Essentials of

Marketing
Research
Part 2: Designing the
Marketing Research Project

Session 6: Descriptive Research

and Sampling
 Descriptive Research Designs

Selection of a descriptive research design is based on three factors.

• The nature of the initial problem or opportunity.
• The research questions.
• The research objectives.

A descriptive design is appropriate in the following situations.

• When the research problem/opportunity is either to describe
characteristics of existing market situations or to evaluate
current marketing mix strategies.
• If research questions include a who, what, where, when, and how for
target populations or marketing strategies.

2
 Descriptive Research Surveys

There are two general approaches used to

collect data for descriptive research.
• Asking questions.
• Observation. The main goal of
quantitative survey
research methods is to
provide facts and
Survey research methods collect
estimates from a large,
quantitative data from large groups of
representative sample
people through a question/answer process.
of respondents.
Quantitative studies are descriptive in the
sense they use numbers and statistics to
summarize.

3
 Advantages and Disadvantages of Quantitative Survey
Research Designs

Can accommodate large sample

sizes so results can be generalized
to the target population. Questions that accurately
Produce precise enough measure respondent attitudes
estimates to identify even small and behavior can be
differences. challenging to develop.
Easy to administer and record In-depth data difficult to obtain.
answers to structured questions. Low response rates can be a
Facilitate advance statistical problem.
analysis.

4
 Types of Errors in Surveys

Sampling error is the difference between the findings based on the

sample and the true values for a population.
• It can be reduced by increasing sample size and using appropriate
sampling methods.

Errors occurring in survey design

Nonsampling errors have these
not related to sampling are called
characteristics.
nonsampling errors.
• They create “systematic
• Respondent error.
variation” or bias in the data.
• Measurement/questionnaire
• They are controllable.
design errors.

5
 Respondent Errors

Respondent errors occur when

respondents cannot be reached,
are unwilling to participate, or
respond untruthfully. Human memory is a source of
Nonresponse error occurs then a response error.
sufficient number of preselected • Selective perception.
respondents do not participate.
• Time compression.
• Multiple callbacks.
• Use of averaging to overcome
• Follow-up mailings. memory retrieval problems.
• Incentives. • Using a socially desirable
• Enhancing sponsor credibility. manner and introducing bias.
• Indicating length of time
needed to complete.
• Shorter questionnaires.

6
Types of Survey Methods – Person-Administered Surveys

Person-administered survey methods use trained interviewers.

Advantages.
• Adaptability – interviewers quickly adapt
to respondents’ differences.
• Rapport – interviewers establish a Disadvantages.
“comfort zone” for respondents.
• Possible recording error.
• Feedback – interviewers explain
instructions and answer any questions • High expense.
while noting verbal and nonverbal cues.
• Quality of responses – interviewers
screen respondents for target population
and face-to-face interactions elicit truth.

7
 Person-Administered Surveys – In-Home Interviews

An in-home interview is a face-to-face structured question and answer

exchange conducted in the respondent’s home or office.

Advantages.
• Interviewers can explain confusing or Disadvantages.
complex questions and use visual • Unsupervised
aids. interviewers may skip
• Respondents can try new products or homes or fabricate
watch potential ad campaigns and interviews.
evaluate them. • Time-consuming
• Respondents are in a familiar and expensive.
environment and more likely to answer
the survey’s questions.

8
Person-Administered Surveys – Mall-Intercept Interviews

A mall-intercept interview is a face-to-face interview, shoppers are

stopped and asked to complete a survey.

The disadvantages are similar to those of

Mall-intercept interviews in-home interviews but interviewer’s travel
share the advantages of time is reduced.
in-home interviews, but Mall patrons are not likely to be
the environment is not as representative of the target population,
familiar. even if screened.
They are less expensive Mall-intercept interviews typically use
and more convenient for nonprobability sampling, which
the researcher. adversely affects the ability to generalize
results.

9
Types of Survey Methods – Telephone-Administered Surveys

Telephone interviews are less expensive than face-to-face interviews,

and more suitable for gathering data from large numbers of respondents.

Advantages.
• Interviewers are supervised at a Disadvantages.
central work location. • Only audio can be used.
• Less expensive than face-to-face • Complexity of questions.
interviews.
• Respondents may hang up.
• Allows respondents from a wide
• Less than 50 percent of U.S.
geographic area, and data is
collected quickly. households have a landline.

• Call backs are possible. • Refusal rates are high.

• Random dialing selects a random • They annoy many people.

sample.

10
Telephone-Administered Surveys – Mobile Phone Surveys

Mobile phone surveys are growing in use for these reasons.

• High percentage of mobile phone usage.
• Availability of mobile phone applications.
• Rapid decline in landline phone penetration.
Disadvantages.
Advantages over Internet/phone
surveys. • Cost to respondents.
• Immediacy. • Safety is a potential issue.
• Portability. • Not suitable for complex
questions.
• They reach consumers with no
Internet or landline available. • Relatively small sample
sizes.

11
 Types of Survey Methods – Self-Administered Surveys

A self-administered survey has the respondent respond in their own

time, without the presence of an interviewer.

Disadvantages.
Advantages.
• Limited flexibility.
• Low cost per survey.
• High nonresponse rates.
• Respondent control.
• Potential response errors.
• No interviewer-respondent bias.
• Slow data acquisition.
• Anonymity in responses.
• Lack of monitoring capability.

12
 Self-Administered Surveys – Mail Surveys

Mail surveys typically are sent to respondents using the postal service.

Disadvantages.
Advantages. • Lower response rates which
 Inexpensive to implement. creates nonresponse bias.
 Reaches hard-to-interview • Misunderstood or skipped
respondents. questions.
• Slow acquisition of data.

13
 Self-Administered Surveys – Mail Panel Surveys

A mail panel survey is a questionnaire sent to a group of individuals who

have agreed to participate in advance.

Advantages.
Disadvantages.
• Can be tested prior to the survey.
• Members are often not
• High response rate.
representative of the target
• Can be used for longitudinal population at large.
research.

14
 Self-Administered Surveys – Online Survey Methods

The most frequently used survey method is online surveys.

Advantages.
• Less expensive per respondent
than other survey methods.
•
Disadvantages.
Collects data from hard-to-reach
samples. • Internet samples are rarely
• Can randomize question order. representative.
• Missing data can be eliminated. • Nonresponse bias can be high.
• Improved graphic capabilities. • Limited ability to generalize to
the general population.
• Companies can survey
customers using email.

15
 Selecting the Survey Method – Situational Factors

Quality requirements.
The goal is to produce usable
data in as short a time as possible • Completeness of data refers to
at the lowest cost. the depth and breadth of data.
Budget. • Generalizable data accurately
represents the population
• Includes all resources, not just
studied and can be projected
dollar amounts.
to the target population.
Completion time frame.
• Small sample size limits
• Direct mail or interviews generalizability.
require long time frames.
• Data precision.
• Online surveys, telephone
• Mail and online surveys can
surveys and mall intercepts
be precise but not
can be done more quickly.
generalizable.

16
 Selecting the Survey Method – Task Factors

Amount of information needed from

Difficulty of the task. respondent.
• Researchers should make it as • Collecting more data lowers
easy as possible for response rates and increases
respondents to answer. respondent fatigue.

Stimuli needed to elicit response. Research topic sensitivity.

• Common examples are • Topic sensitivity is the degree a
products – as in taste tests. question leads a respondent to
give a socially acceptable
• And promotional visuals – as in response.
advertising research.
• In phone and face-to-face
• Online and personal surveys interaction undesirable behaviors
facilitate visual stimuli. are underreported and desirable
behaviors over-reported.

17
 Selecting the Survey Method – Respondent Factors

Diversity.
Respondent participation.
• The more diverse the
respondents, the fewer • Ability to participate – ability
similarities they share. to get meet with researcher.
• The less diverse the • Willingness to participate –
respondents, the more inclination to share thoughts.
similarities they share. • Knowledge level – if they feel
Incidence rate is the percentage they have the knowledge to
of the general population that is answer questions on the topic.
the focus of the research. • Key informants have high
• When incidence rates are low, knowledge levels.
researchers spend time and .
money locating respondents.

18
 Essentials of
Marketing
Research
Part 3: Gathering and
Collecting Accurate Data

Sampling: Theory and Methods

19
 Value of Sampling as a Part of the Research Process

Sampling is selection of a small number of elements from a larger

defined target group of elements.
• Expecting the information gathered from the small group will allow
judgments to be made about the larger group.
Sampling is often used when it is impossible or unreasonable to conduct
a census.
• A census is a research study that includes data about every member
of the defined target population.
Sampling is less time-consuming and less costly than a census.
• Samples play an important indirect role in designing questionnaires.

20
 Sampling Theory Basics – Population

A population is an identifiable group of elements of interest to the researcher and

pertinent to the information problem.

A defined target population is the complete set of elements identified for investigation
based on the objectives of the research project.

Sampling units are target population elements actually available to be used during the
sampling process.

21
 Factors Underlying Sampling Theory

A major goal of researching small, representative samples is that the

results will help predict or estimate what true population parameters are
within a certain degree of confidence.
The Central Limit Theorem (CLT) describes the theoretical
characteristics of a sample population.
• The sampling distribution of the mean () or the percentage value ()
derived from a simple random sample will be normally distributed, if
the sample size is sufficiently large – when n is ˃ or = 30.
• The mean () of the random sample fluctuates around the true
population mean (µ) with a standard error of and a normal sampling
distribution.
• There is a high probability the mean of any sample () taken from the
target population will be a close approximation of the true target
population mean (µ), as you increase the size of the sample (n).

22
 Tools Used to Assess the Quality of Samples

Sampling error is any bias resulting from mistakes in either the

selection process for prospective sampling units, or in determining the
sample size.
Random sampling error occurs as chance variations in the selection of
sampling units.
Sampling error can be reduced by increasing the size of the sample.
Nonsampling errors
Nonsampling error occurs regardless of
are related to the
whether a sample or a census is used. For
accuracy of the
example:
data, sampling
• The target population may be inaccurately errors relate to the
defined causing population frame error. representativeness
There is no way to assess the impact of of the sample to the
nonsampling errors on data quality. defined target
population.

23
 Probability and Nonprobability Sampling

In nonprobability sampling, the

In probability sampling, each
probability of selecting each
sampling unit in the defined target
sampling unit is not known.
population has a known probability
of being selected for the sample. • Sampling error is not known.
Rule ensure the following: Selection of sampling unit is
based on researcher judgment
• Unbiased selection of the
and may or may not represent the
sampling units.
target population.
Results using probability sampling
• Depends on the sampling
designs can be generalized to the
approach and how well the
target population within a specified
researcher executes the
margin of error.
selection activities.

24
Probability Sampling Designs – Simple Random Sampling

Simple random sampling occurs when every sampling unit has a

known and equal chance of being selected.

There are several advantages.

The primary disadvantage is the
• The technique is easily
difficulty of obtaining a complete
understood.
and accurate listing of the target
• Results are generalized with population elements.
margin of error.
• Simple random sampling
• They produce unbiased requires all sampling units be
estimates of the population’s identified.
characteristics.
• This works best for small
• This method results in a valid populations where accurate
representation of the defined lists are available.
target population.

25
Probability Sampling Designs – Systematic Random Sampling

Systematic random sampling requires the target population be

ordered in some way – inexpensive and can be done quickly.
• Sampling units are selected using a skip interval.
Step 1: Obtain a list of potential sampling units containing an acceptable
frame of the target population elements.
Step 2: Determine the total number of sampling units making up the list of
the defined target population’s elements and the desired sample size.
Step 3: Compute the needed skip interval by dividing the number of
potential sampling units on the list by the desired sample size.
Step 4: Using a random number-generation system, randomly determine
a starting point to sample the list of names.
Step 5: With a name as the first sample unit, apply the skip interval to
determine the remaining names that should be included in the sample.

26
Systematic Random Sampling – Advantages and Disadvantages

Frequently used as an easy way The greatest weakness is the

to draw a sample while ensuring possibility of hidden patterns in
randomness. the list of names, creating bias.
The availability of lists and the The number of sampling units in
shorter time required to draw a the target population must be
sample makes this an attractive, known.
economical method. When the size of the target
population is large or unknown,
identifying the number of units is
difficult, and estimates may not be
accurate.

27
Nonprobability Sampling Designs – Convenience Sampling

Convenience sampling draws samples at the convenience of the

researcher.

Convenience sampling enables a large number of respondents be

interviewed in a short time period.
• Commonly used in the early stages of research and pretesting of
questionnaires.
The data is not generalizable to the defined target population.
The representativeness of the sample cannot be measured because
sampling error estimates cannot be calculated.

28
Nonprobability Sampling Designs – Judgment Sampling

In judgment sampling, or purposive sampling, respondents are selected

because the researcher believes they meet the requirements of the
study.
If researcher judgment is correct, the sample generated by judgment
sampling will be better than one generated by convenience sampling.
However, the representativeness of the sample cannot be measured.

29
 Nonprobability Sampling Designs – Snowball

In snowball sampling or referral sampling, a set of respondents is

chosen, and they help the researcher identify additional respondents.
Typically used when:
• The defined target population is small and unique.
• Or when compiling a complete list of sampling units is very difficult.
This method identifies respondents who are members of small, uniquely
defined, hard-to-reach target populations.
• Most useful in qualitative research.
Like all other nonprobability sampling approaches, the ability to
generalize the results to members of the target population is limited.

30
 Determining Probability Sample Sizes

Three factors play a role in sample sizes with probability designs.

• The population variance, a measure of the dispersion of the population,
and its square root – referred to as the population standard deviation.
• The level of confidence desired in the estimate.
• The degree of precision desired – the acceptable amount of error in
the sample estimate.

When estimating a population mean, the formula for

calculating the sample size is:

When estimates of a population is a concern, the

formula for calculating the needed sample size would
be:

31
 Nonprobability and Other Sample Size Approaches

Formulas cannot be used for nonprobability samples.

• Usually a subjective, intuitive judgment by the researcher.
• Sampling results should not be used to make statistical inferences
about the true population parameters.
Sample sizes are often determined using less formal approaches.
• Such as budget or previous studies’ results.
Decisions on approach, or combination of approaches, requires judgment
of both research experts and managers to select the best alternative.

32
 Steps in Developing a Sampling Plan

A sampling plan is the blueprint

ensuring the data is
representative of the defined • Step 5: Determine necessary
target population. sample sizes and overall
• Step 1: Define the target contact rates.
population. • Step 6: Create an operating
• Step 2: Select the data plan for selecting sampling
collection method. units.
• Step 3: Identify the sampling • Step 7: Execute the operational
frame(s) needed. plan.
• Step 4: Select the appropriate
sampling method.

33