At a glance
Solo or pairs · launched in Unit 2 and due at the digest gallery seven classes later · one program, one pile of data, one page of advice · the first task where the output has to change somebody’s mind
What you are making
A program that reads a pile of data and prints a digest: a short, plain summary that a specific person can act on. Forty numbers are not useful to anybody. “Empty the bin on Friday after lunch” is.
Bring your own pile — attendance counts, practice times, bus arrivals, minutes of something, a tally somebody already keeps on paper — or use one of the sets provided. Forty values is a good size. It must be data somebody actually collected about something that actually happens.
Facts are not advice
Here is a program that computes correctly and helps nobody:
items = [3, 11, 4, 9, 12, 2, 10, 5]
total = 0
highest = items[0]
for count in items:
total = total + count
if count > highest:
highest = count
average = total / len(items)
print(f"Days recorded: {len(items)}")
print(f"Average per day: {average:.1f}")
print(f"Busiest day: {highest} items")It prints:
Days recorded: 8
Average per day: 7.0
Busiest day: 12 items
Every line is true. Not one of them tells the caretaker when to walk over with the shoebox. Your digest has to take the last step — from what the numbers are to what your person should do — and it has to be honest about how confident that step is.
What must be in it
- A named person and their decision. One sentence: [Name] has to decide [what], and right now they decide it by [how].
- Your data, with its source, and permission if somebody else collected it. If the data is about people, it arrives with no names.
- A list holding the values, built or read in one place, so that more data means more values and not more code.
- At least three findings computed by loops over that list — a total, a count, an average, a highest or lowest, a search, or a grouping. Nested repetition earns its keep if your data has rows and columns.
- A recommendation in plain language, printed by the program.
- A “what this does not say” paragraph, written by you.
How to work
- Find your person and their decision before you find your data. Data first is how you end up with a beautiful chart nobody needed.
- Get the pile into a list and print it. Check the count. Data always arrives dirtier than promised — a blank, a typo, a day that is missing entirely.
- Compute one finding. Verify it by hand on paper for five values. Do not compute a second finding until the first one is provably right.
- Write the recommendation last, and try it on your person. If they say “I knew that already”, you have a finding, not advice — go back and look for what surprised you.
- Write the “what this does not say” paragraph honestly. Who is missing from this pile? What would change your recommendation? Which number is doing more work than it should?
How this is assessed
Per How Marks Work, the working periods count and so does the trail in your Code Journal. The gallery is part of the task: you read three classmates’ digests and act on one of them, and what you write about somebody else’s data is assessed alongside your own. The paragraph about limits carries real weight — a digest that oversells itself scores below one that says clearly what it cannot support.
Success criteria
| Quality | What it looks like in your digest |
|---|---|
| A decision served | The person and the decision are named up front |
| Data handled properly | Values live in a list; the count is verified |
| Findings that hold up | At least three, each checkable by hand |
| From facts to advice | The output tells them what to do, in plain words |
| Honest limits | You name who or what is missing, without hedging |
| Readable output | A tired person reads it once and understands it |
Reflect
A Code Journal entry: which finding surprised you, and what did you almost round away? Then the harder question, the one When Code Hurts keeps asking — whose experience does your average erase, and would they recognise themselves in your recommendation?
If your data is boring
Good. Boring data about a real chore beats exciting data about nothing. A number that changes what a person does on Friday is worth more than a spectacular chart about a topic nobody in the room has any power over. And if your pile genuinely says nothing, that is a finding too — write the digest that says so, with the evidence, and tell your person they can stop collecting it.
Curriculum connection
A1.6
write programs that declare, initialize, modify, and access one-dimensional arrays.
Link to original
A2.3
write algorithms with nested structures (e.g., to count elements in an array, calculate a total, find highest or lowest value, or perform a linear search).
Link to original
B2.5
design user-friendly software interfaces (e.g., prompts, messages, screens, forms).
Link to original