This assignment builds on your previous work through Milligan, Chapter 9, and takes a deeper dive into two key data visualization techniques: clustering and distribution analysis. Clustering can be a key technique for visualizing the relationships between two or more variables, moving from single dimensions into multidimensional analysis. Distribution analysis can highlight central tendencies in a dataset (such as mean or median) and allow visualization to show outliers clearly.
In this assignment, you will apply clustering and distribution analysis between several factors in a healthcare dataset by following these steps:
· Go through this document and use Tableau to answer all the questions listed below. Where applicable, paste screenshots into the template below.
· When you are ready, complete the online quiz, which verifies your homework. Use the answers you found in this document to answer the questions.
· When you have completed the online quiz, submit the Word document.
· Remember, you can always ask your instructor for help if needed.
· If you need to adjust the size of your visualizations to match the options in the questions, use the “Format”-> “Cell size” options. For example, “Ctrl+Shift+B” on a Windows computer will make the visualization bigger, and “Ctrl+Up” will make it taller.
· Assignment Instructions: Unit 3 Self-Check Assignment 2 Milligan Chapter 9 Clusters and Distributions
· Dataset:
· Pima Indians Diabetes Database (https://www.kaggle.com/datasets/uciml/pima-indians-diabetes-database)
Struggling with where to start this assignment? Follow this guide to tackle your assignment easily!
Below is a student-friendly, step-by-step guide (tutor tone) that tells you exactly how to structure and write your submission, what you must do in Tableau, how to capture screenshots that meet the template, and which resources to cite.
Step-by-step guide to structure and write your paper (and complete the Tableau work)
1) Understand the assignment and prepare
-
Read the entire assignment document once. Highlight action verbs: “use Tableau,” “answer the questions,” “paste screenshots,” “complete the quiz,” “submit the Word doc.”
-
Download the dataset immediately: Pima Indians Diabetes Database from Kaggle. Save a local copy (CSV).
-
Open Milligan Chapter 9 and skim the sections on clustering and distribution so you can reference methods and terminology.
2) Plan your Word document structure — use this clean template
(Place your screenshots in the indicated spots. Keep it professional and uncluttered.)
-
Title Page
-
Assignment title, your name, course, instructor, date.
-
-
Introduction (≈ 1 paragraph)
-
One-sentence dataset description (Pima Indians Diabetes Database).
-
One-sentence objective: e.g., “Apply clustering and distribution analysis to assess relationships between [variables you will use] and identify patterns relevant to diabetes outcomes.”
-
-
Data & Methods (≈ 1–2 paragraphs)
-
Data source & size (# rows, key variables used).
-
Preprocessing steps: missing values handling, filters applied, variable transformations (e.g., log, normalization).
-
Tableau methods: types of visualizations used (scatter plots + k-means clustering, boxplots, histograms, density plots, jitter + cluster color).
-
-
Visualizations & Results (one subsection per figure)
For each figure:-
Figure title (short, descriptive).
-
Embedded screenshot from Tableau (use the template screenshot sizes).
-
Short caption: what the chart shows.
-
Interpretation: main finding (2–4 sentences), statistical observations (clusters, centers, outliers), and any metric reported (cluster centroids, percent in cluster, median value).
-
-
Discussion (≈ 1–2 paragraphs)
-
Synthesize findings across visuals.
-
Clinical/public-health relevance — what do clusters suggest about risk groups?
-
Limitations: sampling bias, missing data, features not included.
-
-
Conclusion (≈ 1 paragraph)
-
One concise takeaway and one recommended next step (e.g., further modeling, collecting additional variables).
-
-
References
-
At least Milligan Chapter 9 + Tableau documentation + Kaggle dataset + any other articles you used. Use APA format.
-
-
Appendix (optional)
-
Code snippets (if you used any preprocessing outside Tableau), data dictionary.
-
3) Preprocess the CSV (quick checklist)
-
Open CSV in Excel or Python/pandas. Inspect for zeros where physiologically impossible (some Pima features use 0 as missing). Decide how to treat them (replace with median, remove rows, or flag). Document your choice.
-
Consider creating a derived categorical variable (e.g., Age groups) if helpful for cluster interpretation.
-
Save a cleaned CSV for import to Tableau.
4) Import into Tableau and set up workbook
-
Start a new Tableau workbook → connect to your cleaned CSV.
-
Rename fields: make sure numeric fields are numbers, categorical as dimensions.
-
Create calculated fields if needed (e.g., BMI category, age group).
5) Build the clustering visualizations (Milligan Chapter 9 focus)
-
Scatter plot with k-means clustering (typical example):
-
Rows: choose one continuous variable (e.g., Glucose). Columns: choose another (e.g., BMI).
-
Marks: Shape = Circle, Size = maybe Age or nothing, Color = Cluster.
-
Analytics pane → Clustering → Add clusters (start with k = 2 or 3). Document chosen k and reason (elbow/interpretability).
-
Display cluster centers: use labels or tooltip enrichment.
-
-
Multi-dimensional view:
-
Use Scatter Plot Matrix (small multiples) or use Dual Axis plus color/size to encode additional variables.
-
-
How to decide k: visually inspect, use silhouette or elbow heuristics (explain qualitatively in methods if not computing exact metrics).
6) Build distribution visualizations
-
Boxplots:
-
Example: Boxplot of Glucose by Outcome (diabetes status). Show median, IQR, outliers.
-
-
Histograms / Density plots:
-
Single variable distributions (e.g., Age, Glucose).
-
Overlay density curves by group (Outcome = 0 vs 1).
-
-
Violin plots (if available via custom shapes) show full distribution shape.
-
Show central tendencies:
-
Add reference lines for mean and median (Analytics → Reference Line → Per Pane or Per Cell).
-
Annotate the numeric values in captions.
-
7) Capture screenshots that meet the template requirements
-
Resize tableau view: Format → Cell Size, or use the keyboard tips in the assignment (Ctrl+Shift+B to make bigger etc.). Aim for consistent screenshot dimensions across figures.
-
Recommended export method: Worksheet → Export → Image (choose high resolution). Or use Snipping Tool/Command+Shift+4 and crop to the viz only.
-
Save images with descriptive names: e.g.,
Fig1_BMI_vs_Glucose_k3.png.
8) Write figure captions & succinct interpretations
-
Use this short formula: What + How (method) + Key finding + One implication.
-
Example: “Figure 2. Scatter plot of BMI vs Glucose with 3-cluster k-means (Tableau). Cluster 1 (n=150) shows high glucose & high BMI — likely higher diabetes risk; Cluster 2 shows low glucose & low BMI; Cluster 3 mixed.”
-
9) Answering the quiz & submission
-
Use the exact numbers you reported in the Results to answer the quiz. Double-check values match screenshots.
-
Save the Word document as
Lastname_U3SelfCheck.docx. Attach the images inline (not as separate files unless asked). -
Submit via your LMS per instructions.
10) Writing tips & scoring signals (what graders look for)
-
Clarity: Each figure must have a caption and a one-sentence interpretation.
-
Methods transparency: Document any imputation/filters.
-
Consistency: Use consistent color schemes for groups across figures.
-
Clinically relevant interpretations: Don’t overclaim — say “suggests” or “may indicate.”
-
References: Always cite dataset and Milligan + Tableau docs.
Quick Tableau how-to checklist (commands & where to click)
-
Import CSV → Data pane.
-
Drag variables to Columns and Rows for scatter/histogram.
-
Marks card → Color to assign cluster color; Size to encode quantity.
-
Analytics tab → Drag Cluster onto view → set k (number clusters).
-
Analytics tab → Drag Reference Line → set to Mean/Median.
-
Export image: Worksheet → Export → Image (choose “Entire View”).
Example visual workflow (minimal reproducible steps)
-
Load cleaned CSV into Tableau.
-
Create scatter: Columns = BMI, Rows = Glucose.
-
Marks: Color = Outcome (initially), then replace Color with Clusters after adding cluster.
-
Analytics → Clusters → Select 3 clusters → Apply.
-
Analytics → Reference Line → add median lines for BMI & Glucose.
-
Export image.
-
Add screenshot to Word → caption and interpretation.
Short checklist to hand in (before you submit)
-
Word doc has Title page, Intro, Methods, Results with screenshots, Discussion, Conclusion, References.
-
Each screenshot is labeled (Figure 1, Figure 2, …).
-
You reported cluster counts and any central tendencies used.
-
You documented preprocessing choices.
-
References in APA format (Milligan Ch.9, Kaggle dataset link, Tableau clustering doc).
Resources & links (use these in your References and for help)
-
Kaggle — Pima Indians Diabetes Database (Dataset): https://www.kaggle.com/datasets/uciml/pima-indians-diabetes-database
-
Tableau — Clustering in Tableau (official docs / how-to): https://help.tableau.com/current/pro/desktop/en-us/clustering.htm
-
Tableau — Reference lines, boxplots, histograms guidance: https://help.tableau.com/current/pro/desktop/en-us/buildexamples_distributions.htm
-
Milligan, M. (Chapter 9) — Clusters and Distributions (use your course text/reading)
-
Quick guide to exporting images from Tableau: https://help.tableau.com/current/pro/desktop/en-us/export_images.htm
-
A short tutorial on preprocessing the Pima dataset (example walkthrough, useful): https://www.kaggle.com/code/ (search for Pima Diabetes tutorials — pick one you used and cite it)
-
Screenshot tips (Windows & Mac): https://support.microsoft.com & https://support.apple.com
Suggested APA citations (examples you can adapt)
-
Milligan, M. (Year). Chapter 9: Clusters and Distributions. In Title of Book. Publisher.
-
Tableau Software. (Year). Clustering in Tableau. Retrieved from https://help.tableau.com/current/pro/desktop/en-us/clustering.htm
-
UCI Machine Learning Repository / Kaggle. (Year). Pima Indians Diabetes Database. Retrieved from https://www.kaggle.com/datasets/uciml/pima-indians-diabetes-database
If you want, I can now:
-
Fill in a specific example (I’ll create the precise sequence in Tableau and mock screenshots descriptions / filenames) — or
-
Produce a sample completed Word document with dummy screenshots and formatted text you can edit, or
-
Give specific sentence-level examples for captions and interpretations you can copy into your submission.
The post Unit 3 Self-Check Assignment 2 — Milligan Chapter 9: Clusters and Distributions appeared first on Skilled Papers.