BUSN2037_S1_2025 Lab 07 - Week 08 instructions

Reminder: There will be a 10-minute Wattle Quiz conducted at the end of the lab session.

Lab 07 Part 1 Predictive Analytics Estimating Cost Behavior Using Regression Analysis

Background: Advanced Environmental Recycling Technologies (AERT) sells plastic/wood composite decking through
its brand name, Choicedek. AERT sells its product through Lowe's and sells most of its products in the spring and
summer, when consumers install or remodel their decks.

Data needed: BUSN2037_S1_2025_Lab07_Choicedek_Costs.xlsx

To estimate the cost behavior for AERT, we will run a basic regression analysis. Because the product sells primarily
in the spring and summer, AERT builds up its inventory from the beginning of the year through the first part of June.

Microsoft | Excel
1. Open BUSN2037_S1_2025_Lab07_Choicedek_Costs.xlsx.
2. To prepare the data for the "Busy" variable, insert a "1" in the cells that corresponds to the busy season
and a "O" in the cells that correspond to the non busy season. To do this, in column "C," starting with cell
C2, insert a 1 for those weeks from the beginning of the year until (and including) 4/06/2025
[dd/mm/yyyy]. Insert a "0" (zero) for those weeks starting 11/06/2025 until the end of the year. Assume
we have a huge amount of data, so use an Excel formula to do this. Note that "if" statements in Excel that
use dates as the logical qualifier must use a DATEVALUE function. such as
IF(A2<DATEVALUE ("5/06/2025"), 1, 0)
3. This tells Excel to look at the date in column A and if the date is less than 5/06/2025, put a "1" into
column C; otherwise put a 0 in column C. Insert this formula in C2, =IF(A2<DATEVALUE("5/06/2025"),1,0),
and copy it to the bottom of the data in column C.

Regression analysis is a way of using mathematics and statistics to determine which of several variables has
an impact on an outcome and how big that impact is on that outcome. We can think about this like an
algebraic equation where y is the dependent variable and x is the independent variable, where y = f(x). In
other words, we will run an equation where y, or Total production costs = f(Production), where the
dependent variable is total production costs for that production week, and the independent variable is the
amount of board feet of production. Through regression analysis, we can assess the fixed cost and variable
cost components, where the fixed cost component is the intercept and the variable cost component is the
coefficient on the weekly production.

4. To perform the regression, we'll use the Data Analysis Toolpak in Excel. Please select Data > Analysis >
Data Analysis.

5. A dialog box will open. Select Regression and click OK.

1
to know what affect total production cost

6. Select the dependent variable (or Y Range) Total Production Cost, so enter $D$1:$D$53 in the Input Y
Range.
7. Select the independent variable (or X Range). We expect that production board feet (variable name:
Production) is associated with total production costs, so include Production as the independent variable,
entering $B$1:$B$53 in the Input X Range. Because we highlighted the labels in the data range, click on
"Labels". Also select "New Worksheet Ply" to get the results in a new worksheet.

8. We now select OK to get the regression results.

9. Repeat Steps 6-8 but include BOTH product board feet (variable name: Production) and the Busy variable
as independent variables under Step 8. select OK to get the regression results.

Var cost board feet of production: coefficient of the production = 0.427

Formula of calculating total cost: coefficient of intercept(fixed cost) + production *0.427 + busy * 4858
Discussion Questions:

Q1. Using the final regression estimates (including busy season variable), what is the variable cost per
board feet of production? What is the formula to calculate an estimate of total production costs
during the busy season?Cost incurred regardless of the production level.
if production =0, busy=0, coefficient still have value, so is fixed cost
The cost is mixed cost as it has variable and fixed component
Q2. What is the fixed costs? Based on the results of the analysis, which would best characterize the
cost behavior for AERT?

Q3. What are some potential reasons that the busy season of production might be more expensive?

2
 Q4. What other variables would you consider that would explain total production costs?

Lab 07 Part 2 Forecasting Future Performance (Sales and Earnings for IBM)

Background: There are many reasons to forecast future performance for a firm. It may be predicting sales to
determine needed manufacturing capacity. It may be predicting cash flows to determine if the company will need
a loan or have sufficient cash on hand without getting a loan. An investor might be forecasting earnings to
determine whether the company's stock will be a good investment. The goal of this lab exercise is to forecast
sales and earnings for IBM using Tableau's Forecasting options.

Data: BUSN2037_S1_2025_Lab07_IBM.xlsx

Tableau | Desktop
1. Open Tableau and connect to Excel File BUSN2037_S1_2025_Lab07_IBM.xlsx.
2. Click on Sheet 1. Rename this as Sales_Forecasting. To start the forecast analysis, double click IBM Sales
from Measures; then double click Date of Sales from Dimensions.

3. Tableau should have defaulted the visualization type to a line chart, but if it did not, you can change it from
the Show Me tab (see above).

Tableau interacts with dates in two ways, as either "Continuous" or "Discrete." Regardless of the way
Tableau chose to display your dates when you added them to your visualization, you can change the field
to the right display quite easily.

4. From the Columns shelf, right click the Date of Sale pill, and select Quarter Q2 2015.
5. To expand this line chart to show a Forecast, select the Analysis Tab > Forecast > Show Forecast.
6. From the Marks card, change the default mark to Circle.
7. From the Analysis Tab, select Forecast > Forecast Options.
8. Adjust the Forecast Length to Exactly 3 Years. Click OK.
9. You will notice that the Forecast is different from those generated during the lecture using Excel. This is
because Tableau has defaulted to leaving out the last quarter from its analysis. Select Analysis Tab >
Forecast > Forecast Options and adjust the default from 1 to 0 in the Ignore Last Quarter.

3
 10. Create a new sheet (Rename this as Income_Forecasting) and perform the same analysis for Income before
Extraordinary Items, repeating steps 2-9 with the new data.

Discussion Questions:
15539
Q5. What is the IBM forecast sales for the 2022 Q1? (enter your answer in millions)
-424
Q6. What is the IBM forecast income before extraordinary items for the 2022 Q2? What is the IBM
forecast lower confidence bound income before extraordinary items for the 2022 Q2?

Q7. Why would a company want to forecast its own earnings?

Q8. What other sources of data besides a company’s own sales might be useful in forecasting sales?
Macro-economic factor
countries' GDP/market sales combination
industry wide: industry prospect/regulations/competitors performance

Purpose of predicting earning

- Predict sales to determine future production capacity
- determine whether the company has sufficient funds to cover future cashflow
- Track and plan inventory level to maxmize profit
- Investor: make sure they invest in company that has good prospects and maximize future return in investment
- determining the dividend to be paid to investors