MATH42715: Introduction to Statistics for Data Sciences
Assignment 2
Aim:
• Perform statistical analysis on a real data set, and
• Produce a report presenting the statistical analysis and its conclusions.
Description:
The objective of this assignment is to encourage you to develop your statistical and R computing skills by
analysing a fresh dataset (or a set of datasets), and to prepare a statistical report. You must NOT use the
following datasets we have used in class: Auto, Boston, Smarket, Default.
A list of suggested datasets is included but you may use any publicly available datasets you find
interesting.
The task is to describe and explore one or more datasets, and submit a written report summarizing your
investigation. Your analysis should include:
1. numerical and (if appropriate) graphical summaries or plots,
2. a brief assessment of the adequacy of the model(s) you have used,
3. an investigation on how predictors and features are associated (e.g. correlation);
4. your interpretations of the data and models, highlighting key features;
5. use a resampling method to validate one of your models (preferably the model you think is
best).
You may choose one dataset dataset suitable for either classification or regression.
Suggested report outline and marking scheme
1. Introduction (15%)
- Motivation and research questions.
- Data description.
- Include exploratory data analysis (of the variables, relationships etc).
2. Data Analysis (40%)
- A brief description of your modelling process. Describe how you chose the modelling
approach, how you conducted model selection, interactions you considered, and any
variable transformations.
- Include the final model(s) and a brief discussion of the model assumptions, diagnostics,
and any model fit statistics.
3. Resampling and Validation (10%)
- Briefly describe the resampling method you used and outline your validation process.
- Include a discussion on how you have used this method to support model selection.
4. Discussion & Limitations (15%)
- This includes any relevant prediction and conclusions from your model. It should be in a
narrative style to help others who want to understand the empirical findings and
implications of the model results.
- A brief discussion about any issues concerning the reliability and validity of your data or
appropriateness of the data analysis.
5. Conclusion (10%)
- In this section, you should summarize your project and highlight any key points you
want the reader to take away from the project.
6. Overall presentation and Formatting (10%)
- 8 to 12, plus an appendix if necessary
- Font: Arial 11pt or similar
- Spacing: single or 1.5
- Margins: 2 to 2.54 cm
- Figures/Tables: 8pt minimum font size, all tables and figures should be captioned.
7. Appendix
- This includes anything that is not included in the main body of the report. This could be
additional - exploratory data analysis, plots, other models you have tried and any
additional analysis. The Appendix is NOT included in the page count, but it should be
well organized and presented.
Report Submission
Please submit your report in PDF format via Gradescope. You do not have to make your R code available
but if you wish to do so, please include it in your appendix or upload it to your University personal