MODULE 2 – DATA COLLECTION & SAMPLING

1️⃣ Meaning of Data in Research

Data refers to raw facts, figures, observations, or responses collected for the purpose of analysis to
answer a research problem.

In business research, data becomes the foundation for decision-making.

Key Idea

Without data, research is only assumption.

With data, research becomes evidence-based.

Example (Crazy Café)

Research Problem:
“Why have sales declined at Crazy Café?”

To answer this, the café owner/researcher needs data such as:

 Customer visit frequency

 Price perception

 Satisfaction levels

 Competitor preferences

This collected information becomes data for analysis.

2️⃣ Types of Data

Data is broadly classified into:

A. Primary Data

B. Secondary Data

A. Primary Data

Meaning

Primary data is original data collected directly by the researcher for a specific research purpose.

It is collected first-hand and is tailored to the research problem.

Features of Primary Data

1. Original in nature
 2. Specific to the research problem

3. More reliable and relevant

4. Time-consuming and costly

Methods of Collecting Primary Data

 Survey

 Interview

 Observation

 Questionnaire

 E-questionnaire (Google Forms, apps)

Example (Crazy Café)

The café owner conducts:

 A survey of 100 students asking:

o How often do you visit cafés?

o What factors influence your choice?

 Interviews with regular customers

This data is primary data because it is collected directly.

B. Secondary Data

Meaning

Secondary data is data that has already been collected by someone else for another purpose but is
used by the researcher.

Sources of Secondary Data

 Company reports

 Government publications

 Research papers

 Websites and databases

 Literature (like books)

Features

1. Already available

2. Less costly and time-saving

3. May not be fully relevant

4. Needs verification

Example (Crazy Café)

The owner uses:

 Previous year sales records

 Online reviews of café

 Competitor pricing data

This is secondary data.

📊 Comparison: Primary vs Secondary Data

Basis Primary Data Secondary Data

Nature Original Already collected

Purpose Specific General

Cost High Low

Time More Less

Accuracy High May vary

3️⃣ Methods of Data Collection

Now we go deeper into how primary data is collected.

3.1 Survey Method

Meaning

A survey is a method of collecting data by asking questions to a large number of respondents.

Features

 Structured questions
  Large sample size

 Quantitative data

Example (Crazy Café)

Survey questions:

 How often do you visit cafés?

 Rate satisfaction (1–5)

When to Use

 When large data is required

 When statistical analysis is needed

3.2 Interview Method

Meaning

An interview involves direct interaction between researcher and respondent to collect detailed
information.

Types

 Structured interview

 Unstructured interview

Example (Crazy Café)

Interview question:

“Why have you stopped visiting Crazy Café?”

Advantage

 Provides deep insights

 Helps understand customer emotions

3.3 Observation Method

Meaning
Observation involves watching and recording behaviour without asking questions.

Example (Crazy Café)

Owner observes:

 Number of customers entering café

 Time spent inside

 Peak hours

Advantage

 Real behaviour (not influenced by responses)

3.4 Questionnaire Method

Meaning

A questionnaire is a method of collecting data by sharing the questions (hard copies) to a

large/medium number of respondents.

Features

 Structured questions

 Large/medium sample size

 Quantitative data

Example (Crazy Café)

MCQ questions:

 How often do you visit cafés? A. Once a week B. Twice a month …

 Rate satisfaction (1–5): A. Agree E. Disagree …

When to Use

 When large/medium data is required

 When statistical analysis is needed

3.5 E-Questionnaire
Meaning

A digital questionnaire shared through:

 Google Forms

 Mobile apps

 Email

Example (Crazy Café)

Questionnaire link shared to students via WhatsApp.

Data Selection Insights (Very Important)

Method When to select it

When you want to collect quantitative data from a large sample quickly, to
Survey find patterns, frequencies or general opinions (e.g., customer satisfaction,
attitudes, preferences).

When you need deep, qualitative insights from a smaller group, for exploring
Interview “why” and “how”, complex experiences, or sensitive topics, where follow-up
and probing are important.

When you want to study actual behaviour in natural or structured settings

Observation without relying on self-reports (e.g., classroom behaviour, shop-floor activities,
social interactions).

When you need standardised, written responses from many respondents

Questionnaire (often distributed in person) and can guarantee literacy and understanding of
questions.

When you want online, fast, low-cost distribution to a large, geographically

E-Questionnair
dispersed sample, and have access to internet-enabled respondents (e-mail,
e
web links, apps).

Classroom Activity

Task:

“Crazy Café wants to understand why customers stopped visiting.”

👉 Which method will you use and why?

Expected answers:

 Survey → large data

 Interview → deep insights

 Observation → actual behaviour

Questionnaire Design, Interviewing, Scheduling & E-Questionnaire

1️⃣ What is a Questionnaire?

A questionnaire is a systematic set of questions used to collect information from respondents for
research purposes.

It is the primary instrument for survey-based data collection.

🔍 Role in Research

Research Problem → Variables → Questionnaire → Data → Analysis

If the questionnaire is poor, data becomes unreliable, and the entire research fails.

Crazy Café Link

Research Problem:
Why are customers not visiting Crazy Café?

Variables:

 Price perception

 Service quality

 Ambience

 Wi-Fi availability

 Customer satisfaction

👉 Questionnaire must capture all these variables.

2️⃣ Types of Questions in Questionnaire

A. Open-Ended Questions

Meaning
Respondents answer in their own words. Open-ended questions are queries that cannot be
answered with a simple "yes," "no," or one-word answer, requiring descriptive, detailed responses
that encourage deeper thought.

Example

“Why did you stop visiting Crazy Café?”

Advantages

 Rich insights

 Useful in exploratory research

Limitations

 Difficult to analyse

 Time-consuming

B. Closed-Ended Questions

Meaning

Close-ended questions are structured queries that restrict respondents to specific, predefined
answers—such as "yes/no," multiple-choice or rating scales. Respondents choose from the given
options.

Types:

 Dichotomous Questions

 Multiple Choice Questions

 Likert Scale Questions (Rating Scale)

 Ranking Scale Questions

 Checklist-Style Questions

1. Dichotomous Questions

Binary choices, most commonly "Yes/No" or "True/False".

o Example 1: "Cleanliness of café needs improvement?” (True/False)

o Example 2: “Do you visit Crazy Café regularly?” (Yes/No)

2. Multiple Choice Questions

Provides several options, where respondents can select one or more answers

o Example: “What influences your café choice?”

  Price

 Ambience

 Service

 Location

3. Rating Scale Questions (Likert Scale)

Measures attitudes or feelings by asking the extent to which they agree or disagree, typically on a 5
or 7-point scale.

o Example: “Rate the service quality”

1 – Poor
2 – Average
3 – Good
4 – Excellent

5 – Brilliant

4. Ranking Questions

Respondents rate a product, service, or experience on a numerical or visual scale.

o Example: “Rank the following factors”

 Price

 Quality

 Location

5. Checklist-Style Questions: Allows respondents to check all applicable options from a list.

o Example: "Which of the following menus do you like to eat? (Select all that apply)"

 Idli-Wada
 Upama
 Avalakki
 Rice Bath
 Masala Dosa

3️⃣ Questionnaire Structure (Very Important for Students)

A good questionnaire must follow a logical sequence.

Standard Structure

Section Content

Introduction Purpose of the study

Screening Questions Identify relevant respondents

Main Questions Core data collection

Classification Questions Demographics

Example (Crazy Café)

Section 1: Introduction

“This survey is for academic research…”

Section 2: Screening

“Do you visit cafés?”

Section 3: Main Questions

 Frequency of visits

 Satisfaction rating

 Price perception

Section 4: Demographics

 Age

 Gender

 Student status

4️⃣ Guidelines for Good Questionnaire Design

1. Clarity

Questions must be simple and clear.

❌ Bad:
“Evaluate the service effectiveness dynamics.”
✅ Good:
“How would you rate the service?”

2. Avoid Ambiguity

❌ “Do you visit often?”

✅ “How many times per week do you visit?”

3. Avoid Leading Questions

❌ “Don’t you think prices are too high?”

✅ “How do you rate the price level?”

4. Logical Order

 General → Specific

 Easy → Difficult

5. Keep it Short

Long questionnaires reduce response quality.

5️⃣ Pretesting of Questionnaire

Meaning

Pretesting (Pilot testing) means testing the questionnaire on a small group before final use.

Purpose

 Identify unclear questions

 Check response time

 Improve reliability

Example

Test questionnaire on 10 students before surveying 100.

6️⃣ Interview Method

Meaning

A structured conversation where the researcher asks questions directly.

Types

Type Description

Structured Fixed questions

Unstructured Flexible discussion

Example (Crazy Café)

“Why do you prefer other cafés?”

Advantages

 Detailed responses

 Clarification possible

Limitations

 Time-consuming

 Interviewer bias

7️⃣ Schedule Method

Meaning

The researcher fills in responses instead of the respondent.

When Used

 Low literacy respondents

 Complex questionnaires

8️⃣ E-Questionnaire (Modern Context)

Meaning

Online questionnaire shared via:

  Google Forms

 Mobile apps

 Email

Advantages

 Fast data collection

 Low cost

 Easy analysis

Limitation

 Low response rate sometimes

9️⃣ Panel Research

Meaning

Data collected repeatedly from the same group of respondents over time.

Example

Tracking the same students’ café preferences for 6 months.

Use

 Study trends

 Study behaviour changes

🔟 Integrated Example – Full Questionnaire (Crazy Café)

Sample Questions

1. How often do you visit cafés?

2. Which café do you prefer?

3. Rate Crazy Café service (1–5)

4. How do you perceive the price?

5. What improvements do you suggest?

🎓 Classroom Activity

Task:

“Design 5 questions to study why customers stopped visiting Crazy Café.”

Expected Outcome

Students create:

 2 closed questions

 2 scale questions

 1 open-ended question

Scaling Techniques & Measurement (Nominal, Ordinal, Interval, Ratio)

1️⃣ Meaning of Measurement in Research

Measurement is the process of assigning numbers or labels to characteristics of objects or

respondents according to specific rules.

Measurement converts abstract concepts (like satisfaction, preference) into analysable data.

🔍 Why Measurement is Important

In business research, many variables are intangible:

 Satisfaction

 Attitude

 Preference

 Perception

These cannot be directly measured unless we use scales.

🧠 Crazy Café Example

Research problem:
“Why are customers not visiting Crazy Café?”

We cannot directly measure “satisfaction,” so we use:

👉 “Rate your satisfaction from 1 to 5”

This is a measurement using scaling.

2️⃣ Meaning of Scaling

A scale is a tool used to assign numbers or labels to variables so that they can be measured and
analysed.

Scaling helps convert qualitative responses into quantitative data.

3️⃣ Types of Scaling Techniques

There are four fundamental measurement scales:

1. Nominal Scale

2. Ordinal Scale

3. Interval Scale

4. Ratio Scale

📘 3.1 Nominal Scale

Meaning

Nominal scale classifies data into categories without any order or ranking. You can count responses
but cannot rank or average

Features

 Only labels

 No order

 No mathematical operations

Examples

Variable Categories

Gender Male / Female

Payment mode Cash / UPI / Card

Café choice Café A / Café B

🔍 Crazy Café Example

“Which café do you prefer?”

 Crazy Café
  Café X

 Café Y

Statistical Use

👉 Chi-square test

📘 3.2 Ordinal Scale

Meaning

Ordinal scale classifies data into ordered categories, but the difference between values is not
measurable. Ordinal scale arranges data in order or ranking, but the difference between ranks is not
equal.

Features

 Order exists

 Difference not equal

Examples

Rank Meaning

1 Poor

2 Average

3 Good

4 Excellent

🔍 Crazy Café Example

“Rank the following factors”

 Price

 Service

 Ambience

Statistical Use

👉 Median, Rank correlation

📘 3.3 Interval Scale

Meaning

Interval scale measures data with equal intervals between values, but no true zero.

Features

 Equal distance

 No absolute zero

Examples

 Temperature (°C)

 Satisfaction rating (1–5 scale)

🔍 Crazy Café Example

“Rate your satisfaction from 1 to 5”

Rating Meaning

1 Very Dissatisfied

2 Dissatisfied

3 Neutral

4 Satisfied

5 Very Satisfied

Statistical Use

👉 Mean, Standard deviation, Correlation

📘 3.4 Ratio Scale

Meaning

Ratio scale has all properties of interval scale + a true zero point.

Features
  Equal intervals

 True zero

 All mathematical operations possible

Examples

Variable Example

Sales ₹5000

Number of visits 5 visits/week

Waiting time 10 minutes

🔍 Crazy Café Example

“How many times do you visit per week?”

 0

 1

 2

 3

Statistical Use

👉 All statistical tools (Regression, ANOVA, etc.)

📊 Comparison Table (Very Important)

Feature Nominal Ordinal Interval Ratio

Nature Categories Ordered Equal intervals Absolute measure

Order ❌ ✔ ✔ ✔

Equal distance ❌ ❌ ✔ ✔

True zero ❌ ❌ ❌ ✔

Example Gender Ranking Rating scale Sales

🧠 Easy Memory Trick

Nominal → Name
Ordinal → Order
Interval → Equal intervals
Ratio → Real zero

4️⃣ Linking Scales to Statistical Tests (Very Important)

Scale Test

Nominal Chi-square

Ordinal Rank correlation

Interval Correlation

Ratio Regression, ANOVA

🎓 Classroom Activity

Task:

“Identify the scale type”

Variable Scale

Gender

Satisfaction rating

Monthly spending

Rank of café

Expected Answers

 Gender → Nominal

 Satisfaction → Interval

 Spending → Ratio

 Rank → Ordinal

Variable Scale

Customer Membership Nominal

Satisfaction level (Poor, Average, Good) Ordinal

Variable Scale

Temperature of café (°C) Interval

Number of customers per day Ratio

Payment method (Cash/UPI/Card) Nominal

Rank of service quality Ordinal

Sampling Techniques, Sample Size, Confidence & Sampling Error

INTRODUCTION TO SAMPLING

Meaning and Definition

Sampling is the process of selecting a subset of elements from a defined population for the purpose
of making inferences about the entire population.

Kerlinger (1986): Sampling is the process of selecting a number of individuals for a study in such a
way that the individuals selected represent the larger group.

Conceptual Understanding

In business research, it is often impractical, costly, or impossible to study the entire population.
Therefore, researchers rely on a representative subset (sample).

The key notes:

“A properly selected sample reflects the characteristics of the population.”

“Good sampling leads to good data, and good data leads to good decisions.”

☕ Crazy Café Illustration

Population: All students near Crazy Café (2000 students)

Sample: 120 students surveyed

👉 If sampling is done correctly, conclusions drawn from 120 students can represent all 2000.

🔍 Why Sampling is Needed

Studying the entire population (census) is:

 Time-consuming
  Costly

 Sometimes impossible

🧠 Crazy Café Example

Population:

All students near the café (say 2000 students)

Sample:

100 students selected for survey

👉 These 100 students represent the population.

2️⃣ Key Terms in Sampling

Term Meaning

Population Entire group (all customers)

Sample Subset of population

Sampling Unit Individual respondent

Sampling Frame List of population

3️⃣ Types of Sampling Techniques

Sampling is divided into:

A. Probability Sampling

B. Non-Probability Sampling

Basis Probability Sampling Non-Probability Sampling

Selection Random Non-random

Bias Low High

Accuracy High Moderate

Cost Higher Lower

Use Descriptive & causal research Exploratory research

PROBABILITY SAMPLING TYPES

Probability sampling ensures that each element in the population has a
known and equal probability of selection. This is achieved through different
structured methods.
Each method is designed to:
 Improve representativeness
 Reduce bias
 Enable statistical testing

2 SIMPLE RANDOM SAMPLING

2️⃣
Meaning
Simple Random Sampling is a technique in which every element of the
population has an equal and independent chance of being selected.

Conceptual Explanation
 Selection is purely random
 No bias or grouping
 Each unit is treated equally

Methods of Selection
 Lottery method
 Random number tables
 Computer-generated randomization

Crazy Café Example

Population: 2000 students
Sample: 100 students
Method: Assign numbers (1–2000) → randomly pick 100 numbers
👉 Every student has equal chance

Advantages
 Simple and easy to understand
 Eliminates selection bias
 Statistically valid

Limitations
 Requires complete population list
 Not suitable for large, scattered populations

3 SYSTEMATIC SAMPLING
3️⃣
Meaning
Systematic sampling involves selecting every kth element from the population
after a random start.

Formula

N
k=
n
Where:
N = Population size
n = Sample size

Crazy Café Example

Population = 2000
Sample = 200
k=2000/200=10
👉 Select every 20th student

Conceptual Insight
 First element is selected randomly
 Remaining are selected systematically

Advantages
  Easy to implement
 Saves time
 Ensures uniform coverage

Limitations
 Risk of periodic bias
 Not suitable if pattern exists in data

4 STRATIFIED SAMPLING
4️⃣
Meaning
Stratified sampling involves dividing the population into homogeneous groups
(strata) and selecting samples from each group.

Conceptual Logic
 Population is heterogeneous
 Each subgroup is homogeneous

Crazy Café Example

Population divided into:
 UG students
 PG students
Sample taken proportionately from each group

Advantages
 High accuracy
 Ensures representation of all groups
 Reduces sampling error

Limitations
 Requires detailed population information
 More complex to implement
5 CLUSTER SAMPLING
5️⃣
Meaning
Cluster sampling involves dividing the population into clusters (groups) and
selecting entire clusters randomly.

Conceptual Logic
 Each cluster is a mini population
 Used when population is geographically dispersed

Crazy Café Example

Clusters = Colleges
👉 Select 2 colleges randomly
👉 Survey all students in those colleges

Advantages
 Cost-effective
 Useful for large populations
 Reduces travel/time

Limitations
 Lower accuracy than stratified sampling
 Higher sampling error

6 COMPARISON OF PROBABILITY SAMPLING TYPES

6️⃣

Method Key Idea Best Use Limitation

Simple
Equal chance Small population Needs full list
Random

Risk of pattern
Systematic Every kth unit Ordered data
bias

Group-based
Stratified Heterogeneous population Complex
sampling

Cluster Group selection Large/geographical Less accurate

Method Key Idea Best Use Limitation

population

NON-PROBABILITY SAMPLING
Meaning and Definition

Non-probability sampling is a technique in which the probability of selecting each element is

unknown, and selection depends on convenience, judgment, or accessibility.

Key Idea: Not all elements have an equal chance of being selected.

Conceptual Understanding

Unlike probability sampling, this method is:

 Non-random

 Based on practical considerations

 Widely used in exploratory and field studies

☕ Crazy Café Example

Surveying only customers present inside the café during evening hours:

👉 Easy to collect data

👉 But may not represent all customers

Types:

1. CONVENIENCE SAMPLING
Meaning

Convenience sampling involves selecting respondents who are easily available and accessible.

Conceptual Logic

 Based on ease and availability

 No systematic selection

Crazy Café Example

Surveying:

 Students sitting in the café

 Friends or classmates

Advantages

 Quick and inexpensive

 Easy to implement

Limitations

 High bias

 Low representativeness

 Cannot generalise results

2. JUDGMENT SAMPLING

Meaning

Judgment sampling involves selecting respondents based on the researcher’s expertise and
judgment.

Conceptual Logic

 Researcher selects relevant and knowledgeable respondents

Crazy Café Example

Selecting:
  Regular customers

 Café staff

 Frequent visitors

Advantages

 Focused data

 Useful for expert opinions

Limitations

 Subjective

 High risk of bias

3. QUOTA SAMPLING

Meaning

Quota sampling involves selecting a fixed number of respondents from different categories.

Conceptual Logic

 Ensures representation of groups

 Similar to stratified sampling but non-random

Crazy Café Example

Sample of 100 students:

 50 UG students

 50 PG students

Selection within groups is non-random

Advantages

 Ensures group representation

 More balanced than convenience sampling

Limitations
  Selection bias within groups

 Not fully scientific

4. SNOWBALL SAMPLING

Meaning

Snowball sampling involves selecting respondents through referrals from initial participants.

Conceptual Logic

 One respondent leads to another

 Sample “snowballs” in size

☕ Crazy Café Example

Surveying:

 One regular customer

 That customer refers 3 more

 They refer others

Advantages

 Useful for hidden populations

 Expands sample quickly

Limitations

 Highly biased

 Limited control

7️⃣ COMPARISON OF NON-PROBABILITY METHODS

Method Basis Best Use Limitation

Convenience Availability Student surveys High bias

Judgment Researcher decision Expert studies Subjective

Quota Group proportion Balanced sample Non-random

Method Basis Best Use Limitation

Snowball Referral Hidden population Bias accumulation

4️⃣ Sample Size Determination

Meaning

Sample size is the number of respondents selected for the study.

Factors Affecting Sample Size

1. Population size

2. Research objective

3. Time & cost

4. Accuracy required

Rule of Thumb (UG Level)

Population Sample Size

Small (<500) 50–100

Medium (1000–5000) 100–200

Large 200+

🧠 Crazy Café Example

Population = 2000 students

Sample = 120 students

5️⃣ Confidence Level

Meaning

Confidence level indicates how sure we are about the results.

Common Levels

Confidence Level Meaning

90% Moderate confidence

Confidence Level Meaning

95% Standard level

99% Very high confidence

Example

95% confidence means:

👉 Results are correct in 95 out of 100 cases

6️⃣ Sampling Error

Meaning

Sampling error is the difference between the sample result and the actual population result.

Example

Sample result: 60% satisfied

Actual population: 65% satisfied

👉 Error = 5%

Causes

 Small sample size

 Poor sampling method

 Bias

A. Confidence Level

Confidence level indicates how sure we are about our results.

Common Values

Level Z-value

90% 1.645

95% 1.96

99% 2.576
B. Margin of Error (e)

Margin of error is the acceptable level of error.

Example:

±5% means results may vary by 5%.

C. Proportion (p)

Estimated percentage of the population having a characteristic.

Example:

50% customers are satisfied → p = 0.5

6️⃣ Standard Result (Must Remember)

At:

 Confidence = 95%

 Error = 5%

👉 Sample size ≈ 384

🧠 Interpretation

If the population is large:

👉 Minimum sample = 384 respondents

8️⃣ Practical Rule for Students

Population Sample Size

< 500 50–100

500–2000 100–200

Large (>5000) 200–400

Sampling Error, Types of Errors, Practical Decisions & Integration

1️⃣ Meaning of Sampling Error

Definition

Sampling error is the difference between the sample result and the actual population result.

Since we study only a sample (not the entire population), some error is inevitable.

🔍 Simple Example (Crazy Café)

Actual population:

👉 65% customers are satisfied

Sample result:

👉 60% customers are satisfied

👉 Sampling Error = 5%

🧠 Key Insight

Sample ≠ Population

Difference = Sampling Error

2️⃣ Types of Errors in Research

Errors are broadly classified into:

A. Sampling Errors

B. Non-Sampling Errors

A. Sampling Errors

These occur because only a sample is studied, not the entire population.

Causes

1. Small sample size

2. Poor sampling method

3. High variability in population

Example

Surveying only 10 students instead of 100

👉 Results may not represent all students

How to Reduce Sampling Error

 Increase sample size

 Use probability sampling

 Ensure proper representation

B. Non-Sampling Errors

These errors occur due to mistakes in data collection, measurement, or analysis.

Types of Non-Sampling Errors

1. Response Error

Respondents give incorrect answers

Example:

Student says “I visit café daily” but actually visits once a week

2. Measurement Error

Poor questionnaire design

Example:

Confusing or ambiguous questions

3. Interviewer Bias

Interviewer influences responses

4. Data Processing Error

Mistakes in data entry or analysis

🔍 Example (Crazy Café)

 Wrong data entry in Excel

 Misinterpretation of satisfaction scale

📊 Comparison

Basis Sampling Error Non-Sampling Error

Cause Sample selection Human/method errors

Basis Sampling Error Non-Sampling Error

Control Can be reduced Can be avoided

Example Small sample Wrong question

A CASE STUDY FOR BRM NOTEBOOK

ABC College launched a mobile app called Campus Connect to help students with:

• Academic updates

• Event notifications

• Assignment submissions

• Communication with faculty

Initially, the app had high engagement. However, after 3 months:

• Active users dropped by 37%

• Students stopped checking notifications

• Event participation reduced

Research Problem

“Why has student engagement declined on the Campus Connect App?”

Sampling Challenge

The college wants to conduct a survey of students, and the total number of students are 5500

Q: Try all methods under probability and non-probability sampling. Which sampling method will
provide the best results for effective decision-making in this case? Justify your answer.

QUESTIONS BANK:

1. Explain the meaning of data in research and its types with suitable examples.
2. Describe different methods of data collection in business research with examples.
3. Explain the concept and types of questionnaires with examples and explain the
guidelines for preparing a good questionnaire.
4. Explain the meaning of scaling and describe different types of scaling
techniques with suitable examples.
5. Explain the meaning of sampling and discuss the types of sampling in business
research.