Statistics Basics: Data, Types, Collection & Uses

TAKEAWAYS:

• Statistics is the science of collecting, analyzing, and interpreting data to make informed business decisions.
• Data can be classified as primary (original) or secondary (existing), and qualitative (descriptive) or quantitative (numerical).
• Variables are categorized as nominal, ordinal, interval, or ratio, each requiring different statistical treatments.
• Sampling techniques (random, stratified, systematic) ensure representative data for accurate analysis.
• Ethical data collection prioritizes privacy, accuracy, and transparency in business research.

1. Introduction to Statistics

Statistics is the backbone of decision-making in business. It helps in collecting, organizing, analyzing, and interpreting data to derive meaningful insights. Business statistics is particularly useful in:

• Market research (e.g., customer preferences, demand forecasting).
• Financial analysis (e.g., risk assessment, investment decisions).
• Operational efficiency (e.g., supply chain optimization, quality control).

Key Definitions

• Population: The entire group being studied (e.g., all customers of a bank).
• Sample: A subset of the population used for analysis (e.g., 500 customers surveyed out of 10,000).
• Parameter: A numerical value describing a population (e.g., mean height of all Nepali adults).
• Statistic: A numerical value describing a sample (e.g., mean height of 1,000 surveyed Nepali adults).

Why Study Statistics?

• Reduces uncertainty in decision-making.
• Identifies trends (e.g., sales growth, customer behavior).
• Optimizes resources (e.g., inventory management, advertising spend).

2. Types of Data

Data can be classified based on source and nature:

A. Based on Source

B. Based on Nature

3. Types of Variables

Variables are the characteristics being measured. They are classified as:

A. Classification of Variables

Visual: Measurement Scales

4. Data Collection Methods

Data collection is the first step in statistical analysis. Methods include:

A. Primary Data Collection

1. Surveys/Questionnaires
   • Directly ask respondents for information.
   • Example: E-Sewa surveys customers on service satisfaction.
2. Experiments
   • Manipulate variables to observe effects.
   • Example: Testing two ad campaigns to see which drives more sales.
3. Observations
   • Record behavior without intervention.
   • Example: Counting foot traffic in a Daraz store.

B. Secondary Data Collection

1. Published Sources
   • Government reports, company annual reports.
   • Example: NEPSE data for stock market analysis.
2. Internal Records
   • Company databases, sales reports.
   • Example: Ncell’s customer call logs for network performance.

Advantages and Disadvantages

5. Sampling Techniques

Sampling ensures that data is representative of the population. Common techniques:

A. Probability Sampling

1. Simple Random Sampling
   • Every member has an equal chance of selection.
   • Example: Randomly selecting 100 customers from a list of 10,000.
2. Stratified Sampling
   • Population divided into subgroups (strata), then sampled.
   • Example: Surveying students from different faculties (Management, Engineering) proportionally.
3. Systematic Sampling
   • Select every k-th member from a list.
   • Example: Surveying every 10th customer entering a Pathao store.

B. Non-Probability Sampling

1. Convenience Sampling
   • Select easily accessible members.
   • Example: Surveying students in a TU classroom.
2. Purposive Sampling
   • Select members based on specific criteria.
   • Example: Interviewing top 10 Daraz sellers for case studies.

Visual: Sampling Techniques

graph TD
    A["Sampling Techniques"] --> B["Probability Sampling"]
    A --> C["Non-Probability Sampling"]
    B --> B1["Simple Random"]
    B --> B2["Stratified"]
    B --> B3["Systematic"]
    C --> C1["Convenience"]
    C --> C2["Purposive"]
    B1 --> B1a["Example: Lottery"]
    B2 --> B2a["Example: Divide by Gender"]
    B3 --> B3a["Example: Every 10th Person"]
    C1 --> C1a["Example: Convenient Students"]
    C2 --> C2a["Example: Expert Judgment"]

6. Ethical Considerations in Data Collection

Ethics ensures data is collected fairly, accurately, and respectfully:

• Informed Consent: Participants must know how their data will be used.
• Confidentiality: Personal data should be anonymized.
• Accuracy: Avoid bias in questions or sampling.
• Transparency: Clearly state the purpose of data collection.

Example: E-Sewa’s Ethical Practices

• Uses opt-in consent for customer surveys.
• Anonymizes transaction data for security.
• Avoids leading questions in feedback forms.

7. Worked Examples

Example 1: Classifying Variables

Data: Customer ratings for a Khalti app (1-5 stars).

• Variable Type: Ordinal (ordered categories with no equal intervals).
• Why? Ratings are ordered (1 < 2 < 3), but the difference between 1 and 2 isn’t numerically equal to 2 and 3.

Example 2: Sampling in Real Life

Scenario: NTC wants to survey customer satisfaction for its broadband service.

• Population: All 5 million broadband users.
• Sample: 500 users selected via stratified sampling (dividing by urban/rural areas).
• Why Stratified? Ensures representation from all regions.

Example 3: Primary vs. Secondary Data

8. In the Real World

1. eSewa’s Data Collection

   • Uses primary data (customer transactions) to analyze spending patterns.
   • Applies stratified sampling to segment users by age, location, and transaction frequency.
   • Ethical Note: Ensures GDPR-compliant data handling for user privacy.

2. Pathao’s Ride Demand Prediction

   • Collects secondary data (historical ride requests, weather data) to predict peak hours.
   • Uses ratio variables (distance, fare) to optimize pricing algorithms.
   • Real-World Impact: Reduces wait times by 30% during rush hours.

3. NEPSE’s Index Calculation

   • Relies on secondary data (company stock prices, trading volumes).
   • Uses interval variables (stock prices) to compute daily index changes.
   • Example Calculation:
     • If NEPSE index rises from 1,200 to 1,250, the percentage change is: $$\frac{1250-1200}{1200}\times 100=4.17\%$$

9. Exam Tip

• Focus on definitions: Know the difference between population vs. sample, parameter vs. statistic, and variable types.
• Practice classification: Given a dataset, classify variables (nominal/ordinal/interval/ratio) and identify data types (primary/secondary).
• Sampling questions: Expect questions on when to use stratified vs. random sampling.
• Ethics: Always justify why a method is ethical (e.g., "stratified sampling ensures representation").
• Real-world tie-ins: Relate examples to Nepali businesses (e.g., Daraz’s inventory management, Ncell’s network data).

10. Common Mistakes to Avoid

• Confusing nominal and ordinal data: Remember, ordinal has a meaningful order.
• Ignoring sampling bias: Always justify why a sample is representative.
• Overlooking ethics: Never collect data without consent or anonymization.
• Misapplying scales: Ratio data allows ratios (e.g., "twice as tall"), but interval does not (e.g., 20°C is not "twice as hot" as 10°C).

11. Practice Questions

1. Classify the following variables:

   • Blood type (A, B, AB, O) → Nominal
   • Customer loyalty tier (Bronze, Silver, Gold) → Ordinal
   • Temperature in °C → Interval
   • Annual salary → Ratio

2. A bank wants to survey customer satisfaction. Suggest a probability sampling method and justify why.

3. Differentiate between primary and secondary data with examples from Nepali businesses.

12. Summary Table

13. Final Visual: Data Collection Flowchart

Exam Alert: Always label axes, justify sampling methods, and tie examples to real businesses (e.g., Daraz, NEPSE, Ncell). Use tables and diagrams to organize answers clearly.