8 Lesson 6: Source-Grounded Data Stories
Ask questions, check evidence, and create honest charts with Copilot
Duration: 2 hours
You are in: Lesson 6
Before this: Lesson 5
After this: Lesson 7
Workbook page: Student Workbook -> Lesson 6 - Source-Grounded Data Report
8.1 Start Here
In Lesson 5, you answered questions from a written source. Today, you will use the same careful reading skills with a spreadsheet.
8.2 Why This Matters
Maria sees this social-media post:
“Food recalls are increasing, and packaged food is becoming more dangerous.”
The post has an attractive chart. It does not explain:
- where the data came from;
- whether the current year is complete;
- whether one row means one separate incident;
- whether categories overlap;
- whether the list contains every FDA recall; or
- whether more records mean more danger.
Maria opens the official FDA recall page. She downloads the spreadsheet and asks Microsoft Copilot Chat to help her check the claim.
Your mission: Be a data detective. Do not only make a chart. Decide whether the chart supports the claim.
How can we use real data without creating a misleading answer?
A chart can be mathematically correct but still communicate a misleading conclusion.
8.3 Core Idea
8.3.1 The C.H.A.R.T. Check
Name it and record the date.
Name the rows, columns, and counts.
Look for blanks and limits.
Choose and label the right chart.
Do not guess causes.
8.4 Today You Will
By the end of this lesson, you will:
- explain why data visualizations are useful;
- read chart titles, axes, values, units, sources, and limitations;
- identify a dataset’s source, title, date range, rows, and columns;
- ask Copilot to use only an uploaded dataset;
- separate a dataset fact from an interpretation;
- find missing or unclear information;
- choose an appropriate bar, line, pie, or scatter chart;
- explain what a chart shows and does not prove;
- repair a misleading conclusion; and
- produce a short source-grounded data report.
Theme: Honest data stories need a source, evidence, and a limitation.
AI Focus: Inspecting, filtering, summarizing, and charting an uploaded public dataset.
ESL Focus: Describing charts, explaining evidence, and correcting strong claims.
Content Objective: Students will create and evaluate source-grounded answers from structured data.
Language Objective: Students will use frames such as The chart shows..., The dataset does not include..., and A more accurate conclusion is....
8.5 Key Vocabulary
Key Vocabulary Match
Complete two short rounds. Choose the word that matches each meaning and example.
Not saved yet
8.6 Sentence Frames
Describe a chart
A data visualization helps me see __________.
The chart shows __________.
The horizontal axis shows __________.
The unit is __________.
The highest category is __________.
The number increased from __________ to __________.
The number decreased from __________ to __________.
Explain evidence
The source contains __________ records.
I used the __________ column.
The evidence for my answer is __________.
I verified the result by __________.
Explain a limitation
The dataset does not include __________.
A blank value does not mean __________.
One record may belong to more than one category.
The chart cannot prove that __________.
Correct a conclusion
The original conclusion was too strong because __________.
A more accurate conclusion is __________.
The source supports __________, but it does not support __________.
8.7 Lesson Agenda
| Time | Activity | Purpose |
|---|---|---|
| 20 min | Vocabulary, visual foundations, and chart choice | Match key words, read a chart, and choose one for a question. |
| 20 min | Meet and inspect the FDA source | Download, upload, and check the file. |
| 20 min | Four source-grounded challenges | Find evidence and missing details. |
| 20 min | Chart laboratory | Apply four chart types honestly. |
| 15 min | Chart Detective and repair | Find and correct misleading claims. |
| 20 min | Final investigation and partner report | Create a grounded data story. |
| 5 min | Exit ticket | Show the key habits. |
8.8 Part 1 — Visual Foundations and Chart Choice
8.8.1 What Is a Data Visualization?
A data visualization shows data in a visual form. Charts can help us:
- compare categories;
- see change over time;
- understand parts of one total; and
- notice a possible relationship between two numbers.
Visualizations are important because a reader can often notice a pattern faster in a chart than in a long list of numbers. However, a chart can be accurate and still support a misleading conclusion. We must read the chart and its limits.
- Question or title: What question is the chart answering?
- Labels and axes: What does each category or axis show?
- Values and units: What do the numbers measure?
- Source and date: Where did the data come from, and when?
- Limitation: What can the chart not show or prove?
The four examples below use fictional practice data. They do not describe this class, the FDA spreadsheet, or all students.
Practice source: Lesson 6 fictional data created for chart-reading practice. Date: Not applicable; these values are fixed examples, not a live dataset.
8.8.2 Bar Chart: Compare Categories
Question: Which vocabulary-practice activity received the most votes?
What to notice: Flashcards received the most votes. A bar chart makes the four independent categories easy to compare.
Limitation: These fictional votes do not prove that flashcards are the best learning activity.
| Practice activity | Votes |
|---|---|
| Flashcards | 7 |
| Practice quiz | 6 |
| Partner talk | 4 |
| Notes | 3 |
Your observation: The highest category is ______________________________.
8.8.3 Line Chart: Show Change in Order
Question: How did the number of words reviewed change over five class days?
What to notice: The count decreased from Day 2 to Day 3, then increased through Day 5. A line chart keeps the days in order.
Limitation: Five fictional days cannot show a long-term learning trend.
| Class day | Words reviewed |
|---|---|
| Day 1 | 4 |
| Day 2 | 7 |
| Day 3 | 6 |
| Day 4 | 10 |
| Day 5 | 12 |
Your observation: The number increased from __________ to __________.
8.8.4 Donut Chart: Show Exclusive Parts
Question: How did 20 fictional learners choose one final-practice activity?
What to notice: The three choices total 20 votes, and every learner chose only one activity. The slices are exclusive parts of one total.
Limitation: A donut chart would be misleading if learners could choose more than one activity.
| Final-practice choice | Votes | Percent |
|---|---|---|
| Chart practice | 9 | 45% |
| Source check | 6 | 30% |
| Partner explanation | 5 | 25% |
Your observation: The largest part of the total is ______________________.
8.8.5 Scatter Plot: Explore Two Numbers
Question: Is there an association between practice minutes and correct answers?
What to notice: The points show a generally positive association, with variation. Each point represents one fictional learner and two numeric values.
Limitation: The chart does not prove that practice time caused the scores. Other information is missing.
| Learner | Practice time (minutes) | Correct answers (out of 10) |
|---|---|---|
| A | 5 | 4 |
| B | 8 | 3 |
| C | 12 | 5 |
| D | 15 | 6 |
| E | 18 | 5 |
| F | 22 | 7 |
| G | 26 | 8 |
| H | 30 | 7 |
Your observation: The points show ____________________, but they do not prove ________________________________.
8.8.6 Choose a Chart for the Question
Use the research question to choose a chart.
- Bar chart: Compare categories.
- Line chart: Show change over time.
- Pie or donut chart: Show exclusive parts of one total.
- Scatter plot: Explore the relationship between two numbers.
Chart-Selection Challenge
0 of 4 questions checked
Not every dataset supports every chart. Do not force a scatter plot from the FDA file only because Copilot can make one.
8.9 Part 2 — Meet the Source
8.9.1 Download and Upload
- Open the FDA Recalls, Market Withdrawals, & Safety Alerts page.
- Find Download XLSX below the recall table.
- Download the Excel file.
- Save it with a clear name.
- Open Microsoft Copilot Chat and start a new conversation.
- Upload the spreadsheet.
- Tell Copilot to use only the uploaded source.
Do not upload personal, private, medical, school, or company-confidential information. This FDA spreadsheet is public.
8.9.2 Source Check
Meet the Source
Record what you see in your downloaded file.
Not saved yet
8.9.3 Copilot Master Prompt
Use only the uploaded dataset.
For every answer:
1. identify the columns used;
2. explain any filters or grouping rules;
3. distinguish facts from interpretation;
4. identify blank, missing, duplicated, or inconsistent values;
5. state the date range;
6. state what the dataset cannot answer;
7. do not use outside knowledge unless I explicitly ask for it;
8. do not invent causes for patterns;
9. do not treat correlation as causation;
10. ask me before making assumptions.
When creating a chart:
- recommend the most appropriate chart type;
- explain why it is appropriate;
- label the axes and units;
- use a neutral title;
- include a source note;
- include important limitations;
- provide the summary table used to create the chart.
8.10 Guided Practice
Part 3 — First Copilot Inspection
Use only the uploaded FDA recall spreadsheet.
Before answering questions, inspect the file and report:
1. the worksheet names;
2. the column names;
3. the number of data rows;
4. the earliest and latest recall dates;
5. the number of blank values in each column;
6. any columns that contain multiple categories in one cell;
7. any values that appear inconsistent or duplicated.
Do not make conclusions about safety yet.
Copilot may make a mistake. Check at least two results directly in the spreadsheet. Check a cell, a count, or a date.
Verify Two Results
Record the result, your spreadsheet evidence, and any correction.
Not saved yet
8.11 Independent Practice
Part 4 — Source-Grounded Questions
8.11.1 Challenge A: Find the Newest Record
Use only the uploaded FDA recall spreadsheet.
What is the newest recall record in the file?
Do not assume the first row is the newest.
Include:
- recall date;
- brand;
- product description;
- company;
- recall reason.
Teaching point: Row order is not evidence. Sort or calculate the latest date.
8.11.2 Challenge B: Brand Is Not Product
Use only the uploaded FDA recall spreadsheet.
Find every record containing the brand GreenWise.
Answer: “Are all GreenWise products recalled?”
Use the exact product descriptions and dates.
Explain why a brand match does not prove that every product from the brand is affected.
Teaching point: Similar names are not exact matches. In the classroom copy, the two GreenWise rows describe different products.
8.11.3 Challenge C: How Many Recalls Are Active?
Use only the uploaded FDA recall spreadsheet.
How many active recalls are in this spreadsheet?
Do not assume that a blank Terminated Recall field means active.
Explain:
1. how many records are marked Terminated;
2. how many records are blank;
3. why the spreadsheet may not contain enough information to calculate the active total.
Teaching point: Blank does not automatically mean “No,” “active,” or zero.
8.11.4 Challenge D: Is My Exact Package Recalled?
Use only the uploaded FDA recall spreadsheet.
I have a product with the same brand and description as one row in the spreadsheet.
Can this spreadsheet confirm that my exact package is recalled?
List information that may be missing, such as:
- lot code;
- UPC;
- package size;
- expiration or best-by date;
- distribution location.
Explain whether the complete FDA recall notice is necessary.
Teaching point: An index record may identify a notice but may not identify one exact package.
Can the File Answer It?
Your existing Lesson 6 answers remain saved here.
Not saved yet
8.12 Part 5 — Chart Laboratory
You have practiced reading four chart types with simple fictional data. Now apply the same chart-reading checks to a real public source. For every chart, name the question, labels, values, units, source, date, and limitation.
8.12.1 Chart 1: Bar Chart
Good for: Comparing or ranking categories.
Avoid: Counting a comma-separated cell as one category or silently merging different meanings.
Use only the uploaded FDA recall spreadsheet.
The Product-Types column may contain more than one comma-separated category in one cell.
1. Split the categories.
2. Normalize extra spaces.
3. Show the grouping rules.
4. Do not silently merge categories with different meanings.
5. Create a horizontal bar chart of the 10 most frequent individual labels.
6. Provide the count for each bar and the summary table.
Explain why the category totals can add up to more than the total number of records.
State what the chart shows and what it does not prove.
Misleading: “Food & Beverages causes most recalls.”
Better: “Food & Beverages is the most frequently assigned product-type label in this dataset.”
Why: Categories overlap. A label count does not measure danger or prove a cause.
Try the exercise before opening this panel. This reference uses the supplied FDA snapshot. Hover over a bar to check its exact count.
What to notice: Food & Beverages has 709 label assignments. One record can have several labels, so the label totals can be greater than 972 records.
Limitation: The chart counts labels in this file. It does not measure danger, severity, or the number of separate incidents.
| Individual product-type label | Count |
|---|---|
| Food & Beverages | 709 |
| Foodborne Illness | 237 |
| Allergens | 198 |
| Drugs | 157 |
| Medical Devices | 60 |
| Animal & Veterinary | 54 |
| Dietary Supplements | 34 |
| Contaminants | 32 |
| Pet Food | 20 |
| Produce | 19 |
Quick Graph Check
0 of 2 questions checked
8.12.2 Chart 2: Line Chart
Good for: Showing change across dates, months, or years.
Avoid: Comparing a partial current year with complete years without a clear warning.
Use only the uploaded FDA recall spreadsheet.
Count the records by month and create a line chart.
Requirements:
1. sort the months chronologically;
2. include every month in the dataset range;
3. display zero when a month has no records;
4. clearly mark incomplete years;
5. do not claim that the dataset contains every FDA recall;
6. provide the summary table.
Describe the largest increases and decreases without inventing causes.
Misleading: “Recalls decreased in 2026.”
Better: “The file contains fewer 2026 records through its latest date, but 2026 is incomplete. The chart does not establish a full-year decrease.”
Try the exercise before opening this panel. Hover over the line to check a month. Drag across the chart to zoom. Double-click the chart to reset it.
Open the complete monthly reference table
| Month | Records |
|---|---|
| October 2017 | 1 |
| November 2017 | 0 |
| December 2017 | 1 |
| January 2018 | 0 |
| February 2018 | 1 |
| March 2018 | 1 |
| April 2018 | 0 |
| May 2018 | 0 |
| June 2018 | 0 |
| July 2018 | 0 |
| August 2018 | 0 |
| September 2018 | 0 |
| October 2018 | 0 |
| November 2018 | 0 |
| December 2018 | 1 |
| January 2019 | 1 |
| February 2019 | 1 |
| March 2019 | 0 |
| April 2019 | 0 |
| May 2019 | 4 |
| June 2019 | 4 |
| July 2019 | 5 |
| August 2019 | 2 |
| September 2019 | 2 |
| October 2019 | 4 |
| November 2019 | 0 |
| December 2019 | 1 |
| January 2020 | 2 |
| February 2020 | 1 |
| March 2020 | 1 |
| April 2020 | 3 |
| May 2020 | 0 |
| June 2020 | 3 |
| July 2020 | 13 |
| August 2020 | 18 |
| September 2020 | 3 |
| October 2020 | 2 |
| November 2020 | 0 |
| December 2020 | 1 |
| January 2021 | 0 |
| February 2021 | 1 |
| March 2021 | 2 |
| April 2021 | 1 |
| May 2021 | 0 |
| June 2021 | 1 |
| July 2021 | 1 |
| August 2021 | 0 |
| September 2021 | 2 |
| October 2021 | 2 |
| November 2021 | 3 |
| December 2021 | 2 |
| January 2022 | 4 |
| February 2022 | 2 |
| March 2022 | 2 |
| April 2022 | 3 |
| May 2022 | 3 |
| June 2022 | 5 |
| July 2022 | 4 |
| August 2022 | 2 |
| September 2022 | 6 |
| October 2022 | 2 |
| November 2022 | 0 |
| December 2022 | 4 |
| January 2023 | 3 |
| February 2023 | 2 |
| March 2023 | 7 |
| April 2023 | 4 |
| May 2023 | 4 |
| June 2023 | 6 |
| July 2023 | 6 |
| August 2023 | 5 |
| September 2023 | 6 |
| October 2023 | 4 |
| November 2023 | 15 |
| December 2023 | 15 |
| January 2024 | 31 |
| February 2024 | 40 |
| March 2024 | 21 |
| April 2024 | 28 |
| May 2024 | 23 |
| June 2024 | 29 |
| July 2024 | 27 |
| August 2024 | 24 |
| September 2024 | 16 |
| October 2024 | 25 |
| November 2024 | 24 |
| December 2024 | 32 |
| January 2025 | 22 |
| February 2025 | 16 |
| March 2025 | 7 |
| April 2025 | 23 |
| May 2025 | 32 |
| June 2025 | 24 |
| July 2025 | 25 |
| August 2025 | 26 |
| September 2025 | 25 |
| October 2025 | 34 |
| November 2025 | 26 |
| December 2025 | 33 |
| January 2026 | 23 |
| February 2026 | 18 |
| March 2026 | 12 |
| April 2026 | 18 |
| May 2026 | 43 |
| June 2026 | 24 |
| July 2026 | 16 |
What to notice: May 2026 has the highest monthly count, 43. The supplied file ends on July 24, 2026, so the last month and year are incomplete.
Limitation: These are records in one downloaded file. The chart does not prove a cause, a complete annual change, or a change in danger.
| Year in supplied snapshot | Records |
|---|---|
| 2017, beginning October 19 | 2 |
| 2018 | 3 |
| 2019 | 24 |
| 2020 | 47 |
| 2021 | 15 |
| 2022 | 37 |
| 2023 | 77 |
| 2024 | 320 |
| 2025 | 293 |
| 2026, through July 24 | 154 |
Quick Graph Check
0 of 2 questions checked
8.12.3 Chart 3: Pie or Donut Chart
Good for: A few mutually exclusive parts that form one meaningful total.
Avoid: Product-type slices. Those labels overlap, so the slices are not exclusive parts of one whole.
Use only the uploaded FDA recall spreadsheet.
Create a pie or donut chart with two categories:
- Explicitly marked Terminated
- Termination status not provided in this spreadsheet
Do not label the blank category “Active.”
Display both counts and percentages.
Provide the summary table.
Explain why “status not provided” is not the same as “active.”
8.13 Why Is a Product-Types Pie Chart Misleading?
One record may belong to several product types. The slices would not represent exclusive parts of one whole.
Try the exercise before opening this panel. Hover over a slice to check both the count and percentage. Read the category names carefully.
What to notice: 284 of 972 records are explicitly marked Terminated. The other 688 cells are blank, so the status is not provided in this spreadsheet.
Limitation: A blank cell does not mean active. This chart cannot calculate the number of active recalls.
| Terminated Recall field | Records | Percent |
|---|---|---|
| Explicitly marked Terminated | 284 | 29.2% |
| Termination status not provided | 688 | 70.8% |
| Total | 972 | 100.0% |
Quick Graph Check
0 of 2 questions checked
8.13.1 Chart 4: Scatter Plot
A scatter plot needs two meaningful numeric variables for every observation. The FDA spreadsheet does not have a good pair. Do not create artificial row numbers only to make a chart.
Use the FuelEconomy.gov download page for the public source. The reference graph below uses the supplied fixed 2025 workbook snapshot and its FEguide worksheet. Ask your teacher for the classroom copy if you need an exact match. A newer official file may have different rows.
Use only the uploaded 2025 FuelEconomy.gov workbook snapshot.
Use only the FEguide worksheet.
Create a scatter plot with:
- x-axis: City FE (Guide) - Conventional Fuel;
- x-axis unit: miles per gallon;
- y-axis: Annual Fuel1 Cost - Conventional Fuel;
- y-axis unit: dollars per year;
- one point per usable 2025 FEguide row.
Identify the source and columns.
Keep only rows where Model Year is 2025 and both numeric values are present.
Verify that Fuel Unit Desc - Conventional Fuel is “miles per gallon.”
Show manufacturer, division, and carline when I hover over a point.
Explain any rows removed.
Label both axes and units.
Identify notable outliers.
Describe association only. Do not claim that one variable causes the other.
Try the exercise before opening this panel. Hover over a point to see the manufacturer, division, carline, city MPG, and estimated annual fuel cost.
What to notice: The 868 points show a generally negative association. Annual fuel cost tends to be lower as city MPG increases. Hovering identifies each row.
Limitation: The chart does not show one driver’s actual cost. Driving, location, fuel prices, and other information may differ. The graph alone does not prove a simple cause.
| Reference check | Result |
|---|---|
| Worksheet | FEguide |
| Usable 2025 rows | 868 |
| City fuel-economy range | 8–57 MPG |
| Estimated annual fuel-cost range | $900–$7,100 |
| Worksheets not mixed into this graph | EV, PHEV, FCV |
Quick Graph Check
0 of 2 questions checked
Interpret and Limit Each Chart
For each chart, say what it shows and what it cannot prove.
Not saved yet
8.14 Part 6 — Chart Detective
Classify each conclusion as Supported, Partly supported, Misleading, or Not supported.
Two Rounds of Four Claims
0 of 8 questions checked
8.15 Part 7 — Repair a Misleading Chart
Copilot creates a chart titled “FDA recalls are falling in 2026.”
The original chart did not state that 2026 ends on July 24 in the classroom copy.
Find the problems:
- unequal time periods;
- incomplete 2026 data;
- a source that does not claim complete recall coverage;
- a title that states more than the chart proves;
- no source note; and
- no date-range note.
Repair the Chart
Make every label and conclusion match the evidence.
Not saved yet
Title: Recall records in the downloaded FDA file by year
Subtitle: The 2026 value includes records only through the latest date available in the downloaded file.
Conclusion: The file currently contains fewer records dated 2026, but the year is incomplete. The chart does not establish that full-year recalls decreased.
8.16 Part 8 — Final Data Investigation
Choose one question:
- Which product-type labels appear most frequently?
- Which companies have multiple records?
- Which brands appear in more than one record?
- How have record counts changed over time?
- Which recall-reason descriptions appear most frequently?
- What percentage of records is explicitly marked Terminated?
- Which values need cleaning or normalization?
8.16.1 Return to Maria’s Claim
Maria’s opening statement had two claims:
“Food recalls are increasing, and packaged food is becoming more dangerous.”
Read each claim separately.
- “Food recalls are increasing” is only partly supported by this snapshot. The
Food & Beverageslabel appears more often in 2025 than in 2024. However, two years do not prove a general trend. The file may not contain every food recall, and 2026 is incomplete. - “Packaged food is becoming more dangerous” is not supported. The spreadsheet does not measure danger. It does not include injury or illness rates, severity, consumer exposure, sales volume, or another measure of risk.
| Year | Records with the Food & Beverages label |
Comparison note |
|---|---|---|
| 2024 | 240 | Complete year in this snapshot |
| 2025 | 243 | Complete year in this snapshot |
| 2026 | 120 | Partial year through July 24 |
Important: These are records in the supplied spreadsheet snapshot. The existing line-chart exercise counts all FDA records in the file, not only records with the Food & Beverages label.
Write your verdict in the final field of your report. Use this frame:
The spreadsheet partly supports __________ because __________.
However, it does not support __________ because __________.
One limitation is __________.
Source-Grounded Data Report
Your existing Lesson 6 source-check fields remain compatible. New fields complete the data report.
Not saved yet
Try writing your own verdict before opening this panel.
The spreadsheet alone does not support the full claim. It has 243
Food & Beveragesrecords in 2025 and 240 in 2024, but this does not prove that food recalls are generally increasing, and 2026 is incomplete. The file does not measure danger, so it cannot show that packaged food is becoming more dangerous.
8.16.2 Assessment Rubric
| Category | 2 points | 1 point | 0 points |
|---|---|---|---|
| Source | Correct source and date | One source detail missing | Source not named |
| Question | Dataset can reasonably answer it | Needs a small revision | Dataset cannot answer it |
| Chart | Appropriate and clearly labeled | Mostly appropriate | Inappropriate or unlabeled |
| Evidence | Correct columns, filters, and counts | Some evidence is unclear | Evidence is missing or incorrect |
| Interpretation | Careful conclusion and limitation | One is incomplete | Exaggerated or unsupported |
Total: 10 points
8.17 Pair Speaking or Presentation Task
Tell a partner:
My research question is __________.
I used the __________ column.
My chart shows __________.
I verified the result by __________.
One limitation is __________.
The chart cannot prove that __________.
My verdict on Maria's opening claim is __________.
The first part is __________, but the second part is __________.
8.18 Reflection
Reflection
Connect your old source-check habit to today’s data work.
Not saved yet
8.19 Exit Ticket
Exit Ticket
Show your six answers and source title to your teacher.
Not saved yet
8.20 Homework
Estimated time: 30-45 minutes.
Tasks:
Find another public CSV or XLSX dataset from an approved source such as FDA, Data.gov, NYC Open Data, New York State Open Data, the Bureau of Labor Statistics, College Scorecard, FuelEconomy.gov, or NOAA.
- Record the source and download date.
- Upload the public file to Copilot.
- Inspect its columns and limitations.
- Ask one practical question.
- Create one appropriate visualization.
- Verify one result manually.
- Write three grounded findings.
- Identify one unsupported conclusion.
- Explain what another source would be needed to confirm.
What to complete:
- Student Workbook page: Lesson 6 - Source-Grounded Data Report
- one chart or screenshot submission note with an accessible description
- three grounded findings, one limitation, and one revised conclusion
Privacy reminder: Do not upload personal or confidential data.
This lab is optional. Use it only if your teacher asks or if you want extra practice. You do not need to complete the lab to finish your portfolio.
Open the Lessons 5-6 Source-Grounded Lab in Colab:
Use the built-in safe sample sources only. Do not paste private documents or upload personal data.
The lab uses Colab AI automatically when it is available. If it is unavailable, the regular lab steps still work.
After any AI answer, say: “I checked the AI answer before I used it.”
Complete:
Student Workbook -> Lesson 6 - Source-Grounded Data Report