Week 4 Data Lab: Do Longer Flippers Mean Heavier Penguins?

Part 1 — The Big Picture

This week you learned to look at two variables at once: a scatterplot shows how two measurements move together, the correlation r puts one number on a straight-line pattern, and a two-way table read as conditional distributions shows whether two categories are linked. In this lab you open the real penguin field log yourself, draw flipper length against body mass, compute r and check that it ignores a change of units, turn the species-by-island counts into conditional percentages, and then decide what the data can and cannot say about cause. Last, you check a chatbot's account of the same data.

Time budget: Plan for about 90 minutes.

Part 2 — The Task

Guiding question: How strongly do flipper length and body mass move together in the penguin field log, how are species and island linked, and what can an observational study like this one say about cause?

Workshop data: palmer-station-penguins

Get the data: palmer-station-penguins landing page or the direct data file — free, no login required.

Source: Measurements of 344 penguins of three species on three islands of Antarctica's Palmer Archipelago, recorded by field researchers and republished in the palmerpenguins package.

• Rows: 344 penguins
• Columns: species, island, bill_length_mm, bill_depth_mm, flipper_length_mm, body_mass_g, sex, year
• Missing values: Numeric measurement columns and sex carry missing values for a small number of penguins

Part 3 — Materials & Procedure

Tools and materials

• Google Sheets (free) or Excel
• The palmerpenguins CSV file
• A document or notebook for your written answers

Ways of working

• interpret-real-data
• compute-and-plot
• structured-observation

Procedure

1. Open a blank sheet. In Google Sheets, click cell A1 and type =IMPORTDATA("https://raw.githubusercontent.com/allisonhorst/palmerpenguins/main/inst/extdata/penguins.csv"). In Excel, save the file from that address, then use Data, then From Text/CSV, and load it starting in cell A1. Tool: Google Sheets (free) or Excel. Check: Row 1 shows the eight column names, and the penguins fill rows 2 through 345; flipper_length_mm is column E and body_mass_g is column F.
2. Select E1:F345. In Google Sheets use Insert, then Chart, and set Chart type to Scatter chart; in Excel use Insert, then Insert Scatter (X, Y), then the first Scatter option. Check that flipper length runs across the bottom and body mass runs up the side, and give both axes titles with their units. Tool: Google Sheets (free) or Excel. Check: One dot per measured penguin, rising from lower left to upper right in a band; the two rows marked NA draw no dot.
3. In empty cells, type =COUNT(E2:E345), =MIN(E2:E345), =MAX(E2:E345), and =CORREL(E2:E345,F2:F345). Format the correlation to show four decimals. Record all four in the first scaffold table, with your four-word description of the scatterplot. Tool: Google Sheets (free) or Excel. Check: The count is 342, the shortest flipper is 172 mm, the longest is 231 mm, and the correlation shows as 0.8712.
4. Test the units property. In cell I2 type =IF(ISNUMBER(E2),E2/10,"") and fill it down to I345, so column I holds flipper length in centimeters. Then type =CORREL(I2:I345,F2:F345) in an empty cell and compare it with your first correlation. Tool: Google Sheets (free) or Excel. Check: Column I runs from 17.2 to 23.1, and the new correlation also shows as 0.8712.
5. Build the two-way table of species by island. In Google Sheets use Insert, then Pivot table, with species under Rows, island under Columns, and species under Values summarized by COUNTA; in Excel use Insert, then PivotTable, and drag the same fields. Then change Show as (Google Sheets) or Show Values As (Excel) to % of column, or % of Column Total, and copy the percentages into the second scaffold table. Tool: Google Sheets (free) or Excel. Check: Each island's column adds to 100%, and the Grand Total column shows each species' share of all 344 penguins.
6. Answer the analysis questions in your document, then paste your two tables into an approved chatbot with the prompt in Part 6 and check its answer against the scripted checks, or complete the equivalent route instead. Tool: A document or notebook for your written answers. Check: You have five written answers and a list of the chatbot's errors, or the printed analysis's errors, with corrections.

Fallback route

When to use it: Use this route if the file will not import or no spreadsheet is available.

Work from the model data below, which reports the summaries and the species-by-island counts from the real file. Complete both scaffold tables, sketch a scatterplot on paper from the description of the pattern, convert the flipper range to centimeters by dividing by 10, turn each island's counts into percentages of that island's total, and answer every analysis question and the critique step from them; submit the same items for equal credit.

This route earns the same credit as the primary route.

These summaries of the real file give you every count and value the lab needs without importing it.

Fallback model data

Quantity from the file	Value
Penguins with both flipper length and body mass	342
Shortest and longest flipper	172 mm and 231 mm
Lightest and heaviest body mass	2,700 g and 6,300 g
Pattern on the scatterplot	Points rise from lower left to upper right in a band that follows a straight line
Correlation between flipper length and body mass	0.8712
Biscoe penguins	168: 44 Adelie and 124 Gentoo
Dream penguins	124: 56 Adelie and 68 Chinstrap
Torgersen penguins	52: all Adelie
All penguins in the file	344: 152 Adelie, 68 Chinstrap, 124 Gentoo
Flipper range in centimeters

Part 4 — The Scaffold

Record what your scatterplot and your formulas show about flipper length and body mass, so you can describe the relationship with evidence.

Student scaffold table 1

Quantity	Your value or description
Number of penguins plotted (n)
Shortest and longest flipper (mm)
Direction of the pattern
Form of the pattern
Strength of the pattern
Any unusual points you notice
Correlation r, flipper in mm (four decimals)
Correlation r, flipper in cm (four decimals)

Turn the species-by-island counts into percentages of each island's total, so you can compare the islands with the whole file.

Student scaffold table 2

Species	Biscoe (% of island)	Dream (% of island)	Torgersen (% of island)	All 344 penguins (%)
Adelie
Chinstrap
Gentoo
Column total

Part 5 — Analysis Questions

1. Using your second table, show the division that gives the Gentoo share of the Biscoe penguins and the division that gives the Gentoo share of all 344 penguins. Which one is a conditional distribution and which one is a marginal distribution?
2. Describe your scatterplot in the four describing words, in order: direction, form, strength, and unusual points. Then say how the value of r backs up your description of the direction and the strength.
3. Your two correlations, one with flipper length in millimeters and one in centimeters, should match. Explain why changing the units does not change r. Then describe a pattern that r would miss even though the two variables are strongly related.
4. Compare the Gentoo, Adelie, and Chinstrap percentages across the three islands with the All 344 penguins column. Are species and island associated in this file? Use at least two numbers from your table as evidence.
5. Choose one and answer it in 3–4 sentences. (a) A friend says your correlation proves that growing longer flippers makes a penguin heavier. Explain why this observational study cannot show that, name one lurking variable that could drive both measurements, and describe the kind of study that could test cause. (b) Suppose one bird in the file had been recorded with a flipper length or a body mass far from the pattern. Explain what a single unusual point can do to r, and how you would check for one before reporting the correlation.

Part 6 — AI-Critique Moment

Choose one equal-credit critique route. Route A: ask an approved AI tool to critique your completed work, then check its response against your own evidence. Route B is the no-AI equivalent printed below.

Approved tools: Claude, Gemini, ChatGPT, Grok, Microsoft Copilot.

Prompt to paste:

I've pasted my completed Week 4 Data Lab tables above. They come from the Palmer penguins file, which has 344 rows and 8 columns: species, island, bill_length_mm, bill_depth_mm, flipper_length_mm, body_mass_g, sex, and year. It is an observational field study. Please do six things. First, describe the relationship between flipper length and body mass, and give the correlation r. Second, say whether these data show that growing longer flippers makes a penguin heavier. Third, say what share of the Biscoe penguins are Gentoo and what that tells us about species and island. Fourth, as a review question, name the skew of the body mass distribution. Fifth, as a stretch question for later in the course, say what the chance is that exactly two of three penguins picked at random from the file are Gentoo. Sixth, also as a stretch question, say how sure we can be about the correlation for all penguins in the archipelago.

Test the response for these likely failures:

• Chatbots often slide from a strong correlation to a cause, saying that longer flippers make penguins heavier or that flipper growth drives weight gain because r is close to 1.
• Chatbots often report a correlation or a share they never computed, such as a rounded r of 0.9 or 0.95, a sample size other than 342, or a Biscoe Gentoo share like 75% or 80%.
• On the review question, chatbots often name the body-mass skew by where the pile of values sits, calling it left-skewed because most penguins sit at the lighter end.
• On the stretch chance question, chatbots often multiply the chances for one ordering of picks, such as Gentoo, Gentoo, then non-Gentoo, and stop, forgetting that the non-Gentoo could be the first, second, or third pick.
• Asked how sure we can be, chatbots often invent an interval for the correlation and say there is a 95% chance the true value is inside it.

Write 3–5 sentences naming one thing the chatbot got right and at least two places it went wrong or overreached, quoting its words and correcting each with evidence from your own tables, your scatterplot, or the field log.

Equivalent route (no AI required)

No AI account is required for this lab. Instead of the chatbot, critique this flawed analysis, printed here in full: "The penguin file has 350 penguins, and flipper length and body mass have a correlation of r = 0.95, so 95% of the penguins follow the pattern. This proves that growing longer flippers makes penguins heavier. On Biscoe, 36.0% of the penguins are Gentoo, the same as in the whole file, so species and island are not linked. Body mass is left-skewed, because most of the penguins are bunched at the lighter end. The chance that exactly two of three random penguins are Gentoo is the Gentoo chance times the Gentoo chance times the non-Gentoo chance, and nothing more. We can be 95% sure that the true correlation for all penguins is exactly 0.95." Mark it up: quote each flawed sentence, name what is wrong (an invented statistic, cause read from association, a misread conditional distribution, a flipped skew direction, a dropped number of ways, or a misread confidence statement), cite the evidence from your tables, scatterplot, or the field log, and write a corrected sentence beneath it; for the last two sentences, say what the course will check later. This route assesses the same objective (Objective 3 — describe the relationship between two variables using scatterplots, the correlation coefficient, and two-way tables, and explain why association alone does not establish causation) as the chatbot critique, takes comparable effort (about 30 minutes and 200–300 words), is scored by the same scoring criteria for the same credit, and its own submission artifact is the marked-up flawed analysis, posted in place of the chatbot share link and report.

Part 7 — What to Submit

• Both completed scaffold tables
• Your scatterplot with titled axes, or a labelled sketch of it
• Written answers to all five analysis questions
• The chatbot critique report with its share link, or the equivalent route's marked-up flawed analysis

Deliverable:

A short Data Lab report on the real Palmer penguins field log: two completed tables (the scatterplot summary with n, the flipper range, the four-word description, and r in millimeters and centimeters, and the species-by-island conditional percentages), a scatterplot with titled axes or a labelled sketch of it, written answers to five analysis questions, and a critique of a chatbot's account of the same data, or, on the equivalent route, a marked-up correction of the printed flawed analysis. Submit in any accessible file type or as a link; a recorded response may be submitted as audio or video with a transcript, or as text. Choose a written, recorded, or visual form for your answers; every form is graded by the same scoring criteria.

Rubric

The rubric below shows how each submitted part is evaluated; bands are full, partial, or none.

Workshop evaluation criteria

Criterion	Full	Partial	None
Getting the data and the correlation	The file is opened or the fallback model data is used, n is 342 with the flipper range recorded, r is computed and reported to four decimals, and the centimeter test is run or worked out and shows the same r.	The data are opened and r is reported, but n, the range, or the centimeter comparison is missing or wrong.	No correlation is recorded, or the values are invented rather than computed.
Scatterplot and its description	A scatterplot or labelled sketch puts flipper length across and body mass up with titled axes, and the description names direction, form, strength, and unusual points in order, backed by the value of r.	The scatterplot is present, but the axes are swapped or untitled, or the description leaves out one or two of the four describing words.	No scatterplot or sketch, and no description of the pattern, is submitted.
Reading the two-way table	Analysis questions 1, 3, and 4 are answered accurately: each conditional percentage divides by the island's total, the Gentoo share on Biscoe is compared with the share of all penguins to name the association, and the unit result and a curve r would miss are explained.	The questions are attempted, but one or two answers divide by the wrong total, confuse marginal with conditional, or leave out a reason.	The analysis questions are missing or do not address the penguin data.
Critique of the chatbot or of the printed flawed analysis	On either route, at least two errors are quoted and corrected with evidence from the student's own tables, scatterplot, or the field log, and one correct point is acknowledged; both routes earn this band equally.	On either route, errors are named but corrected without evidence, or only one error is caught.	No critique appears, or pasted output is submitted without any checking.
Reflection on cause and unusual points	The chosen reflection option is answered in 3–4 sentences: option (a) separates association from causation, names a specific lurking variable, and describes a randomized experiment; option (b) explains that one unusual point can change r a lot and names a concrete check such as looking at the scatterplot first.	The reflection is answered but stays general, or it names no specific lurking variable or no concrete check.	No reflection is submitted.

Quality Gate

• Every count, percentage, and correlation in the tables comes from your own formulas or the fallback model data, never from memory or a chatbot.
• The scatterplot puts flipper length across and body mass up, with both axes titled and their units named.
• Each island's percentages add to 100% apart from rounding, and every conditional percentage divides by that island's total, not by 344.
• The two correlations, in millimeters and in centimeters, are both recorded and compared.
• No statement in the submission claims that one variable causes the other from this observational study.
• No statistic, count, or source is invented anywhere in the submission.
• The critique quotes the chatbot or the flawed analysis and corrects at least two errors with evidence from your tables or scatterplot.