Evaluate quality of information (accuracy, relevance, age, detail, completeness)

1. Data Processing and Information

1.1 Data and Information

Data are raw, unprocessed facts – numbers, text, images, sounds etc. Information is data that has been organised, interpreted or presented so that it is meaningful to a user.

• Sources of data
  • Direct – collected first‑hand (questionnaires, sensors, interviews, experiments).
  • Indirect – obtained from existing records or secondary sources (databases, reports, the Internet).
• Advantages of using data
  • Supports objective decision‑making.
  • Provides a basis for trend analysis and forecasting.
  • Enables automation of routine tasks.
• Disadvantages / risks
  • Data may be incomplete, inaccurate or biased.
  • Collection can be costly or time‑consuming.
  • Privacy, security and legal issues may arise.

1.2 Quality of Information – Factors (Cambridge AS/A Level IT 9626)

Cambridge states that “Factors that affect the quality of information include accuracy, relevance, age (timeliness), level of detail and completeness.” These five criteria must be applied to every data set before it is processed, analysed or reported.

Inter‑relationships

• Low relevance can mask errors in accuracy because the data may never be used where the mistake matters.
• Very detailed data (detail) that is out‑of‑date (age) may be less useful than a concise, recent summary.
• Missing elements (completeness) often lead to misleading conclusions about accuracy and relevance.

Key Quality Criteria

Criterion	What to Look For	Impact of Poor Quality
Accuracy	Correctness of facts, figures and statements; matches original source.	Mis‑interpretation, faulty calculations, loss of trust.
Relevance	Degree to which information meets the current need or problem.	Wasted time, irrelevant conclusions.
Age (Timeliness)	How recent the information is; whether it reflects the latest data.	Out‑of‑date decisions, missed opportunities.
Detail (Granularity)	Level of specificity; sufficient depth for the task.	Over‑generalisation, inability to perform detailed analysis.
Completeness	All necessary parts are present; no critical gaps.	Partial answers, need for additional sources, biased outcomes.

Assessing Each Criterion

1. Accuracy
   • Cross‑check data against original records or trusted references.
   • Look for typographical errors, rounding mistakes, or inconsistent units.
   • Validate calculations (e.g., recompute \( \displaystyle \text{Average} = \frac{\sum_{i=1}^{n} x_i}{n} \)).
2. Relevance
   • Identify the specific question or problem the information is intended to address.
   • Discard data that falls outside the scope, even if it is accurate.
   • Consider the audience’s needs – technical vs. managerial.
3. Age (Timeliness)
   • Check publication or last‑updated dates.
   • Determine whether the subject area changes rapidly (e.g., stock prices) or slowly (e.g., historical facts).
   • Use version control or timestamps where possible.
4. Detail (Granularity)
   • Assess whether the level of detail matches the task (summary vs. full report).
   • Look for missing breakdowns (e.g., totals without category breakdowns).
   • Check if additional attributes (location, time, category) are required.
5. Completeness
   • Compare the information set against a checklist of required items.
   • Identify any omitted fields, records or explanatory notes.
   • Check for logical gaps – e.g., a sales report that lacks returns data.

Exam‑style Checklist (AO2 Apply / AO3 Evaluate)

Criterion	Typical AO2 Question Prompt	Typical AO3 Question Prompt
Accuracy	“Explain how you would verify the correctness of the data.”	“Evaluate the impact of inaccurate data on the final report.”
Relevance	“Select the most appropriate data for the given problem.”	“Assess whether the chosen data set adequately addresses the problem.”
Age	“State how you would check that the data is up‑to‑date.”	“Discuss the consequences of using outdated information.”
Detail	“Describe the level of detail required for a profit‑margin calculation.”	“Critique the level of detail supplied and its effect on analysis.”
Completeness	“Identify any missing elements in the data set.”	“Evaluate how missing data could bias the conclusions.”

1.3 Encryption

Encryption protects information from unauthorised access by converting plain‑text into ciphertext.

Why encryption is needed

• Confidentiality – only intended recipients can read the data.
• Integrity – combined with hashing, it can detect unauthorised changes.
• Authentication – digital signatures confirm the sender’s identity.

Types of encryption

Type	Key Management	Typical Use	Pros	Cons
Symmetric (e.g., AES, DES)	Same secret key for encryption and decryption.	Bulk data transfer, file encryption.	Fast, low computational cost.	Key distribution problem.
Asymmetric (e.g., RSA, ECC)	Public key encrypts; private key decrypts.	Secure key exchange, digital signatures, TLS/SSL.	Solves key distribution; supports authentication.	Slower, requires larger keys.

Common protocols

• TLS/SSL – secures web traffic (HTTPS).
• IPsec – secures IP‑level communications, often used in VPNs.
• PGP – encrypts email and files.

Example (HTTPS)

1. Browser obtains the server’s public key via a digital certificate.
2. Browser generates a symmetric session key, encrypts it with the server’s public key, and sends it.
3. Both parties now share the session key and use it for fast symmetric encryption of the HTTP data.

1.4 Validation & Verification

Both processes ensure that data entered into a system is fit for purpose, but they focus on different aspects.

Validation (checking *what* is entered)

• Presence check – field is not left blank.
• Range check – value falls within permitted limits.
• Type check – correct data type (numeric, date, text).
• Format check – conforms to a pattern (e.g., postcode).
• Consistency check – related fields agree (e.g., start date ≤ end date).

Verification (checking *how* it was entered)

• Double‑entry verification – two operators enter the same data; mismatches are flagged.
• Checksum or hash verification – ensures file integrity after transfer.
• Audit trail – records who entered/changed data and when.

Difference

Validation is about *logical correctness* of the data itself; verification is about *trustworthiness* of the entry process.

Spreadsheet example

1. Apply a data‑validation rule to a column: “Whole number between 1 and 5”.
2. Use conditional formatting to highlight cells that violate the rule.
3. Enable “track changes” so the examiner can see who altered a cell and when.

1.5 Data Processing Methods

How data moves from input to output determines the design of an information system.

Batch processing

• Data are collected over a period and processed together at a scheduled time.
• Typical examples: payroll, monthly sales reports.
• Advantages: efficient use of resources, easy to schedule.
• Disadvantages: results are not available until processing is complete; not suitable for time‑critical tasks.

Online (transaction‑oriented) processing

• Data are processed immediately as each transaction occurs.
• Examples: ticket booking, online banking.
• Advantages: up‑to‑date information, immediate feedback.
• Disadvantages: requires continuous system availability; higher hardware cost.

Real‑time processing

• Data are processed within a strict time limit (often milliseconds) so that the output can influence the ongoing process.
• Examples: smart‑home climate control, industrial PLCs.
• Advantages: enables automatic control, improves safety and efficiency.
• Disadvantages: complex design, must guarantee response time.

Comparison Table

Method	Typical Use‑case	Timing of Output	Resource Implications
Batch	End‑of‑month payroll	After the batch run (hours‑days later)	Can run on low‑priority servers; scheduled downtime acceptable.
Online	E‑commerce order entry	Immediately (seconds)	Requires high‑availability hardware and network.
Real‑time	Smart‑thermostat temperature control	Within milliseconds	Deterministic processing, often embedded hardware.

Suggested Flowchart (Batch → Online → Real‑time)

1.6 Algorithms (A‑Level only)

An algorithm is a step‑by‑step procedure for solving a problem or performing a computation.

Characteristics

• Clear and unambiguous.
• Finite – must terminate after a limited number of steps.
• Input and output defined.
• Effective – each step is simple enough to be carried out.

Representations

• Flowcharts – use standard symbols (process, decision, input/output, connector).
• Pseudocode – plain‑language description that mirrors programming logic.

Example: Simple sorting (bubble sort)

FOR i ← 1 TO n‑1
    FOR j ← 1 TO n‑i
        IF A[j] > A[j+1] THEN
            SWAP A[j] AND A[j+1]
        END IF
    NEXT j
NEXT i

This algorithm repeatedly steps through a list, compares adjacent items and swaps them if they are in the wrong order. It illustrates iteration, comparison and swapping – core concepts examined in the syllabus.

Practical Example 1 – Market Survey

A team receives a CSV file containing customer‑satisfaction scores for a new smartphone.

• Accuracy: Verify each score is between 1 and 5 and matches the original questionnaire.
• Relevance: Confirm the survey targets the smartphone model under review, not a previous version.
• Age: Note the collection date – data older than three months may not reflect recent firmware updates.
• Detail: Ensure demographic fields (age, region, usage frequency) are present for segment analysis.
• Completeness: Check that every respondent has a score; missing entries indicate incomplete data.

Practical Example 2 – Smart‑Home Sensor Data

A smart‑home system logs temperature, humidity and motion readings every minute. The data will be used to optimise heating schedules.

• Accuracy: Compare a sample of sensor readings with a calibrated reference thermometer.
• Relevance: Ensure the data covers the rooms that contain heating zones; readings from the garden are irrelevant.
• Age: Use only the most recent week of data because seasonal changes affect heating needs.
• Detail: Minute‑level granularity is required; hourly averages would be too coarse for fine‑tuned control.
• Completeness: Look for gaps in the timestamp sequence – a missing minute may indicate a sensor outage.

Mini‑Exercise – Spot the Quality Issues

Below is a short data set (CSV format) that records weekly sales of three products.

Week,Product,Units Sold,Revenue (£)
1,Alpha,120,2400
1,Beta,85,— 
2,Alpha,115,2300
2,Beta,90,1800
3,Alpha,?,2600
3,Beta,95,1900

Using the five quality criteria, answer the questions that follow.

1. Identify any accuracy problems.
2. Is the data relevant for analysing product performance over the quarter? Explain.
3. Comment on the age of the data if today is week 5.
4. Does the data provide sufficient detail for a profit‑margin calculation? Why or why not?
5. List the completeness issues and suggest how they could be resolved.

Quick‑Evaluation Checklist

Criterion	Questions to Ask	Evidence Needed
Accuracy	Is the data correct and free from errors?	Source documents, verification logs.
Relevance	Does it answer the current question?	Problem statement, user requirements.
Age	When was it created or last updated?	Timestamp, version number.
Detail	Is the level of detail sufficient?	Data dictionary, field definitions.
Completeness	Are any required elements missing?	Requirement checklist, completeness report.

Summary

Evaluating information quality is a systematic, syllabus‑driven activity. By deliberately applying the five criteria—accuracy, relevance, age, detail and completeness—students can ensure that the data they process, analyse and report is fit for purpose. Mastery of these concepts supports the examination requirements (AO2 Apply, AO3 Evaluate) and prepares learners for real‑world IT practice, where poor‑quality data can compromise security, decision‑making and system reliability.