One question will be drawn from the following. This is the only material you need to know from the first two units except for material that has carried over into Unit 3. For instance, things like response types, population, sample, sampling distribution, etc. were covered in Unit 2. These concepts are important to understanding the Unit 3 material, so you need to know them.
Studying real organizations is sometimes the most effective way to understand some marketing research concepts. In this course, class material has been illustrated through over fifty examples of real organizations.
Most of the examples and cases have been covered in the first two exams. These possible long answer questions address examples and cases that have not been covered–there aren’t that many of that haven’t been covered!
1. In the Diageo Captain Morgan Gold case, what did management choose to do and why? (4 pts) What was the outcome, and why did it happen? (4 pts) What is the main lesson to take away from the case? (2 pts)
2. In the cloth vs. disposable diapers case, describe the background and results of the two studies. (8 pts) What lesson does this illustrate about using secondary data for marketing research? (2 pts)
3. In the Whirlpool case, what did marketing research studies show, and what did management decide to do? (6 pts) While management made a mistake in hindsight, their reasoning made sense from the production side—why? (2 pts) There are several takeaway lessons from this case. Name one. (2 pts)
Unit 3 – There is only one possible long answer question, and here it is:
Do people in New York, Chicago, Los Angeles, and Houston spend the same average amount on furniture each year, or are there differences between the cities? To answer this, a furniture company gathered data from people in the four cities. The supervisor proposes that they compare each pair of cities. So they would compare NYC vs. Chicago, NYC vs. LA, NYC vs. Houston, Chicago vs. LA, Chicago vs. Houston, and LA vs. Houston. If any of those pairs reveals a significant difference with 95% confidence (i.e., you can be 95% confident that the two groups are different), then they can conclude that the cities are not all the same.
a. Briefly, why isn’t this a good way to analyze the data? (5 pts) The problem with running 6 pair tests is that there is still a 5% chance that the z- value we calculate will be a fluke that leads to a wrong conclusion. For each calculation done, there is an increased chance of error, thus we are six times more likely to get the wrong conclusion. This gives you a total of 1-(95/100) ^6 = 0.265 = 26.5% chance of improperly rejecting at least one of your six calculations.
b. What is a better method? You only need to give the name of the method. (2 pts)
The better method to use is called analysis of variance aka ANOVA
When conducting a chi-square test, the expected frequencies are equal to
(Row total x Column total) ÷ Grand total
How is this formula derived from mathematical and probability rules? Be detailed. If it helps to explain it by referring to an actual table, you can use the table below. (10 pts) | This formula is derived by each individual amount being assigned to each other individual amount. The probability of being in row A is A/E = 150/253 = .5929 = 59.29% The probability of being in column C is C/E = 135/253 = .5336 = 53.36% Thus when mathematically combining the probability of being in row A and column C is A/E x C/E = 150/253 x 135/253 = (150×135)/253 = 80.04 which is the same as
.5929 x .5336 = .3164 x 253 = 80.04