Using data science

DASKEREL wrote this teaching and practice using Oak's sequence as a source. View the source lesson at Oak National Academy. Oak has not endorsed this material. Check Oak's terms and separate rights for third-party works.

Learning outcomes

Compare two options using labelled evidence
Explain uncertainty and a reasonable next test

Vocabulary: criterion · dataset · trade-off · uncertainty · decision

Retrieve previous learning

Why does a graph alone not prove a cause?

Check: It may show an association without ruling out other factors or data-quality problems.

Learn the concept

A decision begins with a question and criteria. For a fictional school club room, compare two rooms by seating capacity, accessible route and available times. A table makes different units visible; a single high capacity does not automatically override accessibility or scheduling constraints.

Check where every number came from and whether observations are comparable. A seat count from last year and a current booking calendar describe different dates. Label missing information instead of inventing a value. A recommendation can be provisional when the evidence does not settle a criterion.

A fair conclusion states the option, the evidence used and its limit. For example, room B may hold more pupils but its step-free access needs confirmation. Recommend verifying that access before booking. Do not include named pupils' access needs in a public classroom dataset.

Worked model

Write three criteria for a fictional club venue and identify units and dates.
Compare room A (20 seats) and B (28 seats) while marking B's access information unknown.
State that capacity alone favours B but the overall decision remains provisional.
Request a current accessibility and timetable check before recommending a booking.

Guided practice

Room A has 20 seats and a verified step-free route; B has 28 seats but access is unknown. Is B certainly better?

Hint: Use all decision criteria.

Check the guided answer

No. B has higher stated capacity, but access and availability still need checking. A defensible recommendation is provisional and states the missing evidence.

Independent practice

Write your own response to every question before viewing the model answers. Your writing stays in this page session; this activity does not submit, store or grade it. Avoid including private information.

Teacher observed practical evidence

A teacher must apply the school's risk assessment, allergy policy and any individual healthcare plan before the practical. Observe and record the criteria during supervised work. Written answers cannot certify practical safety or skill.

Teacher checks the learner's saved decision table or labelled chart against its fictional source and dated units.
Learner explains one missing value and a limitation using the actual artefact rather than only a written claim.

Find and correct a misconception

“The option with the biggest number must win.”

Correction: A decision may depend on several criteria, data quality and unknowns, not a single maximum.

Extend and reflect

Build a decision table with a third option and a reasoned tie-break.

Guidance: Define criteria before scoring and justify any weights rather than choosing them to force an answer.

Where this lesson could lead · Computer Science GCSE → Daskerel Computing and AI

This lesson practises: Compare two options using labelled evidence. Algorithms, digital systems and safe practice support later software and technology work.

Next school subject: Computer Science GCSE. Check the subject and examination board that fit your school.
Later DASKEREL learning: Daskerel Computing and AI. Explore further study and build evidence from practical work.
Later credential exploration: AWS Certified Developer - Associate is an example in this later specialism. Check current awarding-body requirements before planning an exam.
Example career direction: Software Engineer. Explore organisations and local support.

This is an exploratory route, not a promise of admission, a qualification, certification eligibility or employment. Years 7–9 learning is an early foundation; any later credential may require substantial study, experience or local approval.

Year 9 · Data representation · original lesson pack

DASKEREL beta · internal AI review pending

Compare global datasets without hiding context

DASKEREL wrote this teaching and practice using Oak's sequence as a source. View the source lesson at Oak National Academy. Oak has not endorsed this material. Check Oak's terms and separate rights for third-party works.

Learning outcomes

Check denominator, time and geography before comparison
Distinguish absolute counts from rates

Vocabulary: rate · denominator · population · time series · comparability

Retrieve previous learning

What does a missing table value mean?

Check: It is unknown or unavailable until checked, not automatically zero.

Learn the concept

Global data often combines observations collected by different organisations, years and methods. Before comparing countries or regions, check the indicator's definition, date, source and geographic boundary. A chart without those details can make unlike quantities look comparable.

Counts and rates answer different questions. If fictional regions P and Q have 20 and 10 cases but populations of 1,000 and 100, their rates per 100 people are 2 and 10. P has more cases; Q has a higher rate. Neither figure tells you why the difference exists or what each individual experiences.

A missing national value may reflect measurement or reporting limits. Avoid treating it as zero and avoid ranking people or places from incomplete data. For a classroom analysis, use a public aggregate source and cite the original indicator rather than uploading any personal records.

Worked model

Write the fictional counts and denominators for P and Q.
Calculate P: 20/1,000 × 100 = 2 per 100; Q: 10/100 × 100 = 10 per 100.
State the separate count and rate comparisons.
List date, definition and method checks before relating the example to real datasets.

Guided practice

Country A has 50 events among 5,000 people; B has 20 among 1,000. Which has the higher rate per 1,000?

Hint: Divide by population before multiplying by 1,000.

Check the guided answer

A has 10 per 1,000 and B has 20 per 1,000. B's rate is higher despite its lower absolute count; the values are fictional.

Independent practice

Write your own response to every question before viewing the model answers. Your writing stays in this page session; this activity does not submit, store or grade it. Avoid including private information.

Teacher observed practical evidence

A teacher must apply the school's risk assessment, allergy policy and any individual healthcare plan before the practical. Observe and record the criteria during supervised work. Written answers cannot certify practical safety or skill.

Teacher checks a saved comparison chart using public aggregate or fictional data, with source, year, denominator and axes visible.
Learner explains the count-versus-rate difference and flags a missing or non-comparable value in the actual artefact.

Find and correct a misconception

“More total events always means a higher individual rate.”

Correction: Population denominators can reverse the ranking; calculate and label the rate.

Extend and reflect

Explain why an age-standardised rate may be used instead of a crude rate.

Guidance: Different age structures can affect comparisons; explain the purpose without trying to derive an unprovided standardisation.

Where this lesson could lead · Computer Science GCSE → Daskerel Computing and AI

This lesson practises: Check denominator, time and geography before comparison. Algorithms, digital systems and safe practice support later software and technology work.

Next school subject: Computer Science GCSE. Check the subject and examination board that fit your school.
Later DASKEREL learning: Daskerel Computing and AI. Explore further study and build evidence from practical work.
Later credential exploration: AWS Certified Developer - Associate is an example in this later specialism. Check current awarding-body requirements before planning an exam.
Example career direction: Software Engineer. Explore organisations and local support.

This is an exploratory route, not a promise of admission, a qualification, certification eligibility or employment. Years 7–9 learning is an early foundation; any later credential may require substantial study, experience or local approval.

Year 9 · Data representation · original lesson pack

DASKEREL beta · internal AI review pending

Complete and communicate a data investigation

DASKEREL wrote this teaching and practice using Oak's sequence as a source. View the source lesson at Oak National Academy. Oak has not endorsed this material. Check Oak's terms and separate rights for third-party works.

Learning outcomes

Plan question, data, preparation, analysis and communication
Review whether the result answers the original question

Vocabulary: investigation · cleaning · analysis · visualisation · evaluation

Retrieve previous learning

Why must a country comparison retain denominator and year?

Check: A count and rate or different years cannot be interpreted as directly comparable without those contexts.

Learn the concept

An investigative cycle starts with a bounded question, chooses appropriate data, prepares it, analyses it and communicates a conclusion. Then the conclusion is tested against the original question and may lead to a better next question. The sequence is iterative because a missing variable or unexpected outlier can require a return to collection or cleaning.

For a fictional club-booking question, collect a dated, non-personal count of room use on comparable weekdays. Check duplicate entries, inconsistent room names and missing dates before calculating a summary. Record what was corrected and why; cleaning should not silently remove inconvenient observations.

Communicate a chart with title, units, scale and source, then state a bounded result and limitation. A visualisation supports reasoning only when its axes and denominators are clear. Ask whether the data answers the question about room use rather than a different question about all school demand. A teacher should inspect the dataset, calculations and saved artefact.

Worked model

Write the room-use question and what a valid observation would be.
Create a fictional dated table, flag one duplicate and one missing date.
Document cleaning decisions, then calculate a comparable weekday count.
Make a labelled chart and conclude only for the observed period; propose a repeat sample.

Guided practice

A table has the same room booking twice on one date. Should the second row be deleted without a note?

Hint: Preserve an audit trail.

Check the guided answer

No. Verify whether it is truly a duplicate, record the evidence and correction, then calculate with the cleaned set. The raw record or a change log should remain available to the teacher.

Independent practice

Write your own response to every question before viewing the model answers. Your writing stays in this page session; this activity does not submit, store or grade it. Avoid including private information.

Teacher observed practical evidence

A teacher must apply the school's risk assessment, allergy policy and any individual healthcare plan before the practical. Observe and record the criteria during supervised work. Written answers cannot certify practical safety or skill.

Teacher checks a saved, non-personal dataset, its cleaning log and chart against the investigation question.
Learner explains one calculation and one uncertainty using the actual file, not only a written description.

Find and correct a misconception

“Data analysis starts by making the nicest chart.”

Correction: First define the question and inspect the data; choose a chart only when it fits the variable and comparison.

Extend and reflect

Propose a follow-up question after one-week room counts.

Guidance: Explain what extra period or variable would make the decision more reliable while protecting privacy.

Where this lesson could lead · Computer Science GCSE → Daskerel Computing and AI

This lesson practises: Plan question, data, preparation, analysis and communication. Algorithms, digital systems and safe practice support later software and technology work.

Next school subject: Computer Science GCSE. Check the subject and examination board that fit your school.
Later DASKEREL learning: Daskerel Computing and AI. Explore further study and build evidence from practical work.
Later credential exploration: AWS Certified Developer - Associate is an example in this later specialism. Check current awarding-body requirements before planning an exam.
Example career direction: Software Engineer. Explore organisations and local support.

This is an exploratory route, not a promise of admission, a qualification, certification eligibility or employment. Years 7–9 learning is an early foundation; any later credential may require substantial study, experience or local approval.

Year 9 · Data representation · original lesson pack

DASKEREL beta · internal AI review pending

Capture comparable, privacy-safe data for a question

DASKEREL wrote this teaching and practice using Oak's sequence as a source. View the source lesson at Oak National Academy. Oak has not endorsed this material. Check Oak's terms and separate rights for third-party works.

Learning outcomes

Define an observation and a collection plan before recording
Check sample, units, consent and missing data

Vocabulary: observation · sampling · variable · unit · privacy

Retrieve previous learning

What comes first in an investigative cycle?

Check: A bounded question that determines what observations would be useful.

Learn the concept

A useful collection plan specifies the question, what counts as one observation, the variable, its unit, when it is recorded and who may access it. For a fictional library queue question, record the number of people waiting at 12:15 on each comparable school day for two weeks. These are queue snapshots, not total daily visitors or individual waiting times.

Choose a sample that can support the claim you want to make. Observing only a rainy Monday may miss ordinary variation. A two-week lunchtime sample can describe those observed lunchtimes, but not weekends or the whole year. Write the date and measurement method alongside each count and mark a missed observation as missing, not zero.

Collect the least personal data needed. Aggregate queue counts can answer the question without names, photos or identifiable records. A teacher must approve the method and location, respect safeguarding rules and provide a simulated dataset if direct observation is unsuitable. Keep a data dictionary so another learner knows whether a value means people waiting or minutes elapsed.

Worked model

State: 'How many people are waiting for the library at 12:15 on comparable school days?'
Define one observation as a headcount at that time, with date, count in people and a note of unusual closure.
Plan ten school-day observations over two weeks, without names or images.
Mark an unobserved day 'missing' and limit the final claim to actual recorded days.

Guided practice

A queue was not observed on Thursday. Should its count be recorded as zero?

Hint: Distinguish no queue from no measurement.

Check the guided answer

No. Record Thursday as missing with a reason if known. Zero would mean the observer checked and found nobody waiting.

Independent practice

Write your own response to every question before viewing the model answers. Your writing stays in this page session; this activity does not submit, store or grade it. Avoid including private information.

Teacher observed practical evidence

A teacher must apply the school's risk assessment, allergy policy and any individual healthcare plan before the practical. Observe and record the criteria during supervised work. Written answers cannot certify practical safety or skill.

Teacher checks the collection method, consent and safeguarding controls before any school observation.
Learner shows a dated aggregate dataset or teacher-approved simulation with units, provenance and explicit missing values.

Find and correct a misconception

“More columns, especially people's names, make every dataset more useful.”

Correction: Collect only task-relevant fields; identifiable data increases risk and is unnecessary for aggregate queue counts.

Extend and reflect

How could a closure day affect a two-week average?

Guidance: Mark it as context or exclude it under a stated rule, then show both the calculation and reason; never silently replace it with zero.

Where this lesson could lead · Computer Science GCSE → Daskerel Computing and AI

This lesson practises: Define an observation and a collection plan before recording. Algorithms, digital systems and safe practice support later software and technology work.

Next school subject: Computer Science GCSE. Check the subject and examination board that fit your school.
Later DASKEREL learning: Daskerel Computing and AI. Explore further study and build evidence from practical work.
Later credential exploration: AWS Certified Developer - Associate is an example in this later specialism. Check current awarding-body requirements before planning an exam.
Example career direction: Software Engineer. Explore organisations and local support.

This is an exploratory route, not a promise of admission, a qualification, certification eligibility or employment. Years 7–9 learning is an early foundation; any later credential may require substantial study, experience or local approval.

Year 9 · Data representation · original lesson pack

DASKEREL beta · internal AI review pending

Clean a small dataset with a visible audit trail

DASKEREL wrote this teaching and practice using Oak's sequence as a source. View the source lesson at Oak National Academy. Oak has not endorsed this material. Check Oak's terms and separate rights for third-party works.

Learning outcomes

Detect missing, duplicate and invalid values
Document justified corrections without changing the raw evidence

Vocabulary: duplicate · missing value · range check · audit trail · raw data

Retrieve previous learning

How should an unobserved queue count be recorded?

Check: As missing, not as a measured zero.

Learn the concept

Cleaning makes data usable while preserving what was recorded. Keep a raw copy and a change log. For queue snapshots, a count must be a non-negative whole number; 'eight' may be converted to 8 after verification, while −2 is invalid and needs checking. A blank is unknown unless a source confirms the actual count.

A repeated date does not automatically mean a duplicate: two legitimate times may have been recorded. If both date and time, method and count are identical, compare source notes before removing one from analysis. Correct a typo only with evidence; never change a value merely because it looks unusual. An extreme count could be a genuine event worth explaining.

After cleaning, recalculate from the retained valid observations and report the denominator. Suppose counts 4, 6, 6 and one missing are recorded, but the second 6 is a verified duplicated row. The clean observed counts are 4 and 6; their mean is 5 people across two observations. Dividing by four or turning missing into zero would misstate the evidence.

Worked model

Copy the raw dated queue table and define valid counts as whole numbers ≥0.
Flag a blank, a −2 and two identical rows; compare each against collection notes.
Leave the blank missing, correct or exclude −2 only with documented evidence, and remove a confirmed duplicate from analysis while retaining the raw row.
Calculate summaries from the final valid observations and report both count and cleaning decisions.

Guided practice

Valid counts are 4 and 6; a third row is missing. What mean can be reported?

Hint: Only measured valid counts enter the denominator.

Check the guided answer

(4+6)/2 = 5 people across two measured observations. The missing row is not zero, and the report should say one observation was unavailable.

Independent practice

Write your own response to every question before viewing the model answers. Your writing stays in this page session; this activity does not submit, store or grade it. Avoid including private information.

Teacher observed practical evidence

A teacher must apply the school's risk assessment, allergy policy and any individual healthcare plan before the practical. Observe and record the criteria during supervised work. Written answers cannot certify practical safety or skill.

Teacher inspects both raw and cleaned files or a worked simulation, including the dated correction log.
Learner explains one flag, one decision and the valid-observation denominator in the saved calculation.

Find and correct a misconception

“Cleaning means deleting every unusual value until the graph looks smooth.”

Correction: Verify values against evidence, preserve the raw record and document every correction or exclusion.

Extend and reflect

How would a mistaken replacement of missing with zero change the mean of 3 and 5?

Guidance: It would give (3+5+0)/3 = 8/3, about 2.67, rather than the supported measured mean 4; explain why the zero is fabricated.

Where this lesson could lead · Computer Science GCSE → Daskerel Computing and AI

This lesson practises: Detect missing, duplicate and invalid values. Algorithms, digital systems and safe practice support later software and technology work.

Next school subject: Computer Science GCSE. Check the subject and examination board that fit your school.
Later DASKEREL learning: Daskerel Computing and AI. Explore further study and build evidence from practical work.
Later credential exploration: AWS Certified Developer - Associate is an example in this later specialism. Check current awarding-body requirements before planning an exam.
Example career direction: Software Engineer. Explore organisations and local support.

This is an exploratory route, not a promise of admission, a qualification, certification eligibility or employment. Years 7–9 learning is an early foundation; any later credential may require substantial study, experience or local approval.

Year 9 · Data representation · original lesson pack

DASKEREL beta · internal AI review pending

Recommend an action proportional to the data

DASKEREL wrote this teaching and practice using Oak's sequence as a source. View the source lesson at Oak National Academy. Oak has not endorsed this material. Check Oak's terms and separate rights for third-party works.

Learning outcomes

Compare a proposed action against bounded evidence
Name uncertainty, access implications and a fair follow-up measure

Vocabulary: recommendation · evidence · baseline · uncertainty · evaluation

Retrieve previous learning

Why keep the raw dataset and change log after cleaning?

Check: They let another person check whether corrections and exclusions were justified.

Learn the concept

A recommendation connects a question, clean evidence, a practical action and a way to check its result. Suppose a fictional library has queue snapshots of 3, 5, 7 and 9 people at 12:15 on four days. The mean is 24/4 = 6, and the maximum is 9. These four days show variation; they do not establish the normal queue in every term.

A low-cost trial might open a second supervised checkout point on two comparable lunchtimes, if staffing and accessibility permit. Define the outcome before the trial, such as the same-time queue count and an optional anonymous wait-time sample. Compare like days and record whether staffing, timetable or a closure changed; do not promise a reduction from the existing four counts alone.

Communicate a recommendation with units, timeframe, source, cleaning decisions, possible downsides and a decision rule. For example, recommend continuing a trial only if measured queues fall without making access worse or overburdening staff. State that the classroom calculation is not a causal proof. Seek pupil and staff input without collecting identities or sensitive needs.

Worked model

Calculate the four-day mean: (3+5+7+9)/4 = 6 people.
Identify that a single maximum of 9 should not be presented as every day's queue.
Propose a short, supervised second-checkout trial and decide how equivalent-day counts will be measured.
Report the result with source dates, limitations, access and staffing checks before deciding whether to continue.

Guided practice

Counts are 3, 5, 7 and 9. Does an average of 6 prove a second checkout will reduce waiting?

Hint: Separate description from causal evidence.

Check the guided answer

No. Six describes the mean of four observed queue snapshots. Trial the change and compare equivalent observations while checking other factors before attributing any difference to it.

Independent practice

Write your own response to every question before viewing the model answers. Your writing stays in this page session; this activity does not submit, store or grade it. Avoid including private information.

Teacher observed practical evidence

A teacher must apply the school's risk assessment, allergy policy and any individual healthcare plan before the practical. Observe and record the criteria during supervised work. Written answers cannot certify practical safety or skill.

Teacher checks the saved chart or decision brief against raw counts, cleaned data and stated time period.
Learner explains the calculation and one limitation orally or through an accessible presentation, with no identifiable data.

Find and correct a misconception

“A lower number after an intervention proves that the intervention worked.”

Correction: Other conditions may have changed; comparable repeated observations and a clear method support a more cautious evaluation.

Extend and reflect

Suggest a fairer way to evaluate the checkout trial across a term.

Guidance: Use comparable days and times, track staffing and closures, report sample sizes and include access feedback without personal identifiers.