AC420 Unit 2 Mixed Cost Behavior Analysis Example

Reviewed by Chester Goodwin, MBA Cost Accounting Purdue University Global Free custom sample in 24 to 48h

A furnace power bill that climbs with tonnage but never reaches zero is split into fixed and variable parts in this AC420 Unit 2 cost behavior analysis. Purdue Global AC 420, Cost Accounting, asks BS in Accounting students to separate a mixed cost and argue for the estimation method they choose. Twelve months of melt data from a composite cookware foundry, Bellhaven Iron Works, are run through the high-low method and two regressions. July's furnace relining distorts the first two estimates. Excluding it on operational grounds gives $17,747 a month plus $62.69 a ton, with an R-squared of 0.998, and the $7,720 relining cost moves to its own budget line.

CourseAC420 Cost Accounting
UnitUnit 2
Paper typeCost behavior analysis
LengthAbout 550 words, 5 pages double-spaced
FormatAPA 7 student paper
SchoolPurdue University Global
ProgramBS in Accounting
UpdatedOctober 2026

Free sample paper for AC420 Unit 2

1

A Furnace Bill That Never Reaches Zero: A Unit 2 Cost Behavior Analysis of Twelve Months of Melt Data

[Student Name]

Purdue University Global

AC420: Cost Accounting

Unit 2 Cost Behavior Analysis

[Instructor Name]

[Date]

Bellhaven Iron Works and every figure are composite; the regressions are computed from the table shown; the textbook and studies cited are real.

What this part is doingNaming the cost, the three methods and the outcome in the lead shows the analysis argues toward a choice.
2

Data

Month, tons melted, induction furnace power cost, furnace power-on hours:

1. January: 512 tons, $50,116, 587 hours.

2. February: 488 tons, $48,258, 527 hours.

3. March: 560 tons, $52,984, 652 hours.

4. April: 541 tons, $51,421, 597 hours.

5. May: 603 tons, $55,566, 705 hours.

6. June: 577 tons, $53,764, 618 hours.

7. July: 318 tons, $45,404, 402 hours. The furnace lining was replaced this month, followed by a cold-start dry-out that ran the furnace for hours with little metal in it.

8. August: 590 tons, $54,910, 646 hours.

9. September: 622 tons, $56,488, 719 hours.

10. October: 655 tons, $58,925, 722 hours.

11. November: 634 tons, $57,543, 728 hours.

12. December: 470 tons, $47,206, 496 hours.

July is flagged before any estimate is computed, because the operational fact, not a statistical rule, is what later justifies treating it apart.

Three Estimates

1. High-low, using October's 655 tons and July's 318: variable cost $13,521 divided by 337 tons, $40.12 a ton; fixed cost $32,645 a month.

2. Regression on all twelve months: fixed $28,708, variable $43.85 a ton, R-squared 0.894.

3. Regression on eleven months, July excluded: fixed $17,747, variable $62.69 a ton, R-squared 0.998, standard error of the slope $0.94.

High-low is weak here for a specific reason, not a general one: its low point is the one abnormal month, so July's dry-out power is read as fixed cost. The method uses only two observations, which textbooks note as its main limitation (Garrison et al., 2021). The full-year regression dilutes July's effect but does not remove it.

Excluding July is justified because the relining changed what the furnace was doing that month, not because the month spoils the fit, and its cost is kept in the budget on a line of its own rather than dropped.

What this part is doingThe highlighted sentence separates an operational reason from a convenient one, the distinction graders look for when data are excluded.
3

Measured against the eleven-month line, July's residual is $7,720, which is the relining cost: $45,404 actual against $37,684 predicted for 318 tons. Anderson et al. (2003) showed that costs do not always move symmetrically with activity, so a line fitted across months with very different conditions can mislead.

Testing the Driver

A regression of power cost on power-on hours, July excluded, gives an R-squared of 0.937, against 0.998 for tons. Tons explain more of the monthly variation, which fits the physics: most power goes into melting metal, and idle hours draw far less. Tons remain the driver.

Forecast at 600 Tons

1. High-low: $32,645 + $40.12 x 600 = $56,718.

2. All twelve months: $28,708 + $43.85 x 600 = $55,018.

3. Eleven months: $17,747 + $62.69 x 600 = $55,364.

The budget carries the eleven-month estimate, $55,364 for a 600-ton month, plus a separate relining line of about $7,700 in any year the lining is due. Datar and Rajan (2021) recommend choosing a cost function on economic plausibility, goodness of fit and the slope's significance; the eleven-month estimate passes all three, and the high-low figure, though simplest, fails the first.

What this part is doingEnding on the budget figure and the three tests shows the method choice is defended, not assumed.
4

References

Anderson, M. C., Banker, R. D., & Janakiraman, S. N. (2003). Are selling, general, and administrative costs "sticky"? Journal of Accounting Research, 41(1), 47-63. https://doi.org/10.1111/1475-679X.00095

Datar, S. M., & Rajan, M. V. (2021). Horngren's cost accounting: A managerial emphasis (17th ed.). Pearson.

Garrison, R. H., Noreen, E. W., & Brewer, P. C. (2021). Managerial accounting (17th ed.). McGraw Hill.

AC420 Unit 2 instructions, in plain terms

The second AC420 unit asks you to split a mixed cost into fixed and variable components and defend the method behind the split. Start with the data in a table, and flag any month whose operations differed before computing anything. Compute the high-low estimate and at least one regression, reporting the intercept, slope and R-squared, plus the standard error where regression output is required. Explain why the methods disagree in terms of the actual business, not just statistics. If you exclude an observation, give an operational reason and show what happens to its cost. Test whether the activity measure you chose is the best driver by trying an alternative. Forecast cost at a stated activity level under each method, then recommend one estimate for the budget using plausibility, fit and significance. Cite the textbook and research on cost behavior in APA 7.

How the AC 420 Unit 2 example is put together

The sample lists all twelve months with tons, power cost and power-on hours, and flags July's relining before any arithmetic. Three estimates follow: high-low at $40.12 a ton and $32,645 fixed, a twelve-month regression at $43.85 and $28,708 with an R-squared of 0.894, and an eleven-month regression at $62.69 and $17,747 with an R-squared of 0.998. A paragraph explains that high-low fails because its low point is the abnormal month, citing Garrison and colleagues. The highlighted sentence separates an operational reason for exclusion from a convenient one. July's $7,720 residual is identified as the relining cost, and Anderson and colleagues are cited on asymmetric costs. A driver test finds hours explain less than tons. Three forecasts at 600 tons close the analysis, with the eleven-month estimate chosen by Datar and Rajan's three tests.

AC420 Unit 2 rubric: what earns full points

AC420 cost behavior analyses are graded on the soundness of the method choice as much as the arithmetic. Each estimate must be computed correctly and its weaknesses explained in terms of the data. Excluding an observation earns credit only when the reason is operational and stated, and when the excluded cost is not simply lost. Regression results should be interpreted, not just reported: what the slope means per unit of activity, what the intercept represents, and how much variation the model explains. Testing an alternative cost driver shows the estimate was defended rather than assumed. Forecasts should be shown under each method so the reader can see the stakes of the choice. A recommendation applying plausibility, fit and significance completes the argument. Organized tables and APA 7 citations to the textbook round out this Purdue Global unit's marks.

AC420 Unit 2 help: mistakes that cost points

Many Unit 2 analyses compute high-low and stop, never asking whether the high and low months were normal. Others drop an awkward month because it lowers R-squared, with no operational reason. The excluded month's cost vanishes from the budget instead of moving to its own line. Regression output is pasted without interpretation, and the intercept is called fixed cost even when the data never approach zero activity. Only one driver is tried. Forecasts are given under the chosen method alone, hiding how much the choice matters. If your section's data cover utilities, maintenance or delivery costs instead of a furnace, our writers can run the estimates and build the argument from your own figures.

Get an AC420 Unit 2 example written to your instructions

Paste in the monthly cost and activity data from the Unit 2 problem, say which estimation methods the instructions require, and attach the rubric. Every method comes back computed on your months, any unusual period identified by its cause and the choice of estimate argued in writing. A first custom sample is free and usually ready within 24-48h. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.

Request this unit free All AC420 units Finished samples

More AC420 papers and BS in Accounting samples

AC420 Unit 2 questions, answered

Is it acceptable to drop a month from the regression?

When the month reflects a different cost structure and the reason is operational, yes, provided the analysis says so and shows the effect. Here the relining is documented, and its residual of roughly 7,700 dollars becomes a separate budget line rather than disappearing. What graders mark down is a month dropped because it spoils the fit, with no reason beyond the statistics.

Why test a second cost driver at all?

Because the unit asks for the estimate to be defended, and a driver chosen without comparison is only an assumption. Furnace hours are a plausible candidate for power cost, so the sample fits both and reports which explains more of the monthly variation. When a problem supplies only one activity measure, the sample notes the untested alternative in a sentence instead.

Does AC420 expect regression output from Excel?

Many sections supply regression output or ask for it; others accept high-low with a written critique. The sample reports intercept, slope, R-squared and the standard error of the slope when regression is required, and explains each in terms of the foundry's power bill. If your instructions call for a scatter plot as well, it is added with the relining month labeled.

Where can I find a free AC420 Unit 2 sample paper?

This page holds a complete AC420 Unit 2 cost behavior analysis of a composite foundry's furnace power cost, with twelve months of data, three estimates, a driver test, forecasts at 600 tons and APA 7 references.

Why does high-low give a poor estimate in AC420 Unit 2?

Its low point is July, the relining month, when the furnace ran for hours with little metal in it. High-low reads that dry-out power as fixed cost, so it overstates fixed cost and understates the cost per ton.