in)    Market Regions [#4 & #32]




In addition to the classification variables, there are seven variables of interest in the dataset.  For a single row in the dataset, there is an observation on each of the variables for each of the four brands.  For example, selling price is one key variable of interest in the dataset.  In week #1 in market #4 at store #186, the dataset has four observations for selling price, one for each brand {A,C,D,X}, represented by variables price.a, price.c, price.d, and price.x.  Thus the total number of columns in the dataset is 7x4=28 plus the above three classifiers (week, store and market), which equals 31.




Out of the seven variables of interest, there are four continuously scaled variables as follows.


Units Sold: Actual number of product units sold in that store for that week (unit sizes are equivalent across brands)


Selling Price: Recorded selling price at the register (includes discounts)


# of UPC’s: Number of UPC’s (Universal Product Coding) for a particular brand stocked in that store for that week. The higher the number of UPC's, the more variety is offered by that brand.


TPR % depth: Percentage discount (industry uses the term temporary price reduction or TPR) from normal selling price due to a price cut at that store for that week




The variables Units Sold, Selling Price, and # of UPC’s are measured on a scale from 0 to ∞.  TPR % depth is defined on the range 0-1 with zero representing normal selling price and .14, for example, representing a 14% discount on normal selling price.  Null observations for a given brand in a store for a given week mean no data recorded, and can be interpreted to mean either that the store didn’t sell any of that brand for that week, or that they simply don’t carry that brand at all.




There are three categorical variables included in the dataset that record the state of certain promotional activities for that specific brand.  If the particular activity was true for that brand for that store during that week, the variable will take the value one, if not it will take the value zero.




Feature: The brand appeared in a coupon booklet inserted into a local newspaper or other circular that week


Display: The brand is presented in some in-store sales display


Feature and Display: The brand was both featured and on in-store display






Data exploration:


Presentation of descriptive statistics, frequencies, graphics, cross tabs for relevant data.  Note that for some of the graphs/tables you will need to use different options from what were used in classroom examples. For example, to obtain the pie chart for the total unit sales of the four brands, you will need to choose "summaries of separate variables" to summarize the four unit sales variables for the four brands and then  "sum of variables" to obtain the total unit sales for each brand. If you need to look at the market share for the different market regions, then you will need to specify the market region variable in the "Panel" section.


  1. Use the unit sales variables to obtain a pie chart that represents market shares of the four brands.

  2. Obtain the market shares for the two different market regions. Do you see any differences?

  3. Frequencies of promotional activities (Feature, Display, Feature&Display) for the four brands. What do you find?

  4. Summarize the selling prices for the four brands using descriptive statistics, and compare the prices side by side using boxplots. What does the comparison reveal to you?

  5. Summarize the #UPC's and the TRP%depth discount for the four brands. What can you tell from the summary tables?

    Hypothesis testing:

    Now, pick a brand out of the four, and assume you are the manager for that brand. For all questions in this section, state the null and alternative for each hypothesis, provide a concise summary of your hypothesis testing in the form of a table, and state your conclusions.  Describe what these results mean beyond the reject, do not reject claim.

  6. Test if there is a relationship between the likelihood of your brand being on in-store display and the market region.

  7. Test if being featured makes a difference on the unit sales of your brand.

  8. Test if there is a difference in average price among the 18 stores.



  9. Use unit sales of your brand as the dependent variable, and pick the explanatory variables that you believe are important in explaining the sales of your brand. Obtain a multiple regression model. Interpret your model. What do the coefficients suggest? How much of the variation of the sales of your brand has been explained by your model? Are all the explanatory variables significant in the model?

    Summary and Conclusions:

  10. Provide a one-paragraph summary of your findings, and suggest a course of action for your selected brand based on your analysis.



    Submit your homework in one WORD document. Paste all relevant tables and graphs as pictures to the word document. You are required to submit your homework in class on 11/02. No late submission is accepted.


Subject Business
Due By (Pacific Time) 11/02/2015 06:00 am
Report DMCA

Chat Now!

out of 1971 reviews

Chat Now!

out of 766 reviews

Chat Now!

out of 1164 reviews

Chat Now!

out of 721 reviews

Chat Now!

out of 1600 reviews

Chat Now!

out of 770 reviews

Chat Now!

out of 766 reviews

Chat Now!

out of 680 reviews
All Rights Reserved. Copyright by - Copyright Policy