3 – Research Design

Introduction

A research design is just a study plan. This serves as a guide for gathering and analysing data. It might be referred to as a blueprint for carrying out the research. It is analogous to an architect’s blueprint for building a house; if research is undertaken without a blueprint, the outcome will likely differ from what was envisioned at the outset. The blueprint comprises

(1) interviews to be performed, observations to be made, experiments to be carried out, and data analysis to be carried out.

(2) Data collection tools, such as questionnaires

(3) What sampling procedures were used?

3.1  Overview

The structure of research can be thought of as research design; it is the “glue” that ties all of the pieces together in a research effort. A successful design results from a collaborative effort that includes solid planning and communication.

There are three study designs: exploratory, descriptive, and causal.

Exploratory research is conducted to understand the general nature of the problem and specify the relevant variables that must be considered. This form of research does not require prior knowledge, and the research methodologies are flexible, qualitative, and unstructured.

Descriptive research is an extensively used method in marketing. In general, a hypothesis is presented in a descriptive study, and questions such as size, distribution, and so on are asked in relation to this hypothesis.

Causal research is concerned with determining the relationship between cause and effect. Typically, experiments are carried out in this form of research.

3.1.1 The Importance of Research Design

• Before beginning the research process, a practical and relevant research design should be developed. A study design is required because it delivers the following advantages:
• It aids in the seamless execution of numerous research processes.
• It necessitates less effort, time, and money; it aids in planning the methodologies and techniques for data collection and analysis; and it aids in achieving the study objectives within staff, time, and money constraints.
• Before developing a research plan, the researcher should examine the following factors:
• The procedure for collecting information source
• The skills of the researcher and the co-ordinating staff, the problem objectives, the nature of the problem, and the time and money available for research activity are all important factors.

3.1.2 Different Types of Research Designs

The most common categories are exploratory, descriptive, and causal research. Exploratory research is done to gain an understanding of the general nature of the problem. It specifies the relevant variables that must be considered. This research has no prior knowledge, and the methodologies are flexible, qualitative, and unstructured. In this strategy, the researcher has no idea “what he will find.”

Descriptive research is an extensively used method in marketing. In general, a hypothesis is presented in a descriptive study, and questions such as size, distribution, and so on are asked in relation to this hypothesis.

Causal research is concerned with determining the relationship between cause and effect. Typically, experiments are carried out in this form of research.

3.2 Exploratory Research

The primary focus of exploratory research is on transforming significant, imprecise problem statements into small, detailed sub-problem statements to create particular hypotheses. A hypothesis is a statement that states, “How are two or more variables related?”

In the early stages of the study, we usually lack an adequate understanding of the problem to establish a specific hypothesis. Furthermore, multiple possible causes frequently exist.

For example, “Sales are down because our pricing is too high,” “Our dealers or salespeople aren’t doing a good job,” “Our advertising is ineffective,” and so on.

In this case, very little information is available to determine the root cause of the problem. Exploratory research’s primary goal is to narrow down the scope of the problem, so it is used in the early stages of research.

Under what conditions is an exploratory research ideal?

The following are the situations in which an exploratory study would be ideal:

• To develop an understanding of the problem
• To come up with new product concepts
• To make a list of all possible outcomes. We must prioritise the possibilities that appear likely among the many that exist.
• Creating hypotheses on occasion

For example, consider a market researcher working for a (new entrant) company for the first time.

• To identify priorities to do additional study.
• Exploratory research can be utilised to explain concepts and assist in formulating specific challenges.

For example, management is considering changing the contract policy in the aim of increasing channel member happiness.

An exploratory study can be used to clarify the current condition of channel member satisfaction and to design a mechanism for measuring channel member satisfaction.

• To put a draught questionnaire through its paces.
• In general, exploratory research is suited for any subject about which little is known. This study serves as the foundation for any subsequent research.

3.2.1 Exploratory Research Characteristics

• Exploratory research is adaptable and versatile.
• No organised forms are employed for data gathering.
• Experimentation isn’t required.
• The cost of conducting the investigation is modest.
• This form of research enables a broad range of viewpoints to be explored.
• Research is interactive and open-ended in nature.

3.2.2 Hypothesis Development at Exploratory Research Stage

• If the situation is being explored for the first time, it may be impossible to establish any hypothesis. This is due to the lack of prior data.
• There may be situations when information is available, and a tentative hypothesis can be developed.
• In other circumstances, most data is available, and solutions to the problem may be possible.

The following examples demonstrate each of the types above:

Example:

Research Purpose	Research Question	Hypothesis
1.   What product feature, if stated, will be most effective in the advertisement?	What benefit do people derive from this Ad appeal?	No hypothesis formulation is possible.
2.   What new packaging will the company develop (concerning a soft drink)?	What alternatives exist to provide a container for soft drinks?	Paper cups are better than any other form, such as a bottle.
3.   How can our insurance service be improved?	What is the nature of customer dissatisfaction?	Impersonalization is the problem.

In Example 1, the study question is, “What benefit do consumers desire from the Ad?” No prior research has examined this product’s consumer benefits, so no hypothesis can be formed.

In Example 2, some information on soft drink packaging is currently available. It is feasible to formulate a speculative theory here. The hypothesis presented here could be just one of several possibilities.

In Example 3, the root reason of consumer unhappiness is identified: a lack of personalised service. In this scenario, it is feasible to determine whether or not this is a cause.

3.2.3 Formulation of Hypothesis in Exploratory Research

Using any of the four strategies listed below is the quickest and cheapest way to create exploratory research hypotheses.

1. Literature Search refers to “consulting the literature to build a fresh hypothesis.” Trade journals, professional journals, market research finding publications, statistical publications, and so on are examples of the literature alluded to. Assume one of the issues is “Why are sales down?” This can be rapidly analysed using publicly available data, which should reveal “whether the problem” is an “industry problem” or a “firm problem.” There are three ways to formulate the hypothesis.

• Although the company’s market share has decreased, industry data remain typical.
• Because the industry is declining, so is the company’s market share.
• The industry’s share is increasing while the company’s share is decreasing.

If our company’s sales are down despite the market’s upward trend, we must examine the elements of the marketing mix.

Example:

• A television manufacturer believes its market share is dropping even though the entire television sector is doing well.
• A trade embargo enforced by a country reduces textile exports, which in turn reduces sales of a company producing garments for export.

The above data can be utilised to identify the cause of dropping sales.

2. Experience Survey: It is preferable to conduct experience surveys with people well-versed in the subject matter under investigation. These individuals could be company executives or individuals from outside the organisation. There is no need for a questionnaire in this case. The approach used in an experience survey should be very unstructured, allowing respondents to express differing opinions.

• A group of housewives might be approached to select a “ready-to-cook product.”
• A publisher may wish to investigate the reason for the low circulation of a freshly introduced newspaper. He might run into:
• Newspaper vendors,
• Public reading rooms,
• Members of the general public,
• Members of the business community, and so on.

These are experienced individuals whose knowledge the researcher can benefit from.

3. Focus Group: Another strategy commonly employed in exploratory research is the focus group. A focus group is a small group of people who study and discuss a specific topic. A moderator facilitates the dialogue. The group usually consists of 8-12 people. When picking these individuals, attention must be paid to ensure they share a shared background and purchasing histories. This is necessary since there should be no disagreement among group members about the shared problems being discussed. During the debate, prospective purchasing attitudes, current purchasing opinions, and so on are obtained.

Most companies that run focus groups examine the candidates first to determine who will make up the specific group. Firms also avoid groups in which some members have friends and relatives present, as this leads to a skewed conversation. Typically, some similar groups are formed, and the conclusions of the various groups are used to formulate the hypothesis. As a result, having similar groups is an essential component in focus groups. There are usually 4-5 groups. Some of them may have as many as 6-8 groupings. The guiding criteria are whether the latter groups are developing new ideas or repeating themselves about the subject under study. When the group’s return began to dwindle, the conversations stopped. A regular focus group lasts between 1 and 30 hours to 2 hours. The moderator of the focus group plays an important role. His job is to steer the group on the proper path.

A moderator/facilitator should have the following characteristics:

• Listening: He must be able to listen well. Due to a lack of attention, the moderator must not overlook the participant’s comment.
• Permissive: The moderator must be permissive while also looking for signs that the group is collapsing.
• Memory: He must have an excellent memory. The moderator must be able to recall the participants’ comments. For instance, consider a conversation centred on a new commercial by a telecommunications provider. The participant may make an early remark and then make a later statement that contradicts what was said earlier. For example, the participant may claim that he or she has never agreed with the opinions of the competitor’s advertisement but then claim that the “current advertisement of competitor is outstanding.”
• Encouragement: The moderator must urge members who are not responding to participate.
• Learning: He should be able to learn quickly.
• Sensitivity: The moderator must be sensitive to direct the group discussion.
• Intelligence: He must be a person of above-average intelligence.
• Gentle/firm: He must balance detachment and empathy.

Four. Case Studies Analyzing a specific situation might sometimes provide insight into the problem under investigation. Case studies of companies that have been through similar circumstances may be accessible. These case studies are ideal for conducting exploratory research. However, case history investigation findings are always regarded as suggestive rather than conclusive. Many case histories may be accessible in the form of past research conducted by rivals if “ready-to-eat food” is preferred. We must carefully investigate the previously published case studies in terms of other variables such as pricing, advertisement, changes in taste, and so on.

3.2.4 Secondary Data

Secondary data is information collected for reasons other than completing a research project. The researcher can access secondary information sources while gathering data about an industry, potential product applications, and the marketplace. Secondary data can also be employed to acquire an initial understanding of the study problem.

Secondary data analysis reduces the time that would otherwise be spent gathering data and, especially in the case of quantitative data, produces larger and higher-quality databases than any individual researcher could collect on their own. Furthermore, social and economic change analysts believe that secondary data is crucial because it is impossible to conduct a fresh poll that fully captures past changes and developments.

Secondary data can be gathered from two types of research:

• Quantitative: Census, housing, social security, and voting statistics, as well as other associated datasets.
• Qualitative: semi-structured and structured interviews, transcripts of focus groups, field notes, observation logs, and other personal, research-related documentation.

Secondary data is classified according to its source, whether internal or external. Internal data, often called in-house data, is secondary information gathered within the business where research is conducted. External secondary data is collected from various sources.

Sources of Internal Data

Internal secondary data is typically a low-cost information source for companies performing research and an excellent place to start for existing operations. Sales and pricing data created internally might be used as a research source. This data is used to define the firm’s competitive position, evaluate a previous marketing strategy, or better understand the company’s best customers.

Internal data is mainly derived from three sources. They are as follows:

1. Reports on sales and marketing: These can include:

• The type of product/service purchased
• The type of end-user/industry segment
• The method of payment
• The product or product line
• The sales territory
• The salesperson
• The date of purchase
• The amount of purchase
• The price
• The application of the product
• End-user location

2. Financial and accounting records: These are frequently overlooked sources of internal secondary information, yet they can be extremely useful in identifying, clarifying, and forecasting specific problems. Accounting data can also be used to assess the success of various marketing tactics, such as direct marketing campaign revenues.

Using accounting and financial data presents some challenges. One is the timing issue; accounting statements are frequently inaccessible for several months. Another consideration is the structure of the documents themselves. Most businesses do not adequately configure their accounts to offer the types of answers to their required research queries. Account systems, for example, should record project/product expenses to determine the company’s most profitable (and least profitable) operations.

Companies should also explore developing financial-data-based performance measures. These can be industry standards or new ones designed to assess essential performance factors, allowing the firm to track and compare its performance over time. Sales per employee, sales per square foot, and expenses per employee are a few examples (salesperson, etc.).

3. Miscellaneous reports: These can comprise inventory reports, service calls, the number (qualifications and salary) of employees, and production and R&D reports. A company’s business plan and a log of client calls (complaints) can also be important sources of information.

Sources of External Data

Today, there is an abundance of statistical and scientific data available. Some examples are:

• Federal government;
• Provincial/state governments
• Statistical organisations
• Trade organisations
• General business periodicals
• Articles in magazines and newspapers
• Annual reports
• Scholarly publications
• Library resources
• Bibliographies on computers
• Services that are syndicated.

Secondary data has two significant advantages in market research: it saves time and money.

1. The secondary research phase can be completed quickly – often in two to three weeks. A skilled analyst can acquire significant amounts of meaningful secondary data in a few days.

2. If secondary data is accessible, the researcher merely needs to find the source and extract the necessary information.

3. In general, secondary research is less expensive than original research. The majority of secondary research data collection does not necessitate the utilisation of expensive, specialised, and highly trained individuals.

4. The creator of the material bears the cost of secondary research. There are also some drawbacks to adopting secondary data. These are some examples:

• Secondary information relevant to the research issue is either unavailable or insufficient.
• Some secondary data may be of dubious accuracy and dependability. Statistics in government publications and trade journals can often be deceptive.
• Data may be in a different format or units than the researcher requires.
• Because much secondary data is several years old, it may no longer reflect current market conditions. Trade journals and other periodicals frequently accept manuscripts six months before they are published in print, so the study could have been conducted months or even years ago.

3.2.5 Qualitative Research

Through the study of unstructured information—such as interview transcripts, e-mails, notes, feedback forms, images, and videos—qualitative research searches out the ‘why’ rather than the ‘how’ of an issue. It is not solely based on statistics or numbers, which are the purview of quantitative researchers.

Qualitative research is used to learn about people’s attitudes, behaviours, value systems, concerns, motivations, goals, culture, and lifestyles. It guides business decisions, policy development, communication, and research. Among the various formal methodologies employed are focus groups, in-depth interviews, content analysis, and semiotics, but qualitative research also includes examining any unstructured material, such as customer feedback forms, reports, or media clips.

Qualitative research is done to better understand how and why people feel the way they do. It is concerned with gathering detailed information by asking probing questions such as “Why do you say that?” Compared to quantitative initiatives with substantially larger samples, sample sizes in qualitative projects are typically much smaller. Depth interviews and group discussions are two prominent ways of gathering qualitative data.

As a result, qualitative research can be classified as a sort of scientific study. In general, scientific research is an examination that:

• Seeks solutions to a question;
• Employs a specified set of methods to answer the question in a systematic manner
• Gathers evidence
• Generates conclusions that were not predetermined
• Generates findings that are applicable beyond the study’s immediate bounds

Qualitative research shares these characteristics. Furthermore, it aims to comprehend a specific study problem or topic from the local community’s views. Qualitative research successfully gathers culturally relevant information on specific groups’ beliefs, attitudes, behaviours, and social contexts.

The ability of qualitative research to produce detailed textual descriptions of how people experience a given research subject is its strength. It provides information regarding an issue’s “human” side, i.e., individuals’ often contradictory behaviours, attitudes, views, feelings, and relationships. Qualitative approaches are also helpful in discovering intangible aspects such as societal norms, socioeconomic status, gender roles, race, and religion, whose involvement in the research issue may not be obvious. When combined with quantitative methodologies, qualitative research can assist us in interpreting and better understanding the complicated reality of a given situation and the implications of quantitative data. Although findings from qualitative data are frequently generalizable to people with similar characteristics to those in the study population, gaining a rich and complex understanding of a specific social context or phenomenon typically takes precedence over eliciting data that can be generalised to other geographical areas or populations. In this regard, qualitative research differs slightly from standard scientific study.

The three most frequent qualitative methodologies are participant observation, in-depth interviews, and focus groups. Each is explained in length in its own module and best suited for obtaining a particular data.

• Participant observation is suitable for gathering data on spontaneously occurring activities in their natural situations.
• In-depth interviews are ideal for gathering information about people’s backgrounds, perspectives, and experiences, mainly when discussing sensitive themes.
• Focus groups are influential in eliciting data on a group’s cultural norms and producing comprehensive overviews of problems relevant to the cultural groups or subgroups represented.

3.3 Descriptive Research Design

The name implies that it is essentially a research project to describe something. It can, for example, explain the features of a group, such as customers, organisations, marketplaces, and so on. A descriptive study reveals an “association between two factors,” such as income, shopping location, age and preferences.

Descriptive statistics provide information about the proportions of high- and low-income clients in a given territory. However, descriptive research cannot reveal a cause-and-effect link between the features of interest, which is a significant drawback.

A descriptive study necessitates a precise statement of the research’s “Who, What, When, Where, Why, and How.” Consider the case of a convenience store (Food World) looking to open a new location. “How do people come to patronise a new outlet?” the corporation wants to know. The following are some of the questions that must be answered before data collection for this descriptive study:

•  Who is considered a shopper accountable for the shop’s success, and whose demographic profile is required by the retailer?
•  What shopper attributes should be measured?
• Is it the shopper’s age, gender, income, or residential address?
• When should we measure?
• Should the measurement be taken while the shopper is buying or later?
•  Where are we going to measure the shoppers?
• Should we call them outside the stores shortly after they visit, or should we contact them at their home?
•  Why do you wish to quantify them?
• What is the function of measurement? Are there any fact-based solutions that will assist the retailer in increasing sales? Is the retailer hoping to forecast future sales based on the information gathered?
• The answers to some of the preceding questions will assist us in developing the hypothesis.
• How do you measure? Is it a ‘structured,’ ‘disguised,’ or ‘undisguised’ questionnaire?

3.3.1 When Should You Use a Descriptive Study?

1. To determine market parameters such as:

• Market size
• Consumer purchasing power
• Product consumption pattern
• To determine the product’s market share(s) and to monitor a brand’s performance.

2. To ascertain the relationship between two variables: advertising and sales.

3. To make a forecast. We may be interested in sales forecasts over the next three years to plan for the training of new sales reps.

4. How can we calculate the proportion of people in a given population who behave in a certain way?

For example, what percentage of the population in a specific geographical place would shop in a specific store?

At the descriptive research stage, hypotheses are investigated (to demonstrate the characteristics of the group).

Management problem	Research problem	Hypothesis
How should a new product be distributed?	Where do customers buy a similar product right now?	Upper-class buyers use ‘Shopper’s Stop’, and middle-class buyers buy from local department stores.
What will be the target segment?	What kind of people buy our product now?	Senior citizens buy our products. Young and married buy our competitor’s products.

3.3.2 Different Types of Descriptive Studies

Descriptive research is classified into two types:

• Longitudinal research
• Cross-sectional research

1. Longitudinal Study: These are studies in which an event or occurrence is repeatedly measured. They are sometimes referred to as ‘Time Series Studies. ‘ Through longitudinal research, the researcher learns how the market changes over time.

Panels are used in longitudinal research. Once formed, certain aspects will be included in the panel. Individuals, businesses, and dealers are examples of these factors. Throughout the period, the panel or sample remains consistent. There could be a few dropouts and additions. The panel’s sample members are being measured regularly. The study’s frequency could be monthly, quarterly, or whenever.

Assume that market research on ready-to-eat food is undertaken at two different intervals in time, T1 and T2, separated by four months. A sample of 2000 households is chosen and interviewed each of the preceding two times. The following are the brands that are most commonly used in households:.

Brands	At T1	At T2
Brand X	500(25%)	600(30%)
Brand Y	700(35%)	650(32.5%)
Brand Z	400(20%)	300(15%)
Brand M	200(10%)	250(12.5%)
All others	200(10%)	250(12.5%)
	200	100%

As can be observed from the comparison of periods T1 and T2, Brand X and Brand M have increased their market share. Brands Y and Z have lost market share while all other categories have increased. This demonstrates that Brands A and M have gained market share at the expense of Brands Y and Z.

Panels are classified into two types:

(a) True panel: This includes repeatedly measuring the same variables. Consider the perception of frozen peas or iced tea. Each panel member is scrutinised at a different time to conclude the subject above.

(b) Omnibus panel: A sample of elements is also selected and stored in the omnibus panel, but the information obtained from the members changes. At some point, the attitude of panel members “towards an advertisement” can be assessed. The same panel member may later be questioned about “product performance.”

Advantages of Panel Data

• We can determine the proportion of people who purchased our brand against those who did not. The brand switching matrix is used to compute this.
• The study also assists in identifying and targeting the group that requires promotional efforts.
• A large amount of data can be collected because panel membership is voluntary.
• The most significant advantage of panel data is its analytical nature.
• Panel data is more accurate than cross-sectional data because it lacks the inaccuracy of reporting previous behaviour. Errors in past behaviour occur due to time passing or amnesia.

Disadvantages of Panel Data

• The sample may not be representative. This is because panels are sometimes chosen for their convenience.
• The panel members who submit the data may be unwilling to continue serving on the panel. There may be dropouts, migration, and so on. Members who replace them may be significantly different from the original members.
• The remuneration of panel members may be unappealing, so some people may not want to serve on a panel.
• Panel members may occasionally appear disinterested and uncommitted.
• Prolonged participation in the panel may lead to respondents considering themselves experts and professionals. They may begin to reply as experts and consultants rather than as respondents. No member should be maintained for more than six months to avoid this.

2. Cross-sectional Study: One of the most essential types of descriptive research, cross-sectional study can be done in two ways:

Field research: This involves an in-depth investigation. A field study is an in-depth examination of a subject, such as the reaction of young men and women to a product.

Example: Indian men’s reaction to a branded ready-to-wear suit. Field research is conducted in real-world settings. Test marketing is a type of field study.

Field survey: The study has a large sample size. The most significant limitations of this survey are financial and timing constraints. Also, if the reply is cautious, he may answer the questions differently. Finally, conducting a field survey necessitates an extensive understanding of questionnaire design, sampling methodologies, etc.

Assume management believes that a geographical component is an essential factor in determining the consumption of a product, such as sales of woollen apparel in a specific place. The claim under consideration is that the urban population is more likely to use the product than the semi-urban population. A cross-sectional study can be used to test this concept. In terms of occupation and product use, measurements can be gathered from a representative sample of the population in both geographical locations. In the event of tabulation, the researcher can count the number of examples that fall into each of the following categories:

• Urban population using the product – Category I
• Semi-urban population utilising the product – Category II
• Urban population not using the product – Category III
• Semi-urban population not using the product – Category IV

In this instance, we should be aware that the sample data must support and test the hypothesis, i.e., the proportion of urbanities using the product must outnumber the proportion of semi-urban populations utilising the product.

3.3.3 Survey

A survey is a research approach in which data is acquired by asking respondents questions. One of the most essential measuring fields in applied social research is survey research. Survey research is a broad term that refers to any assessment process that involves asking respondents questions. A “survey” can range from a little paper-and-pencil feedback form to a lengthy one-on-one, in-depth interview.

Types of Survey 

Surveys are classified into two types: questionnaires and interviews. Questionnaires are typically paper-and-pencil devices that respondents fill out. The interviewer completes the interviews based on what the respondent says. It can sometimes be challenging to recognise the difference between a questionnaire and an interview. Some believe surveys always ask short, closed-ended questions, whereas interviews always ask broad, open-ended ones. However, questionnaires with open-ended questions are standard (albeit often shorter than interviews), while interviews frequently include a sequence of closed-ended questions.

In the last ten years, survey research has evolved tremendously. We offer automated telephone surveys and use random dialling methods. In public spaces, digital kiosks allow users to provide feedback. The focus group approach is a completely new type of group interview. Survey research is being increasingly linked with service delivery. On the desk in your hotel room, there is a survey. With your check, your waiter hands you a small customer satisfaction survey. Several days after your previous call to a computer company for technical assistance, you receive a call for a job interview. You are invited to complete a brief survey when you visit a website.

Selecting a Survey Method

Choosing the type of survey to employ is one of the most important considerations in many social research scenarios. You’ll notice that very few easy guidelines will make the selection for you — you must use your judgment to weigh the benefits and drawbacks of various survey styles. All I want to do here is suggest a few questions you might ask to help guide your selection.

Population Issues

The first set of concerns is related to the population and its accessibility.

1. Is it possible to count the population?

You have a complete list of the units that will be sampled for various populations. For others, compiling such a list is difficult or impossible. For example, there are complete lists of registered voters and those with valid driver’s licenses. However, no one keeps an up-to-date list of homeless people. If you are doing a study that involves input from homeless people, you will almost certainly need to go out and find the respondents yourself. Mail surveys and telephone interviews are out of the question in such cases.

2. Is the general public literate?

Questionnaires necessitate that your respondents be able to read. While this may appear to be a realistic assumption for many adult populations, new research indicates that adult illiteracy rates are frighteningly high. Even if your respondents can read to some extent, your questionnaire may contain complex or technical terminology. Some populations would be expected to be illiterate. Questionnaires would not be appropriate for young children.

3. Are there any language barriers?

We live in a multilingual world. Almost every civilization has people who speak a language other than the prevalent language. Some countries, such as Canada, have official multilingualism. Furthermore, our increasingly global economy necessitates research that transcends countries and linguistic groups. Are you able to create several versions of your questionnaire? Can you know in advance the language your respondent speaks when sending mail instruments, or do you send various translations of your instrument? Can you be sure that your instrument’s essential implications are not culturally specific? Could some key details be missed in the translation of your questions?

4. Will the general public cooperate?

Researchers working on immigration issues face a difficult methodological challenge. They frequently need to speak with undocumented immigrants or others who may know about others who are. Why should we expect those respondents to work with us? Although the researcher may mean no harm, the responders face serious legal consequences if the information they reveal falls into the hands of the authorities. The same can be said of any target group involved in unlawful or unpopular actions.

5. What are the geographical constraints?

Is your community of interest too spread geographically to examine feasibly with a personal interview? You might be able to send a mail instrument to a countrywide sample. You might be able to interview them over the phone. However, if respondents are widely separated, conducting research requiring interviewers to visit them will most likely be more challenging.

Sampling Issues

The sample is the group with whom you will have to communicate. When conducting survey research, various essential sample concerns must be considered.

1. What information is available?

What details do you have on your sample? Do you know where they are now? What are their current phone numbers? Is your contact information up-to-date?

2. Can responders be located?

Can you find your respondents? Some people are extremely busy, some travel a lot, and some work the night shift. Even if you have a precise phone number or location, you may be unable to reach or contact your sample.

3. What is the respondent’s name?

In your study, who is the respondent? Assume you choose a random sample of households in a small city. A household does not qualify as a respondent. Do you wish to interview a specific person? Do you only want to speak with the “head of household” (and how is that person defined)? Are you willing to speak with any member of the family? Do you say you’ll talk to the first adult member of the family who opens the door? What if that person refuses to be interviewed, but someone else in the house is? How do you handle multi-family dwellings? Similar issues emerge when sampling groups, agencies, or businesses. Can you poll any of the organization’s members? Or are you simply interested in speaking with the Director of Human Resources? What if the person you want to interview refuses or cannot participate? Do you make use of another organisation member?

4. Can the entire population be sampled?

You may not be able to sample every member of the population if you have an incomplete list of the population (i.e., sampling frame). Lists of numerous groupings are exceedingly difficult to keep up with. People migrate or have their names changed. Even though they are on your sampling frame list, you may be unable to access them. It’s also possible they aren’t even on the list.

5. Is it conceivable that response rates will be an issue?

You must still deal with response rates even if you solve the other population and sample issues. Some of the people in your sample will refuse to answer. Others have good intentions but can’t find time to complete your questionnaire before the deadline. Others misplace the instrument or forget about the interview appointment. Low response rates are among the most challenging challenges to solve in survey research. They have the potential to derail an otherwise well-planned survey effort.

Question Issues

The type of survey you choose may be determined by the nature of the questions you wish to ask respondents.

1. What kinds of queries are permitted?

Will you be asking personal questions? Will you require a lot of detail in the responses? Can you predict the most common or important types of reactions and formulate appropriate closed-ended questions?

2. How difficult will the questions be?

Sometimes, you have to deal with a complicated subject or topic. The questions you want to ask will be divided into sections. You may need to branch to sub-questions.

3. Will there be any screening questions?

A screening question may be required to assess whether the reply is qualified to answer your issue of interest. For example, you wouldn’t ask someone their thoughts on a specific computer software without first “screening” them to see if they’ve used the programme before. You may sometimes need to screen on many variables (e.g., age, gender, experience). The more intricate the screening, the less probable that paper-and-pencil devices will suffice without confusing the respondent.

4. Is it possible to adjust the question sequence?

Is your survey one in which you can plan a suitable sequence of questions ahead of time? Or are you conducting an initial exploratory study in which you may need to ask several follow-up questions you cannot foresee?

5. Will you ask extensive questions?

If your issue is intricate, you may need to provide the respondent with some background information before asking a question. In a phone interview, can you reasonably expect your respondent to sit still long enough to ask your question?

6. Are long response scales going to be used?

Suppose you question them about the various computer equipment they use. In that case, you may need an extensive response list (CD-ROM drive, floppy drive, mouse, touchpad, modem, network connection, external speakers, etc.). Asking about each of them in a brief phone interview may be challenging.

Content Issues

The substance of your study may also provide difficulties for the various survey forms you may employ.

1. Can the respondents be expected to be aware of the problem?

If the respondents do not keep up with the news (for example, by reading the newspaper, watching television news, or conversing with others), they may be unaware of the news problem you wish to ask them about. Or, if you want to study family finances and talk to a spouse who does not pay the bills regularly, they may not know how to answer your inquiries.

2. Will the respondent be required to consult records?

Even if the respondent knows what you’re asking, you may need to let them examine their records to receive an appropriate answer. For example, if you inquire how much money they spent on food the previous month, they may need to consult their personal check and credit card records. In this scenario, you don’t want to be involved in an interview where they have to check stuff up while you wait (they wouldn’t be happy about that).

Bias Issues

People bring their own biases and prejudices to the study project. However, particular survey methodologies may make these biases less of an issue.

1. Is it possible to avoid social desirability?

Respondents wish to “appear good” in the eyes of others. Nobody wants to appear as though they don’t know the answer. We don’t want to say anything that will make us look bad. If you ask individuals about information that could place them in this situation, they may not tell you the truth or may “spin” their response to make themselves look better. This may be more of an issue in an interview where they speak face-to-face or over the phone with a live interviewer.

2. Is it possible to control interviewer distortion and subversion?

Interviewers can also distort an interview. They may refrain from making inquiries that make them feel uneasy. They may not pay close attention to responses on topics they have strong feelings about. They may believe that, based on their previous responses, they already know what the respondent would say to a question, even if this is untrue.

 3. Is it possible to avoid fake respondents?

It may be challenging to determine who responded to a mail survey. Did the home’s head complete the study or someone else? Did the CEO honestly respond, or did he delegate the duty to a subordinate? Is the person on the other end of the phone who they say they are? Personal interviews, however, give you a good idea of who you’re speaking with. This may not be the case in postal surveys or phone interviews.

Administrative Issues

Last but not least, you must assess the viability of the survey method for your research.

• Costs: Regarding survey types, cost is generally the most critical deciding factor. You may like to do personal interviews but cannot justify the high cost of training and compensating interviewers. You may prefer to send a large mailing but cannot afford the postage.
• Facilities: Do you have the resources (or do you have access to them) to process and manage your research? Do you have well-equipped phone surveying facilities for phone interviews? Do you have a suitable and accessible room to host focus groups? Do you have the necessary equipment to record and transcribe responses?
• Time: Some types of surveys require more time than others. Do you need immediate responses (as in an overnight public opinion poll)? Have you scheduled enough time for your study to send out mail surveys and follow-up reminders and receive responses via mail? Have you allotted enough time to do enough personal interviews to validate that strategy?
• Personnel: Different types of surveys require different personnel requirements. Interviews require motivated and well-trained interviewers. People trained in group facilitation are needed to deliver group surveys. Some studies may be in a technical field that requires some skill on the interviewer’s part.

There are numerous factors to consider when deciding which survey to use in your research. There is often no clear and straightforward approach to making this decision. There may not be a single technique that is unquestionably the best. You may have to decide between benefits and drawbacks, requiring some judgment. Two skilled researchers may choose different survey approaches for the same problem or issue. However, if you select a method that isn’t acceptable or doesn’t fit the context, you could do a study before developing the instruments or questions.

3.3.4 Observation Studies

An observational study derives conclusions regarding the likely effect of a treatment on subjects when the assignment of subjects to a treated group versus a control group is outside the investigator’s control. This contrasts with controlled experiments, such as randomised controlled trials, in which each patient is randomly allocated to a treated or a control group before treatment begins.

Natural experiments or quasi-experiments, are terms used to describe observational investigations. These variances in language reflect differences in emphasis. Still, a common feature is that the early stages of planning or designing an observational study seek to replicate some of the qualities of an experiment as closely as feasible.

3.4 Difference between Exploratory Research and Descriptive Research

Aspect	Exploratory Research	Descriptive Research
Purpose	To explore a topic with limited understanding	To describe the characteristics of a
	And generate hypotheses.	Population or phenomenon.
Focus	Generating ideas, uncovering trends,	Providing an accurate portrayal of existing
	it is understanding the nature of a phenomenon.	Conditions, behaviours, or attitudes.
Stage in Research	Often conducted at the initial stages of a	Typically conducted after exploratory
Project	research project.	research, or when existing data is available
Methods	Literature reviews, interviews, focus groups,	Surveys, observational studies, statistical
	Observations, case studies, etc.	Analysis of existing data, etc.
Outcome	Insights, hypotheses, and groundwork for	Quantifiable data, descriptive statistics,
	further investigation.	Factual descriptions of the subject.

3.5 Causal Research Design

Causal Research refers to studies that test hypotheses to explain the nature of specific relationships, identify differences between groups, or establish the independence of two or more components in a situation. It is a study design in which the primary focus is on establishing a cause-and-effect link. The research is used to determine the impact of a particular modification on current standards. It helps market researchers forecast hypothetical situations on which a company’s business plan might be based.

For example, if a clothing company sells blue denim jeans, causal research can assess the impact of altering the product design to white.

Following the completion of the investigation, firm executives will be able to determine whether altering the jeans’ colour to white would be profitable.

To summarise, causal research determines how current activities will affect a business in the future. However, it should be noted that not all causal research hypotheses can be investigated. There are several reasons for this, one of which is that genuine random assignment is not always achievable. The three most important reasons why you can’t test everything are:

• Technology, or the inability of today’s technology to do some tasks, such as gender assignment.
• Ethics, because we can’t randomly assign some people to receive a virus to test its effects, or that some participants must act as slaves and others as masters to test a hypothesis, and
• Resources: A study will not be conducted if a researcher lacks the funds or the necessary equipment.

The study of cause-and-effect linkages between two or more variables is known as causal design.

In Methods in Social Research, William J. Goode and Paul K. Hatt describe cause and effect relationship as “where two or more occurrences of a given phenomenon have one and only one condition in common, that condition may be regarded as the cause and effect of that phenomenon.”

The collection of causes developed to forecast their effects might be deterministic or probabilistic. The deterministic cause is necessary and sufficient for another event to occur. While probability is the most important cause, it is not the only one accountable for stimulating another event’s occurrence.

Causal research must establish which variable may be causing a given behaviour, i.e., whether there is a cause-and-effect link between variables. This type of research is highly complex, and the researcher can never be positive that no other factors impact the causal relationship, especially when dealing with people’s opinions and motivations. There are frequently far deeper psychological issues that even the respondent is unaware of.

All variables allowed for the most accurate casual remarks are not frequently present in marketing decision-making. Yet, marketing managers will nevertheless make causal inferences in these scenarios because they would like to be able to make casual statements about the consequences of their conduct.

For example, a new advertising campaign designed by a corporation resulted in a % rise in sales, or a sales discount strategy implemented by a company resulted in a percentage increase in sales. Marketing managers are making a casual message in each of these situations.

On the other hand, the scientific idea of casualty is sophisticated and differs significantly from the one held by the average person on the street. According to common sense, a single occurrence (the cause) always results in another event (the consequence) occurring. We realise in science that an occurrence has several determining circumstances or causes that function in concert to make the event likely. It is worth noting that the common-sense concept of casualty holds that the effect always follows the cause. This is deterministic causality instead of scientific causation, which describes the result as probable. Probabilistic causality is the word for this. According to scientific theory, we can only infer casualty and never truly establish it. That is, the possibility of making an inaccurate inference is constantly considered. The realm of marketing corresponds to the scientific definition of casualty. Various causes drive marketing impacts, and we can only infer a causal association. The following conditions must be met before we can make a casual inference:

• The time and order of occurrence of variables.
• Concurrent variation
• Exclusion of any other potential causative factors.

Causal research designs establish a more solid foundation for the presence of a causal relationship between variables. The researcher controls one or more extraneous variables’ influence on the dependent variable. When it is impossible to regulate the impact of an extraneous variable on the dependent variable, the variable is called chaotic.

What is a Synopsis, and How Do I Write One?

A synopsis is an abstract form of research that highlights the research technique and is offered as a guide for evaluating its potentiality. It may be described in one word as a condensed version of the final report. The structure of a summary varies and is also determined by the guides’ preferences. However, for our purposes, a typical structure can be framed as follows:

• Defining the Problem: When describing the problem of the research objective, important terminology, basic background information, restrictions of the study, and sequence of presentation should be briefly described.
• Review of Existing Literature: Under this heading, the researcher should summarise various viewpoints on the subject matter found in books and journals and the approach to be used while writing.
• Conceptual Framework and Methodology: The researcher should first state the hypothesis under this heading. Debates on the research methodology employed and the relationship between the hypothesis and the purpose of the study should occur, as should discussions on the sources and methods of acquiring data. In this section, the researcher should also mention any constraints of technique and any natural crises that the research is sure to undergo due to such apparent limitations.
• Data Analysis: Data analysis entails evaluating hypotheses based on data collected and arriving at crucial conclusions.
• General Conclusions: The researcher should clarify the objectives in the general conclusions. The conclusion should include hypothesis acceptance or rejection, stated goals, proposed areas for further research, and a concluding discussion of the study’s probable consequences for a model, group, theory, and discipline.

Finally, the researcher should discuss the bibliographies and appendices. The above approach is based on a common global framework for preparing a summary. However, in our country, the style and structure of synopsis vary depending on the research goal, and it is pretty common to find that the research guide uses his or her discretion in synopsis writing rather than adhering to some approved international norms. A typical format for preparing a summary that is often used in management and commerce research in India is as follows:

• Introduction: This comprises an explanation of the problem and a historical overview.
• Study Objective: It outlines the research objective and distinguishes it from previous research in the connected sector.
• Literature Review: This section covers, among other things, several sources from which the required abstract is derived.
• Methodology: It is designed to outline the study sequences and the methods and procedures for conducting the survey and compiling the data.
• Hypothesis: A formal declaration relating to the study problem that must be tested based on the researchers’ results.
• Model: It underpins the kind and structure of the model that the researcher will construct based on survey findings.

3.6 Experimentation

Experimentation Causative research is another name for research. A descriptive study will reveal if there is a relationship between the variables, but it will not demonstrate a cause-and-effect relationship. For example, data obtained may indicate that the number of persons who own a car and their income has increased. Despite this, we cannot argue that “the growth in the number of cars is related to an increase in income.” Perhaps improved road conditions or an increased number of institutions offering auto loans have contributed to a rise in car ownership.

The researcher must experiment to determine the causal relationship between the variables.

Example:

1. Which of the following print advertisements is more effective? Is it the first, second, or third page?

2. “Which is more effective” among various promotional measures, such as advertising and personal selling? Can we enhance our product’s sales by gaining more shelf space? What exactly is experimentation? The research method manipulates one or more variables to demonstrate the cause-and-effect relationship. Experimentation is used to determine the influence of one thing on another. The many components of the experiment are described here.

Units for Testing

These are the units on which the experiment is performed. It is carried out with one or more independent variables controlled by a person to determine their effect on a dependent variable.

Explanatory Variable

These are the variables whose influence the researcher wants to investigate. Explanatory variables may include, for example, advertising, pricing, packaging, and so on.

Dependent Variable

This variable is being researched, including sales, consumer attitudes, brand loyalty, etc.

Assume a particular colour TV manufacturer cuts the price of the TV by 20%. Assume that his savings are passed on to the consumer and that sales will increase by 15% next year. During the festival season, leading television corporations do trials of this nature.

The causal investigation investigates whether lowering the price leads to increased sales.

Variables That Aren’t Required

These are also referred to as blocking variables. Extraneous variables influence the outcome of the experiments.

1. Assume a toffee manufacturing company is attempting to measure buyer response to two different styles of packaging in two separate areas. According to the manufacturer, all other aspects must be the same for each customer group. If the manufacturer allows the extraneous variable, “Price,” to vary between the two buyer groups, he will not be able to determine which package the buyers prefer. Price changes are an unimportant factor in this case.

Two approaches can be taken when it comes to unnecessary variables. Extraneous variables can be managed physically. The preceding example is the price.

Extraneous variables in the second category may be outside the researcher’s control. In this instance, we say the experiment has been confused, meaning no inferences can be drawn. A variable like this is referred to as a “confounding variable.”

2. The company launches a product in two cities and wants to know how its advertising is affecting sales. Simultaneously, a competitor’s goods in one of the cities are unavailable during this time due to a factory strike. Now, the researcher cannot conclude that the sales of their product in that city have increased due to advertising. As a result, this experiment is flawed. The confounding variable in this example is strike.

Extraneous Variable Types

The many types are as follows:

• History
• Maturation
• Testing
• Instrument Variation
• Selection Bias
• Experimental mortality

1. History: History refers to occurrences that occur outside of the experiment yet happen simultaneously as the experiment is being carried out. This could have an impact on the outcome. Assume that a manufacturer reduces the price of a product by 20% and watches sales in the coming weeks. The study’s goal is to determine the effect of pricing on sales. Meanwhile, sales will not rise if the product’s manufacturing falls due to a lack of raw materials; sales will not rise. As a result, we cannot conclude that the price cut did not affect sales because external events occurred during the period, and we had no control over the occurrence. The only thing that can be done is to identify the occurrence.

2. Maturity: Maturity is analogous to history. Maturation refers to changes within the test units rather than arising from experimentation. Maturation occurs as a result of the passage of time. Maturation is the result of humans getting older. People could be utilising a product. They may stop using the product or switch to a different product.

• Pepsi is usually consumed when one is young. As one age, the customer may eat or avoid Diet Pepsi.
• Assume that a sales training programme is being run, and the company wishes to assess its effectiveness. If the corporation discovers that sales have increased, it is possible that this is not attributable to the training programme. It could be because salespeople have gained more experience and better understand their customers. Increased sales may result from improved communication between the salesperson and the consumer.

The maturation effect is not restricted to the test unit, which is entirely human. Organizations also evolve; dealers expand, become more successful, diversify, and so on.

3. Testing: When the same respondents are measured more than once, a pre-testing effect occurs. Responses given later in the process will directly impact responses offered earlier.

Example: Consider a respondent given a preliminary questionnaire designed to assess brand awareness. Following his exposure, if a second questionnaire similar to the initial questionnaire is presented to the respondent, he will reply quite differently due to the respondent’s familiarity with the previous questionnaire.

Internal validity is a problem with pre-tests. An example will help you understand this. Assume that a respondent’s opinion is measured before and after exposure to a Hyundai vehicle commercial with Shahrukh Khan as the brand ambassador. When the responder responds for the second time, he may recall how he assessed Hyundai during the first measurement. He may provide the same rating to demonstrate consistency. In that instance, the difference between the two measures tells us nothing about the actual impact.

Alternatively, some respondents may provide a different rating during the second measurement. This may not be because the respondents’ opinions of Hyundai and the brand ambassador have shifted. He has provided a different rating because he does not want to be branded as someone who has not changed his mind on the commercial.

Internal validity suffers in both of the preceding examples.

4. Instrument Variation: When human respondents are involved, the instrument variation effect threatens internal validity. For example, a vacuum cleaner may be left behind for the customer to use for two weeks. The responders are given a questionnaire to complete after two weeks. The response may differ significantly from the respondent’s before the product trial. This could be due to two factors:

• A few of the questions have been revised.
• The interviewer for pre-testing and post-testing differs.

The equipment used to measure in experiments will impact the measurement. In addition, where there are multiple interviewers, the results may vary due to the use of instruments. As a result, ensuring that all interviewers ask the same questions in the same tone and build the same rapport is tough. Because each interviewer performs the interview differently, responses may change.

5. Selection Bias: Selection bias arises when two groups chosen for an experiment are not identical. When the two groups are posed different questions, their responses will differ. This error will occur if many groups participate. There are two advertisements for “Ready to eat meals,” A and B. The goal is to determine the effectiveness of the two adverts. Assume that the respondents exposed to ‘A’ are the product’s dominant users. Assume that 50% of individuals who viewed ‘Advertisement A’ purchased the product, but only 10% of those who saw ‘Advertisement B’ purchased the product. The above does not imply that advertisement ‘A’ is more effective than advertisement ‘B.’ The main difference between the groups may be related to food preference habits; internal validity may deteriorate somewhat in this scenario.

6. Experimental Mortality: Some original group members may leave, and some new members may join the existing group. This is due to the possibility of some members migrating to another geographical area. As a result of this change in membership, the group’s composition will change.

Example: Assume a vacuum cleaner maker wishes to release a new model. He interviews a hundred people who are still using the earlier version. These 100 respondents ranked the current vacuum cleaner on a 10-point scale (1 for lowest and 10 for highest). The average rating of the respondents is 7.

The latest version is now shown to the same hundred respondents, and the equipment is given to them for two months. After two months, only 80 participants responded because the remaining 20 refused to answer. If the mean score of the 80 respondents is eight on a scale of 10, Can we conclude that the new vacuum cleaner is superior?

The answer to the preceding question is determined by the composition of the 20 responders who dropped out. If the 20 respondents who did not complete the survey had an adverse reaction to the product, the mean score would not have been 8. It could be less than 7. The difference in mean rating does not provide an accurate picture. It does not imply that the new product is superior to the old product.

One might wonder why we don’t just leave the 20 initial group respondents out, calculate the mean rating of the remaining 80, and compare them. However, this strategy will not address the mortality effect, which will occur in any experiment, whether or not humans are involved.

Concomitant Variable

The extent to which a cause “X” and an effect “Y” vary simultaneously in a predictable manner is referred to as the concomitant variable.

Example:

1. The electric vehicle is novel in India. People may or may not be enthusiastic about electric cars. Assume the corporation has launched a new advertising campaign “to modify people’s attitudes toward this car” to increase vehicle sales. Assume that after testing the results of this campaign, the corporation discovers that both goals were met, i.e., people’s attitudes toward electric vehicles have improved, and sales have increased. Then, we can argue that there is a relationship between attitude and sales. Both variables are pointing in the same direction.

2. Assume an educational institution proposes a new elective that it claims is job-related. The college’s administration publicises this course in a major newspaper. They want to know how students feel about this course and how many will enrol. If, after testing, it is discovered that the perception of this course is positive and that most respondents are willing to enrol, we can conclude that there is a contemporary variation between perception and enrolment. Both variables are pointing in the same direction.

3.6.1 Experimental Design

The following are the various experimental designs:

• Only after design
• Factorial design
• Before-and-after design
• Latin square pattern
• Design after the fact

After only Design

This design measures the dependent variable after the test units are exposed to the experimental variable. The following example can help you understand this.

Assume M/s Hindustan Lever Ltd. wants to test the “Impact of free samples on the sale of toilet soaps.” A little sample of toilet soap is mailed to a pre-selected group of clients in a given area. After one month, a voucher for 25 paise off one cake of soap is mailed to each consumer who receives a free sample. A similar amount of these coupons are sent to people in other communities. The coupons are coded to maintain track of the quantity of coupons redeemed from each location. Assume 400 tickets from the experimental group were redeemed and 250 coupons from the control group. The 150-point discrepancy is thought to be the result of the free samples. Only after experimenting can a conclusion be formed using this method.

Before After Design

Measurements are taken both before and after using this procedure.

Example: Assume that an experiment is carried out to test an advertisement targeted at lowering alcoholism.

Before being exposed to the advertisement, the participants’ attitudes and perceptions toward liquor consumption are assessed. The group is shown an ad that informs them of the consequences, and their attitudes are measured again after a few days. The difference, if there is one, demonstrates the effectiveness of advertising.

The validity of the above “Before-after” example is jeopardised because of the following:

• Before-measure effect: It informs respondents that they are being researched. Respondents may change their behaviour after discussing the issues with friends and family.
• Instrumentation effect: This can be caused by using two distinct instruments, one before and one after, or by changing the interviewers before and after.

Factorial Design

A factorial design can be used to evaluate two or more variables simultaneously. It aids in determining each variable’s influence and measuring the interactive effect of numerous variables.

Example: A department shop wants to investigate the impact of lowering a product’s price. Given that, there is also promotion (POP) taking place in the stores (a) near the entrance (b) at the usual location, all at the same time. Assume there are two pricing levels, A1 (normal price) and A2 (discount price). There are three kinds of POP: B1, B2, and B3. There are 3 x 2 = 6 possible combinations. B1A1, B1A2, B2A1, B2A2, B3A1, B3A2 are the available combinations. The researcher is interested in which of these combinations is best suited. Assume the business has 60 department stores separated into ten-store clusters. Now, assign the combination above to each of these ten stores at random as follows:

Combinations	Sales
B1A1	S1
B1A2	S2
B2A1	S3
B2A2	S4
B3A1	S5
B3A2	S6

 

S₁ to S₆ reflect the sales generated by each variable. The information acquired will provide information on product sales based on two independent variables.

The following two questions will be addressed:

• Is the discounted price more cost-effective than the standard price?
• Is the entry display more effective than the display in the regular location? The study will also reveal the interaction effect of the two variables.

The following is the outcome of the sales experiment:

• Price reduction with entrance display.
• Price reduction with display in the standard location.
• There will be no display, and the standard pricing will apply.
• Put up a sign at the door with the standard price.

Latin Square Design

The researcher selects three different shelf configurations from three other retailers. He’d like to look at the sales generated at each store at other times of the year. The researcher must ensure that each retailer only uses one shelf arrangement.

Only one variable is investigated in the Latin square design. As an example of Latin square design, suppose a supermarket chain is curious about the impact of in-store promotion on sales. Assume the following three promotions are being considered:

• No advancement
• Complimentary sample with a demonstration
• Display of windows

Which of the three will be most effective? The size of the stores and the period may impact the outcome. If we select three stores and three time periods, the number of possible combinations is 33 = 9. The setup is as follows:

Time period	Store
	1	2	3
1	B	C	A
2	A	B	C
3	A	B	C

Latin Square is concerned with the efficiency of various types of marketing on sales.

Ex-post Facto Design

This is an alternative to “after-only design.” In this design, the groups, such as experiment and control, are only recognised after they have been subjected to the experiment.

Assume a magazine publisher wants to discover the effect of advertising on knitting in the ‘Women’s Era.’ Magazine subscribers are asked if they have seen this advertisement about “knitting.” Those who have read and those who have not read are asked about the product’s pricing, design, etc. The disparity reflects the effectiveness of the advertisement. In this design, the experimental group is assigned to receive the therapy rather than choosing to be exposed to it.

REVIEW QUESTIONS:

1. Can every causal research hypothesis be examined? Why or why not?

2. For each scenario below, determine whether the research should be exploratory, descriptive, or causal, and explain why:

(a) Investigating the relationship between promotion and sales.

(b) Understanding consumer reactions to new, economical detergents.

(c) Identifying the demographic profile of a shopping mall’s target market.

(d) Estimating the sales potential for ready-to-eat food in northeastern India.

3. What are the advantages and disadvantages of panel data in analysis?

4. Why does Latin Square Design typically test only one variable?

5. Are there benefits of factorial design over before-after design? Support your answer.

6. Should researchers include bibliographies and appendices? Why or why not?

7. Demonstrate the advantages of experience surveys with examples.

8. Why is exploratory research commonly used in the initial stages of research?

9. What type of research would you employ to generate new product ideas, and why?

10. Which research approach would you utilize to determine market characteristics?