ECON1061: Forecasting and Quantitative Analysis Assessment

Final Assessment

Instructions

This assignment must be completed individually. You are required to analyse data, estimate forecasting models, conduct diagnostic tests, compare models using accuracy measures, and interpret results.

Submission

Submit two files via the Assignments folder in Canvas:

1. R File containing all clean, error-free coding.

2. Word or PDF File containing written answers in the order asked.

Marks

• Total Marks: 100

• Weight: 40% of overall course grade

Academic Integrity

• This assessment is strictly individual work.

• Submissions will be checked via Turnitin for plagiarism.

• By submitting, you acknowledge your agreement to the RMIT Assessment Declaration: Presentation Requirements

Your written report must:

• Use a professional font (recommended: Arial, size 11)

• Include page numbers

• Clearly label each answer (e.g., Question 1a)

• Present graphs and tables clearly with labels

• Be free of spelling or grammatical errors

• Be written in a new Word document (do NOT copy assessment questions)

Files Provided in Canvas

1. Tourist_data.xls

2. My_data_MEL.xls

3. Final Assessment.R (template)

The dataset includes monthly short-term visitor arrivals to Australia from selected countries (Jan 1991 – Dec 2019). You must identify the country assigned to you using My_data_MEL.xls and extract the corresponding series.

R Coding Requirements

Name your file: FamilyName_StudentID.R

To score well, ensure:

1. Code runs smoothly in one execution

2. All sections are clearly labelled (e.g., Question 1, Question 2)

3. Use meaningful comments explaining steps

Data Understanding & Benchmark Forecasting 

Question 1 

Create appropriate plots to understand your dataset.

• Label axes

• Provide 50-word commentary per plot

• Discuss trends, seasonality, outliers, structural breaks, etc.

Question 2

Evaluate whether the data requires transformation.

• Compare two transformation approaches graphically

• Select the best transformation and justify (100 words)

Question 3

Apply the two most appropriate benchmark methods and justify your selection (100 words).
Examples:

• Naive method

• Seasonal naive

• Mean method

Question 4 

Perform residual analysis for each benchmark model.

• Comment whether residuals resemble white noise (100 words)

Question 5 

Generate and plot 2-year forecasts (with intervals) from each benchmark model.

• You may use a shorter window (e.g., last 5 years) for clarity

• Compare models and discuss merits/limitations (100 words)

ARIMA Modelling 

Question 6 

Inspect the transformed data visually to determine required differencing.

• Use relevant plots (ACF, PACF, differenced series)

• 50-word commentary per plot, justifying decisions

Question 7

Estimate an ARIMA model using and present the results in a table.

Question 8 

Conduct residual diagnostics for the ARIMA model.

• Use ACF plots, Ljung-Box test, etc.

• Provide a 100-word discussion

Question 9 

Plot 2-year ARIMA forecasts with intervals and provide a brief comment (50 words)

Model Evaluation 

Question 10 

Create a training set by holding out the final two years as a test set.
Generate forecasts from:

• Benchmark Model 1

• Benchmark Model 2

• ARIMA Model

Plot forecasts vs. actual data (use last 5 years if clearer).
Discuss performance (100 words).

Question 11 

Compute forecast accuracy for all three models in a table.

• Identify which model performs best and explain why (50 words)

Summary of Assessment Requirements

This final assessment for ECON1061 requires students to independently analyse a time-series dataset of monthly short-term visitor arrivals to Australia. The assessment is structured into three key parts:

1. Understanding and transforming the data, then applying simple benchmark forecasting models.

2. Building and evaluating an ARIMA model, including required diagnostics.

3. Comparing all models using forecast accuracy on a training-test split.

Students must submit two files:

• A clean, executable R script that includes labelled sections and comments.

• A Word or PDF report answering all questions in order, presenting plots, tables, commentary, and interpretations.

The assessment tests the student’s ability to:

• Create and interpret visualisations

• Assess the need for transformations

• Apply benchmark models

• Conduct residual diagnostics

• Fit ARIMA models

• Generate short-term forecasts

• Compare forecasting models using accuracy measures

• Produce a professional analytical report compliant with academic and formatting standards

• Demonstrate academic integrity by completing all work independently

The dataset is selected using the student’s unique Country ID, and all work must be original.

How the Academic Mentor Guided the Student Through the Assessment

The Academic Mentor supported the student by breaking the assessment into clear, manageable tasks and guiding them step by step through each requirement. The mentoring process focused on clarity, interpretation, and correct application of forecasting techniques while ensuring the student maintained ownership of the work.

Step 1: Understanding Assessment Scope and Files

The mentor first ensured the student understood the purpose of the task, the structure of the dataset, and the required file submissions. The mentor clarified how to identify the correct Country ID and extract the relevant series for analysis.

Step 2: Plotting and Interpreting the Data (Part A – Q1)

The mentor instructed the student on creating multiple exploratory plots, such as time-series plots, seasonal plots, and subseries plots.
They explained what to look for, including patterns, changes in level, variations in seasonality, and possible outliers. The student was guided on how to write concise 50-word commentary for each plot.

Step 3: Transformation Decisions (Part A – Q2)

The mentor helped the student compare the Box-Cox transformation and logarithmic transformation visually. The student was shown how to interpret variance stability and select the most appropriate transformation based on the plot outcomes.

Step 4: Selecting Benchmark Methods (Part A – Q3)

The mentor guided the student in choosing two suitable benchmark models, such as Naive and Seasonal Naive, explaining the logic behind each method. They discussed how these models serve as essential comparison points for more complex forecasting techniques.

Step 5: Residual Analysis of Benchmark Models (Part A – Q4)

The student was shown how to interpret residual plots, ACF of residuals, and histogram distributions. The mentor explained the criteria for white noise and helped the student summarise their findings in a structured manner.

Step 6: Forecasting Using Benchmark Models (Part A – Q5)

The mentor guided the student in generating two-year forecasts and plotting these alongside actual data. The student was encouraged to use a shorter recent window for clearer interpretation and compare forecast behaviour across models.

Step 7: Determining Differencing and Stationarity (Part B – Q6)

The mentor demonstrated how to visually inspect the transformed series, difference the data correctly, and evaluate ACF/PACF plots. Each differencing step was explained in terms of trend removal and stationarity testing.

Step 8: Building and Presenting an ARIMA Model (Part B – Q7)

The student learned how to apply the auto-ARIMA function and present ARIMA results in an appropriate table format, including coefficients, standard errors, and model identification.

Step 9: Diagnostic Checks for ARIMA (Part B – Q8)

Through mentor guidance, the student evaluated residuals using the Ljung-Box test, ACF plots, and diagnostic panels. They were taught to articulate model adequacy in a clear 100-word summary.

Step 10: Forecasting with ARIMA (Part B – Q9)

The mentor supported the student in producing two-year ARIMA forecasts and helped them interpret forecast patterns, intervals, and expected behaviour.

Step 11: Training-Test Split and Model Comparison (Part C – Q10)

The mentor showed the student how to create a training set by withholding the last two years of data.
They generated forecasts from all three models and overlaid these on the test data. The student was guided on how to write a concise 100-word performance comparison.

Step 12: Accuracy Measures and Final Evaluation (Part C – Q11)

The student computed accuracy metrics such as MAE, RMSE, and MAPE.
The mentor helped interpret these results and identify the best-performing model with justification.

Final Outcome and Learning Achieved

By completing the assessment with mentor guidance, the student achieved the following learning outcomes:

• Developed strong competence in time-series visualisation and interpretation

• Understood the importance and process of data transformation

• Applied benchmark forecasting methods confidently

• Learned how to diagnose model performance using residual analysis

• Gained hands-on experience in ARIMA modelling

• Produced multi-step forecasts with uncertainty intervals

• Compared forecasting models using both visual and numerical evidence

• Strengthened skills in R coding, analytical writing, and structured reporting

• Demonstrated academic integrity by independently completing all required tasks

The mentor’s step-by-step approach ensured the student could clearly understand the rationale behind each stage, execute all analysis correctly, and produce a coherent, academically sound final submission.

Get High-Quality Academic Guidance the Right Way

If this solution helped you understand how to structure and present your assessment, you can download the full sample to explore the formatting, analysis style, and writing approach in more detail. The sample is provided strictly for learning and reference purposes only. Submitting it as your own work may result in plagiarism penalties or academic misconduct under university policies.

For students who want personalised, reliable academic support, you can also order a fresh, custom-written assignment created specifically for your topic, marking rubric, and deadline. Every solution is prepared by qualified academic writers, checked for originality, and fully tailored to your requirements.

Why Choose a Fresh Assignment?

• Completely original and plagiarism-free work

• Written by subject experts with academic experience

• Structured and referenced according to your university guidelines

• Delivered on time with revision support

• Helps you improve understanding and boost grades

Use the sample responsibly for learning, and choose a fresh custom solution when you need confident, professional academic support.

[Download Sample Solution]     [Order Fresh Assignment]