What is Research Design It is a framework or blueprint for conducting the marketing research project.

. It specifies the precise details of the procedures necessary for obtaining the required information. Research Design Components Design of the research- exploratory, descriptive or causal. Define information needed Define measurement and scaling procedures Construct and pretest the questionnaire (interview form) Specify sampling process and sample size Develop a data analysis plan

Classification of Marketing Research Design (1/2)

Research Design

Descriptive Research

Exploratory

Conclusive

CrossSectional Design Causal

Longitudinal Design

Descriptive

Single Cross Sectional Design

Multiple Cross Sectional

Descriptive research :is characterized by a clear statement of the problem, specific hypotheses, and detailed information needs.

The Six Ws of Descriptive Research 1. Who: who should be considered? 2. Where: where should the respondents be contacted to obtain the required information? 3. When: when should the information be obtained from the respondents? 4. What: what information should be obtained from the respondents? 5. Why: why are we obtaining information from the respondents? 6. Way: the way in which we are going to obtain information from the respondents.

Cross Sectional Designs Cross-sectional designs involve the collection of information from a sample population at a single point in time Single Cross Sectional Design Only one sample of respondents is drawn from the target population and information is obtained from the sample only once. Multiple Cross Sectional Design There are two or more sample of respondents and information from each sample is obtained only once. Cohort Analysis Causal Research: To understand which variables are the cause (independent variables) and which variables are the effect (Dependent variables) of a phenomenon. To determine the nature of the relationship between the causal variables and the effect to be predicted Potential Sources of Error Random sampling error and non-sampling error are the two main sources of error. Random sampling error is the error due to the particular sample selected being an imperfect representation of the population of interest Random Sampling Error It is defined as the variation between the true mean value for the sample and the true mean value of the population. It is usually negligible provided a sample of sufficient size from the relevant population has been taken.

Non-sampling Error No-sampling errors occur due to a variety of reasons including errors in the problem definition, approach, scales, questionnaire design, interviewing methods, and data preparation and analysis. THE MARKETING RESEARCH PROCESS

MARKET RESEARCH

Secondary Research

Internal Sources Company Accounts Internal Reports and Analysis Stock Analysis Retail data - loyalty cards, till data, etc. External Sources Government Statistics (ONS) EU - Euro Stat Trade publications Commercial Data - Gallup, Mintel, etc. Household Expenditure Survey Magazine surveys Other firms research Research documents publications, journals, etc. Sampling Methods

Sampling Methods: Random Samples equal chance of anyone being picked May select those not in the target group indiscriminate Sample sizes may need to be large to be representative Can be very expensive Stratified or Segment Random Sampling Samples on the basis of a representative strata or segment Still random but more focussed May give more relevant information May be more cost effective Quota Sampling Again by segment Not randomly selected Specific number on each segment are interviewed, etc. May not be fully representative Cheaper method Cluster Sampling Primarily based on geographical areas or clusters that can be seen as being representative of the whole population Multi-Stage Sampling Sample selected from multi-stage sub-groups

Snowball Sampling Samples developed from contacts of existing customers word of mouth type approach! Primary Research

Primary Research First hand information Expensive to collect, analyse and evaluate Can be highly focussed and relevant Care needs to be taken with the approach and methodology to ensure accuracy Types of question closed limited information gained; open useful information but difficult to analyse Quantitative and Qualitative Information: Quantitative based on numbers 56% of 18 year olds drink alcohol at least four times a week - doesnt tell you why, when, how Qualitative more detail tells you why, when and how!

Purpose:

Advantages of Market Research Helps focus attention on objectives Aids forecasting, planning and strategic development May help to reduce risk of new product development Communicates image, vision, etc. Globalisation makes market information valuable (HSBC adverts!!) Disadvantages of Market Research Information only as good as the methodology used Can be inaccurate or unreliable Results may not be what the business wants to hear! May stifle initiative and gut feeling Always a problem that we may never know enough to be sure!

MEASUREMENT & SCALING: Measurement is the assignment of numbers or other symbols to characteristics of objects according to certain pre-specified rules. Scaling is an extension of measurement where it involves the generation of a continuum upon which measured objects are located

Level of measurement of a scale There are four basic characteristics: description, order, distance and origin, and together they define the level of measurement of a scale. The level of measurement denotes what properties of an object the scale is measuring or not measuring. Description By description we mean the unique labels or descriptors that are used to designate each value of the scale, e.g., 1. Female, 2. Male. Order By order we mean the relative sizes or positions of the descriptors. There are no absolute values associated with order, only relative values. Order is denoted by descriptors such as greater than, less than, and equal to. Distance The characteristic of distance means that absolute differences between the scale descriptors are known and may be expressed in units. E.g.Number of persons living in your household ___________ Origin The origin characteristic means that the scale has a unique or fixed beginning or true zero point. Thus, an exact measurement of income by a scale such as: What is the annual income of our household before taxes? Rs. ______________has a fixed origin or a true a zero point. ((( Description, order, distance, and origin represent successively higher-level characteristics, with origin being the highest-level characteristic. ))) Primary scales of measurement: Nominal scale. This is used only as a labeling scheme where numbers serve only as labels or tags for identifying and classifying objects. The numbers in a nominal scale do not reflect the amount of a characteristic possessed by the objects, rather they are used only for identification. For example, numbers on baseball players uniforms, street names, or social security numbers. Ordinal scale This is a ranking scale in which numbers are assigned to objects to indicate the relative extent to which some characteristic is possessed. It is then possible to determine whether an object has more or less of a characteristic than some other object. For example, rankings of teams Interval scale Numbers are used to rank objects such that numerically equal distances on the scale represent equal distances in the characteristic being

measured. Examples include time and temperature. The location of the Zero point is not fixed Ratio scale. This is used to identify or classify objects, rank order the objects, and compare intervals and differences. For example, height, age, and income.

two broad scaling measures: Comparative Noncomparative Comparative -A direct comparison of stimulus objects is elicited. Thus, two brands may be compared along a dimension such as quality. Must be interpreted in relative terms and have only ordinal and rank order properties Noncomparative scalesthe respondent provides whatever standard seems appropriate to him/her, thus, only one object is evaluated at a time. In this case, one brand is rated on a scale independent of other brands. comparative scaling techniques: Paired comparison scaling. Here a respondent is presented with two objects at a time and asked to select one object in the pair according to some criterion. The data obtained is ordinal in nature. This is frequently used in marketing when comparisons of products or brands are being made. Rank order scaling. Respondents are presented with several objects simultaneously and asked to order or rank them according to some criterion. This is commonly used to measure preferences for brands as well as the importance of attributes. Rank order scaling. Respondents are presented with several objects simultaneously and asked to order or rank them according to some criterion. This is commonly used to measure preferences for brands as well as the importance of attributes. Q-sort scaling. This technique uses a rank order procedure in which objects are sorted into piles based on similarity with respect to some criterion.

Noncomparative scaling as: The type of scaling which does not compare the object against another object or some standard. Rather, the rater uses whatever standard seems most appropriate to him or her. Continuous rating scale: The respondents rate the objects by placing a mark at the appropriate position on a line that runs from one extreme of the criterion variable to the other. The form of the continuous scale varies considerably depending on the imagination of the researcher. Their use in marketing has been limited because they are not as reliable as itemized scales, the scoring process is cumbersome, and they provide little additional information. Itemized rating scale: The respondents are provided with scales having numbers and/or brief descriptions associated with each category. The respondents are required to select one of the specified categories that best describes the object being rated. Several types of itemized rating scales exist: Likert scale: The respondents are required to indicate a degree of agreement or disagreement with each of a series of statements related to the stimulus objects. The Likert scale is often used in marketing. It is easy to construct and administer, it is easy for respondents to complete, and it is suitable for mail, telephone, and personal surveys. Semantic differential scale: This is a seven-point rating scale with end points associated with bipolar labels that have semantic meaning. Respondents are required to rate objects on a number of itemized, seven-point rating scales bounded at each end by one of two bipolar adjectives. This scale is popular in marketing and has been used in image studies, promotion strategy, and new product development studies. Stapel scale: This is a unipolar rating scale with ten categories numbered from 5 to +5, without a neutral point. Respondents are asked to indicate how accurately or inaccurately each term describes the object by selecting an appropriate numerical response category. Though easier to construct than the Semantic Differential, while giving the same results, the Stapel scale has not been widely applied in marketing. major decisions involved in constructing itemized rating scales: Number of scale categories to use Balanced versus unbalanced scale Odd or even number of categories

 Forced versus non-forced scales Nature and degree of verbal description Physical form of the scale the criteria used to evaluate a multiitem scale: Measurement accuracy Reliability Validity Generalizability RESEARCH REPORT: A research report is: a written document or oral presentation based on a written document that communicates the purpose, scope, objective(s), hypotheses, methodology, findings, limitations and finally, recommendations of a research project to others. The last stage of a marketing research process.; It is more than a summary of findings; rather it is a record of the research process. The researcher has to convince the client [and others who may read the report] that the research findings can be acted on for their own benefit. Types of Research Report I Any research report contains: descriptions on methodology, results obtained, and recommendations made. The basic orientation of a research report depends on its audience. Before writing the report the researcher must know his or her audience; he/she may have to make assumptions about the composition, background and interests of the target readers. Types of Research Report II Two types of reports: Technical Report: suitable for a target audience of researchers, research managers or other people familiar with and interested in the technicalities such as research design, sampling methods, statistical details etc., Popular Report: suitable for:

a more general audience, interested mainly in the research findings as it is non-technical in nature. The writing style is designed to facilitate easy and rapid reading and understanding of the research findings and recommendations. REPORT PREPARATION AND PRESENTATION PROCESS

Problem Definition, Research Design and Methodology

Data Analysis

Pre-report Writing Activities

Interpretation of Research findings

Report Preparation

Report Writing Activities

Oral Presentattion

RESEARCH FOLLOW-UP

Reading of the Report by the client

Post Report Writing

The first step in the process involves..: the interpretation of the results of data analysis in light of: the marketing research problem investigated, and the research design and methodology followed.

The research report is a means of communication that can be understood, believed, trusted by everyone who are likely to be affected by the research, and acted upon by the decision maker. Before writing the report

the researcher should discuss: the major findings, conclusions, and recommendations with the key decision makers. necessary to ensure that the report meets the client's needs and is ultimately accepted.

The entire marketing research project: should be summarized in a single written report or in several reports addressed to different readers. should present the findings in such a way that they can be used directly as input into decision making. Oral Presentation

Generally, an oral presentation supplements the written report. The client should be given adequate time to read the report. If necessary, the researcher should assist the client in understanding the report, implementing the findings, undertaking further research, and evaluating the research process in retrospect. Report Format

No universally accepted standard format or style for research writing. Different researchers may prepare their reports differently. The personality, background, expertise, and responsibility of the researcher and those of the decision maker for whom the report is written interact to give each report a unique character. Report formats are likely to vary with the nature of the project itself. However, the research report closely resembles the steps of the marketing research process.

Most research reports include the following elements: I. II. III. IV. V. VI. VII. VIII. Title page Letter of transmittal Table of contents List of tables List of graphs List of appendices List of exhibits Executive summary a. Major findings b. Conclusions c. Recommendations IX. Introduction a. Background to the problem b. Statement of the problem X. Approach to the problem XI. Research design a. Type of research design b. Information needs c. Data collection from secondary sources d. Data collection from primary sources e. Scaling techniques f. Questionnaire development and pretesting g. Sampling techniques h. Field work XII. Data analysis a. Methodology b. Plan of data analysis XIII. Results XIV. Limitations and caveats XV. Conclusions and recommendations XVI. Appendix a. Questionnaires and forms b. Statistical output c. Lists

The results... may be presented in several chapters of the report. For example, a Malaysian researcher conducting a national survey, may perform the data analysis in two stages. First, he or she may analyze the overall national sample followed by nine separate analysis for each of the states. The results may then be presented in ten chapters [ one overall plus nine state based] instead of one. Report Writing I Effective report writing is an art. Some basic points to note in writing a report. Readers: The report should take into account the level of readers' technical sophistication, their interest in the project, ability to understand as well as the circumstances under which they will read the report and how they will use it.. Adherence to study objectives: A research report must show that the research objectives have been accomplished.. Easy to follow: The most basic characteristic of a good report is that it is easy to follow. It should be well organized, logically structured, and clearly and lucidly written. Headings and sub-headings should be used for different topics and subtopics respectively.. Report Writing II Objective: Report writing should always be guided by objectivity. Should accurately present the methodology, results, and conclusions of the project, without slanting the findings to conform to the expectations of management. Selectivity: A researcher must use his or her discretion in deciding what should be included in the report. Concise: A report should be concise. Yet brevity should not be achieved at the expense of completeness. Presentation: The report should be professionally done with quality paper, good typing, and attractive binding.

Report Writing III Visual aids: Key information presented in the text of a research report should be reinforced with tables, graphs, pictures, maps, and other visual devices. Guidelines for Tables: Statistical tables are a vital part of the report and deserve special attention. Every table should have a number and brief but clear title. Basis or unit of measurement should be clearly stated to facilitate understanding. The arrangement of the data item should emphasize the most important aspect of the data being presented. If necessary, explanations, comments etc. should be provided as footnotes. If the table presents secondary data, the source(s) must be cited clearly. Oral Presentation Should be carefully prepared keeping the audience in mind. A good presentation does not mean a lengthy presentation. Carefully selected visual aids such as graphs, tables, charts, maps etc. help presentation. However, Too many visual aids, particularly statistical tables, could often be boring and may not serve any purpose. During oral presentation, people may seek clarification. The speaker must be patient and should not show signs of anger or frustration. He or she should be natural, establish eye contact with the audience, and interact with them. Body language and descriptive gestures are also quite useful.

MARKET RESEARCH PROCESS

Problem Identification & Formulation

Research Design

Questionnaire

Data Collection

Design Validation

Pretest Analysis Sample design Data collection Descriptive Statistics Data Modeling

Data

Research Report

Generation Presentation

Data driven decisions Why do we need data? What can data tell us? How do we obtain data? How can we use data? How do we know if the data is clean - accurate, reliable, affordable, accessible, STATISTICAL TECHNIQUES DESCRIPTIVE STATISTICS Central Tendencies Measures of Variability STATISTICAL INFERENCE Univariate techniques Multivariate techniques

Descriptive statistics After cleaning up the data Frequency analysis Measures of location or central tendencies such as mean, median, mode are calculated Measures of variability such as the range, inter-quartile range, variance and standard deviation are calculated Skewness and Kurtosis of the data are estimated Hypothesis testing

Univariate techniques

multivariate techniques I

n dependent Samples T-Test to examine mean differences between two exclusive or independent groups on a ratio-scaled or interval-scaled measure. For example, if you want to determine whether males or females have stronger intentions to buy your product based on a survey in which you measured purchase intentions with a Likert-type scale youd run an independent samples t-test. The following are situations in which you may use an independent samples t-test: To determine whether current customers and noncustomers beliefs about your brand differ To determine whether higher-income or lower-income consumers are more loyal to your brand To determine whether people with active or inactive lifestyles view your ads more favorably Paired Samples T-Test to compare the means of non-independent or non-exclusive samples For example we would use a paired sample t- test to determine whether males and females differ in their pre-purchase versus post-purchase beliefs about the brand. Some more examples: To determine whether respondents differ in their attitudes toward print and television ads

To determine whether Caucasians or Hispanics respond more favorably to banner or pop-up ads To determine whether younger or older customers are more likely to dine in or order takeout from your restaurant

One-Way Analysis of Variance (ANOVA) to examine the mean differences in how respondents answer an interval or ratioscaled question (dependent variable). For ANOVA, respondents must be part of a demographic that has three or more independent groups, such as ethnicity (independent variable). This technique is referred to as a one-way analysis because it uses only one independent variable. Some examples: To determine how customers brand perceptions vary by service quality (specified as high, average, and low) To determine how customer loyalty varies by selling price (specified as expensive, reasonable, and cheap) To determine whether marital status influences buying intentions To determine whether return policies (measured by easy, typical, and difficult) influence customers beliefs about your store

Linear Multiple Regression (LMR) to explain variance in a dependent variable with multiple independent variables With LMR, the dependent variable is interval- or ratio-scaled, and the independent variables can be of any scale type

For example, future sales of your product will depend on consumers disposable incomes, your marketing expenditures, competitors marketing expenditures, technological changes, and so on. In addition to predicting a dependent variable, LMR provides an R2 value, which denotes the percent of variance in the dependent variable explained by the independent variables; the more jointly predictive the independent variables, the larger the R2 value. Some examples: To identify marketing strategies that influence purchases by targeted customers To identify profit-maximizing marketing strategies To forecast future sales

Conjoint Analysis Conjoint analysis can determine how people value the different components of a good or service This approach requires respondents to indicate their relative preferences for different combinations of product attributes, viz., a utility score is used to express the relative value of each attribute level Example: Toothpastes have different features, such as teeth whitening, breath freshening, and the like. Hotels offer different amenities, such as free continental breakfast, free newspaper, free phone calls, and pay-per-view television programming. To design the profit-maximizing version of a product, you must discover the monetary value that customers assign to each component.

Some more examples where conjoint analysis is used: To design a new product (provided it shares features with existing products) To fine-tune your marketing strategy, especially as it relates to pricing To identify groups of consumers who have similar product preferences To evaluate new products (For example, if a new brand is introduced, how might it compete against current and likely competitor brands?)

Exploratory Factor Analysis (EFA) to identify a reduced number of factors from a larger pool of measures EFA allows one to summarize the data efficiently by reducing the number of characteristics one needs to analyze. can be used to assess the reliability and validity of the survey questions Where is it used: To identify defining characteristics for grouping consumers To determine product evaluation and acceptance based on brand attitudes To understand media viewing habits of targeted consumers

Multidimensional Scaling (MDS) Multidimensional scaling (MDS) is a statistical technique for mapping objects based on respondents perceptions of either specific object attributes or overall object similarity MDS can help you identify the characteristics that consumers use to differentiate products or retailers. These visual depictions can help you to identify saturated and unaddressed niche markets, which in turn can inform your product positioning strategies Some more examples where conjoint analysis is used: To identify the product aspects that consumers use to evaluate different brands To develop new products that satisfy consumers currently unmet or unsatisfactorily met preferences To assess the efficacy of marketing strategy changes to shift consumers perceptions of your brand in a desired direction

Cluster Analysis to assign people or objects to groups so that intragroup similarity and intergroup dissimilarity are maximized The goal is to identify naturally occurring (rather than predefined) groupings of people or objects based on their characteristics. Unlike regression techniques, cluster analysis includes no predicted (dependent) or predictive (independent) variables Also, cluster analyses rely on computer algorithms rather than statistical formulae, so you can base groupings on nonmetric (for example, sex, race, and religion) and metric (for example, age and household size) data Some more examples where cluster analysis is used: To help select test markets To help identify new product opportunities To identify customer groups seeking certain combinations of product benefits (for example, toothpaste that freshens breath and whitens teeth)

Discriminant Analysis to determine which characteristics discriminate between two or more naturally occurring groups The goal is to assign new people, households, or businesses to a group to predict group membership based on characteristics that accurately assign known group members to the proper group Example: males versus females, or nonusers versus light users versus heavy users of a product Credit-scoring schemes rely on discriminant analysis. Lenders, such as banks and credit card companies, collect detailed customer profiles. Over time, some customers never miss payments, and other customers rarely make payments. By weighing and summing the characteristics that differentiate reliable from unreliable payers creating a model that generates predictive credit scores creditors can reduce their costs by denying new credit to likely financial deadbeats. Some more examples where discriminant analysis is used: To determine whether targeted customer groups view different media To identify the distinguishing characteristics of consumers who often, occasionally, and never respond to online solicitations (for example, e-mail and banner ads) To identify lifestyle preferences that differentiate loyal from non-loyal customers To identify demographic characteristics that differentiate highly price sensitive, somewhat price-sensitive, and price-insensitive customers

LOGISTIC REGRESSION to predict the probability of an event. It models a binary response (dichotomous) outcome variable (for example, will happen or will not happen), which is coded as 1 (probability of success ) or 0 (probability of failure 1 ) as predicted by continuous and/or dichotomous independent variables Logistic regression is comparable to discriminant analysis (with only two groups) and multiple regression. However, unlike these other methods, logistic regression requires no data distribution assumptions on the predictor variables; that is, the predictors dont have to be normally distributed, linearly related, or have equal variance in each group. Some more examples where logistic regression is used: To forecast which salespeople are likely to quit in the next six months To forecast who will purchase a home in the next 12 months To forecast the regular customers who will switch to a new provider