Data analysis project guideline new.pdf
Data analysis project guideline
ECO 595 Applied Business Research
1. You need to review Chapter 23, 24, and time series to be able to finish this assignment.
2. Each group will be assigned a company by the professor. Revenue data and macro data can be found
in Data Analysis ta
folder on D2L.
3. Time: the presentation should be no longer than 10 minutes. One group member can record the
voice-over on behalf of the whole group or all members can share the recording.
4. What to submit: the Excel file that includes all your analysis and data, and the voice-over ppt file
that includes your presentation. The file names should be ‘Last name 1_Last name 2_Last name
3_project name’ for group work and ‘Last name_First name_project name’ for individual work. Only
one copy is needed from each group.
5. Part I Time Series analysis: discuss the revenue data similar to what I have showed in the Sears
example is expected: scatter plot of revenue; linear trend; moving average; structural
eak. Your
company revenue may not exhibit strong seasonality. For practice purpose, you are required to fit a
moving average to your revenue data nevertheless.
First explain what the graph tells you as a manager about the trend, seasonality, and structural
eak in this company data. Then conduct research to find out how management actually made the
decisions in the data.
Note: you do not need to provide the graphs that cannot be obtained using Excel (eg. the four
figures shown in the second to last slide in the Time Series Analysis ppt file)
6. Part II Multiple linear regression using Macro data:
a. Select at least three variables from the macro data file that you believe can be highly
co
elated with the revenue data.
. Create the appropriate seasonal dummy variables.
c. Run the regression using the three macro variables and the seasonal dummy variables as the
independent variables and the revenue data as the dependent variable. Report the
egression results.
d. Note that the dates for the revenue data and macro variables do not necessarily match each
other. You will need to determine the date range that has available data for all variables and
drop the extra data. For example, if the date for GDP ranges from 2000 to 2015 and the
evenue data ranges from 1995 to 2016. Then you would only keep data from 2000 to 2015.
e. Comment on the validity of the model
f. Comment on the significance of individual coefficients. Briefly discuss the relationship
(positive or negative) between the dependent variable and each of the independent
variables with significant coefficient. Are those relationship expected?
g. Report the adjusted r-squared and
iefly comment on the fitness of the model based on
your opinion.
For e and g, since you cannot actually point to the numbers during a voice-over presentation,
you must highlight the relevant excel output numbers on the slide about during your
discussion. You can also use an animated a
ow to point to the numbers if you want to.
7. For presentation:
a. Do not read the ppt slides and do not put everything you want to say on the ppt slides.
. Make sure you to record over every ppt slide. That is, do not give your whole presentation over
the first slide. You need to turn the pages as you go over them one by one.
c. Given an overview of the presentation before formally presenting the detailed contents.
d. Rehearse before recording the presentation.
e. You should include a cover page/slide with you name and the title of the presentation.
f. The font should be large enough for audience in a distance to see (size>18 as a minimum)
g. Unless necessary, restrain from using pictures and cartoons.